Addition of a miscible liquid, a non-solvent for the polymer
Wall polymer. Wall-forming polymers include acrylates, cellulose acetate, cellulose acetate butyrate, ethylcellulose, poly(lactic acid), poly(lactic-co-
glycolide), polystyrene, polyvinyl acetate, polyvinyl chloride and other polymers (Fong, 1988).
Control of agglomeration. Fong (1988) described a number of patents regarding agglomeration. Agglomeration of poly(lactic acid) microcapsules prepared from a dispersion of drug particles in a solution of the polymer using a non-solvent was minimized by conducting the phase separation at a temperature of — 65°C using a dry ice bath. The use of low temperatures made the wall of the microcapsules sufficiently firm during the phase separation process, so that adhesion of the microcapsules was avoided. Another technique for minimizing the agglomeration of microcapsules employs talc. During the addition of the non-solvent, talc, a mineral silicate is added to minimize the adhesion and coalescence of the microcapsules. It is suggested that the talc forms a barrier against adhesion between the microcapsules. Talc has been used to minimize agglomeration in other patents. Polyiso-butylene has been used to minimize aggregation of microcapsules prepared from Eudragit RS (Chun and Shin, 1988) and from Eudragit RS100 (Chattaraj etal., 1991).
Polymer solvent. The polymer solvent must be miscible with the non-solvent and should not dissolve the core. The choice of a solvent for a particular polymer can have an effect on the product and its properties. A methacry-late polymer, Eudragit RLPM, has been used to encapsulate riboflavin. When the polymer is dissolved in benzene and then treated with petroleum ether as the non-solvent, phase separation occurs, with the result that large polymeric droplets are formed which adsorb on the surface of the vitamin as a thick, uniform coat. The product provides a slow release of riboflavin. This is in contrast to the use of isopropanol as the solvent which, after treatment with non-solvent, provides smaller coacervate droplets, a thinner coat and faster release (El Sayed etal., 1982).
Non-solvent. The miscible non-solvent should effect phase separation of the polymer and not dissolve the core material. The non-solvent should be easily removed by evaporation or by rinsing with a volatile solvent possessing similar properties to the non-solvent. Both polar and non-polar non-solvents have been used in the formation of microcapsules. Fong (1988) preferred polar non-solvents such as isopropanol and isobutanol to nonpolar, non-solvents such as heptane in low temperature microencapsulation. He found that a combination of non-solvents, e.g. propylene glycol and isopropanol, produces larger microcapsules (100-125 /¿m) than those prepared from isopropanol alone (25-50 /¿m). In some cases, it is easier to control the wall thickness when the appropriate non-solvent is used to prepare polystyrene microcapsules (Iso etal., 1985a,b).
Core. Generally, the core should be insoluble in both the solvent for the polymer and the non-solvent. A wide variety of cores have been encapsulated by this method, including antibacterial and anticancer agents, steroids, vitamins, antacids, and pharmaceuticals (Fong, 1988). Products with two walls have also been encapsulated (Hiestand, 1966). Methods have been used to encapsulate cores which are soluble in the solvent. For example, soluble thioridazine as the free base was soluble in the polymer solvent, but after converting to the pamoate salt it was insoluble in both the polymer solvent and the non-solvent. Another method for encapsulating a soluble core is to begin phase separation of the polymer before adding the core particles. As an example, after a solution of styrene maleic acid copolymer in ethanol was prepared, the non-solvent isopropyl ether was added until turbidity was first noticed. The drug methylprednisone was then added and the rest of the isopropyl ether was added to complete the process (Fong, 1988).
The non-solvent method of inducing phase separation has been used to prepare products which have two polymers to alter the release of the core. Itoh and Nakano (1980) coated matrix particles composed of an evaporated product of drug and cellulose acetate with ethylcellulose. In a patent, Fong (1988) describes the preparation of microcapsules with a double wall of polylactic acid prepared by essentially repeating the process.
Wall polymer. The polymer selected for this process must have a low solubility at room temperature and a high solubility at elevated temperature where it is completely dissolved. Few polymers possess this property, for example, ethylcellulose, hydroxyethylcellulose, hydroxypropylmethylcellu-lose, methylcellulose (Fong, 1988). The molecular weight of the ethylcellulose affects some of the properties of the final product. Deasy etal. (1980) showed that a higher molecular weight of ethylcellulose (100 cp, 0.1 Pa s) gave finer microcapsules and slower drug dissolution than those microcapsules prepared with lower molecular weight ethylcellulose (10cp, 0.01 Pas).
Control of agglomeration. Koida etal. (1983) found that the agglomeration of microcapsules was affected by the molecular weight. For example, microcapsules in the 149-250 fim size range increased as the molecular weight of the ethylcellulose increased. Agglomeration of microcapsules can be minimized by vigorous agitation, slow cooling near the coacervation temperature and washing with cold solvent (Deasy etal., 1980). Several aliphatic solvents at low temperature, 10°C, have been used to minimize aggregation: pentane, hexane or octane and also cyclohexane (Morse etal., 1978). The addition of a polymer for the purpose of minimizing aggregation has been investigated. Both butyl rubber and polyethylene have been investigated, however, polyisobutylene has been studied more intensely. Samejima etal. (1982) indicated that polyisobutylene prevented aggregation of microcapsules and found it to be equally as effective as butyl rubber and much better than polyethylene. Donbrow and Benita (1977) and Benita and Donbrow (1980) indicated the beneficial effects of polyisobutylene in preventing aggregation and suggested that it was adsorbed onto the surface of the ethylcellulose droplet, probably functioning as a protective colloid. Ethylene vinyl acetate copolymer has also been shown to alter the particle size and perhaps the aggregation of ethylcellulose microcapsules (Lin etal., 1985).
Control of particle size. The particle size of the coacervate drops was shown to decrease as the viscosity of the medium increased as a result of higher concentrations of polyisobutylene in the preparation of ethylcellulose microcapsules in cyclohexane by temperature change (Benita and Donbrow, 1980).
Core. Fong (1988) has provided a list of cores that have been encapsulated by this method; these include aspirin, vitamin C, sodium salicylate, chloramphenicol, isoniazid, phenethicillin potassium, sodium phenobarbital, niacinamide and riboflavin. The core compounds should have a low solubility in the solvent at the coacervation temperature. In addition, the compounds should be stable at the temperature employed. Koida etal. (1986) found that the efficacy of encapsulation was also related to low solubility of the core material.
Addition of an incompatible or non-wall-forming polymer
Incompatible polymers. The incompatible polymer is chosen on the basis of its higher solubility in the solvent than the coating polymer, thus there will be a tendency for the coating polymer to come out of solution first and coat the core, resulting in a core that is surrounded by one polymer only. Polymers that have been used are frequently low molecular weight liquids, mainly polybutadiene, methacrylic polymer and polydimethyl siloxane. It has been indicated that the advantage of using an incompatible polymer is that certain properties of the coacervation phase, namely the viscosity and relative volume, can be controlled. If the viscosity of the coacervation phase is too high, proper coating of the core cannot occur (Fong, 1988). Extensive studies on the system poly(DL-lactic-co-glycolic acid) copolymer dissolved in methylene chloride using silicone oil have been made by Ruiz etal. (1989).
Wall polymers. Some polymers used to form the walls are ethylcellulose, polymethylmethacrylate, polystyrene and poly (lactic-co-glycolide). The wall material may be hardened by adding a non-solvent for the polymer, for example microcapsules of methylene blue hydrochloride prepared from the wall polymer ethylcellulose dissolved in toluene employing the incompatible polymer polybutadiene were hardened by treatment with hexane. Other techniques include solvent evaporation and chemical cross-linking (Fong, 1988).
Control of agglomeration. It has been suggested that agglomeration can be minimized when solvent evaporation is used to harden the wall by using excess liquid paraffin and/or low temperatures during the microencapsulation process (Fong, 1988).
Solvent. The solvent must dissolve the wall polymer, the incompatible polymer and it should also be miscible with the washing non-solvent and should not dissolve the core.
Washing solvent. The function of the washing solvent is to remove the polymer solvent from the microcapsules and any incompatible polymer. Consequently, the washing solvent should be miscible with the polymer solvent, dissolve the incompatible polymer and not dissolve the core material or the polymer wall material.
Core material. Some examples of core materials encapsulated by this method include antibiotics, pharmaceuticals, polypeptides and dyes (Fong, 1988).
Evaporation with a miscible liquid, a non-solvent for the polymer
Wall polymer. Polymers include primarily ethylcellulose, polyethylene, ethylene acrylic copolymers and vinyl polymers (Fong, 1988).
Polymer solvent. The polymer solvent should have a relatively high vapour pressure so that it is readily evaporated. Polymer solvents include aliphatic and aromatic hydrocarbons, ketones, ethers, alcohols and esters.
Miscible liquid, a non-solvent for the polymer. The miscible liquid should not dissolve the core or the polymer, but should be miscible in the concentration used with the polymer and solvent, thus at the beginning of the process the mixture is homogeneous. Furthermore, this liquid should have a low vapour pressure so that during evaporation it will function as the suspending liquid. Examples of suspending liquids include hydrocarbons with a high boiling point, silicone fluid and polyethylene glycols (Fong, 1988).
Control of agglomeration. Moldenhauer and Nairn (1990) indicated that polyisobutylene has other effects, in addition to its protective colloidal action, namely, increasing the viscosity of the system, thereby suspending the core more uniformly, especially larger particle sizes, and decreasing the rate of evaporation, thereby the rate of surface nucleation of the polymer. It has also been suggested that the use of the suspending medium, light liquid paraffin, in this case actually initiates the coacervation process by removing some of the solvent from the polymer.
Coat structure. The rate of evaporation has an effect on the coat structure and thereby the rate of release of the microencapsulated drug (Moldenhauer and Nairn, 1991). Intermediate rates of evaporation provide a dense coat of uniform thickness and a smooth surface, whereas fast evaporation rates produce an irregular, smooth, porous coat and slow rates of evaporation produce a sponge-like coat.
Evaporation with an immiscible polar liquid, a non-solvent for the polymer
Core. The most important factor in selecting a core material is that it has a low solubility in the polar liquid, usually water, otherwise some of the core will partition into the aqueous external phase. A number of cores have been encapsulated, as listed by Fong (1988), such as antibiotics, antineoplastics, anaesthetics, insulin, steroids and other pharmaceuticals. Water-soluble drugs are generally not successfully encapsulated by this method; for example, salicylic acid, theophylline or caffeine could not be encapsulated with polylactic acid from a preparation of the drug in methylene chloride (Bodmeier and McGinity, 1987a).
Several methods have been used to encapsulate water-soluble or partially water-soluble drugs. Weak bases or weak acids in their salt form may be converted to their non-ionic form, thereby reducing their solubility. Chemical modification of a compound can be used to reduce its water solubility, thereby making it easier to encapsulate by this method. The addition of an inorganic salt to the aqueous phase will reduce the solubility of the core. Alternatively, some of the core can be added to the external aqueous phase in order to decrease the partition of the drug from the core to the external aqueous phase. For example, the addition of tetracaine to the non-solvent more than doubled the drug content of the microspheres (Wakiyama eta/., 1981). Other examples of loading in the external aqueous phase with drug in order to obtain a high drug content in the microcapsule include a saturated solution of quinidine sulfate (Bodmeier and McGinity, 1987b), and cisplatin (Spenlehauer etal., 1988).
The solubility of the core in the solvent for the polymer will have an effect on the nature of the final product. If the core material is soluble in the polymer solvent, then the encapsulated product will tend to have a homogeneous structure, as both the core and the polymer will come out of solution as the solvent is evaporated. If the core material is insoluble in the polymer solution, then thought should be given to the particle size before beginning the encapsulation process and milling or micronization should be considered. When the polymer comes out of solution, it will surround the core particles, leading to a heterogeneous product. Finally, large cores have been encapsulated by this method with the result that a membrane covers the drug. This type of product is different from the two types described above which are either homogeneous or heterogeneous in nature. The rate of release from the single core will tend to be constant while that from the monolithic type of microcapsule will tend to decrease with time. Blank microcapsules, that is microcapsules without a core, may also be prepared (Fong, 1988).
The core loading will affect the ratio of polymer to core, the size of the microcapsule, and the rate of release of the core material. The rate of release of dibucaine (Wakiyama etal., 1982), butamben, tetracaine (Wakiyama etal., 1981), ketotifen, and hydrocortisone acetate (Fong, 1988) all increase as the core loading increases. The maximum fraction of core loading depends upon the properties of the microencapsulation system, but may range from 0.4 to 0.75; for example, thioridazine and ketotifen were encapsulated at a fraction of 0.5 to 0.6 (Fong, 1988).
As mentioned in the process section, aqueous solutions have been encapsulated with considerable success, leading to a w/o/w system; however, a water-soluble compound will tend to diffuse into the outer aqueous phase. The loss of water to the external aqueous phase is reduced by using humectants such as glycerin (Fong, 1988). Gelatin has also been used as an internal stabilizer (Kondo, 1979a).
Wall polymers. The polymer selected for this process must be insoluble in water. Some examples include ethylcellulose, polystyrene and cellulose acetate butyrate which are used to prepare microcapsules. A number of biodegradable polymers have been used to prepare microspheres of pharmaceuticals; these include homopolymers and copolymers of lactic acid, glycolic acid, 0-hydroxybutyric acid and caprolactam (Fong, 1988). This phenomenon occurs with poly(DL-lactide) (Spenlehauer et al., 1986). Generally, particle size increases with polymer concentration.
The concentration of the wall-forming polymer solution, that is the polymer to solvent ratio, has an effect on the in vitro release rate of the core material. The release of the core material decreased when the initial concentration of the polymer solution was increased. The significance of this factor depends upon the drug that is used as the core. When thioridazine was used as the core material, the effect was appreciable as the initial polymer concentration in the process was raised from 5 to 10%.
However, only a small change was noticed when the core was hydrocortisone acetate. It has been suggested that the formation of homogeneous microcapsules containing thioridazine, which is soluble in the polymer solution, was more readily affected by the initial polymer concentration in the solvent than for the hydrocortisone acetate which was not soluble in the polymer solution, and thus formed heterogeneous microcapsules (Fong et al., 1986).
Polymer solvents. The solvent for the wall-forming polymer should be immiscible or have only a low solubility in water. Its boiling point must be lower than that of water so that it will evaporate faster than the external phase water. A solvent frequently used in this microencapsulation process is methylene chloride because of its high vapour pressure and because it is a good solvent for many polymers. Methylene chloride is, however, toxic and considerable amounts can remain in the product even after drying. Weight losses of up to 3.5% were determined by thermogravimetric analysis and chlorine content analyses (Benoit etal., 1986). Other solvents for polymers include chloroform, carbon tetrachloride, ethylene chloride, ethyl ether, benzene and methyl acetate (Fong, 1988).
Aqueous phase. The solubility of the polymer solvent in the aqueous phase has been shown to have a significant effect on drug loading. A study of solvent effects on the entrapment of quinidine sulfate showed that high loading was achieved with the solvent methylene chloride, which had the greatest water solubility, whereas very poor loading was achieved using chloroform, which has a lower water solubility. It is suggested that solvents with high water solubility effect rapid deposition of polymer at the droplet interface, creating a barrier at the interface, thus minimizing drug diffusion out of the microsphere to the outer phase water. Alternatively, if water-miscible polymer solvents are used to dissolve the drug and polymer, agglomerates of polymer are formed on mixing. Mixtures of polymer solvents with different water solubilities can be used to obtain microspheres (Bodmeier and McGinity, 1988).
Surfactants and emulsifying agents. The emulsifying agent should be selected so that an emulsion of the appropriate particle size is readily formed, and it stabilizes the emulsion during removal of the volatile polymer solvent preventing coalescence of the droplets. As this method of phase separation involves the formation of an o/w emulsion, the proper HLB (hydrophile-lipophile balance) value is 8 to 18. The specific emulsifying agent and its concentration should be determined by trial and error.
Salts of fatty acids, particularly sodium or potassium oleate, have been found to be useful for the preparation of microcapsules, including polymers that are subject to biodégradation. As an example, sodium oleate produced high yields of biodegradable microspheres which were free of agglomeration. The fraction of the drug incorporated was 80-99%, and core loading was up to 0.5 of the weight of the microsphere. The size of the microspheres was less than 150 /xm diameter, which will pass through a 20 gauge needle (Fong et al., 1986). It is necessary to consider the pH of the aqueous phase if the surface active agents are subject to different degrees of ionization as a result of different pAT values. It will be necessary to maintain a pH at least two to three units above the pK^ of the fatty acid in order for it to be properly ionized.
Surface-active agents, both anionic and non-ionic with an HLB of at least 10 at a concentration of 0.1-1% have been used to prepare microcapsules by this method (Morishita etal., 1976). Watts et al. (1990) have listed a number of surface-active agents: polysorbate 80, sodium oleate and sodium dodecyl sulfate. The use of polysorbate 80 in the aqueous phase at a concentration of 2% produced a small reduction in the content of quinidine in the microcapsules. This was attributed to not only an increased solubility of the drug in the aqueous phase, but also to stabilization of the polymer droplet interface which reduced the rate of solvent loss, thereby reducing the polymer deposition rate and permitting a greater loss of drug from the partially formed microcapsules before a suitable hardened barrier could be formed (Bodmeier and McGinity, 1987b). Low yields of small microcapsules may be obtained if surface-active agents are used, e.g. sodium lauryl sulfate (Jaffe, 1981).
Emulsifiers such as gelatin and polyvinyl alcohol at a concentration of 0.5-2.0% may be used to form o/w emulsions (Morishita etal., 1976). Emulsifiers have other effects on the preparation of microspheres as a result of enhanced solubilization. Lomustine and progesterone crystals were formed on the microsphere surface and in the aqueous phase as a result of using polyvinyl alcohol and methylcellulose. The crystals were eliminated and drug loading improved when the emulsifier was removed half way through the evaporation step (Benita etal., 1984). The use of emulsifiers can alter the rate of release of drugs from the microsphere. For example, the use of a gelatin solution which provides a lower solubility for insulin showed only a 26% burst effect, compared with a solution of polyvinyl alcohol which provides a higher solubility for the insulin, resulting in an 88% burst effect. The difference in the burst effect has been attributed to the difference in the solubility of insulin in the hydrophilic colloidal solution (Kwong etal., 1986).
Wakiyama etal. (1982) investigated the effect of acid-processed gelatin and alkaline-processed gelatin as an emulsifying agent on the yield and efficiency of microencapsulation of basic amino drugs. A greater efficiency of drug incorporation was achieved when alkaline-processed gelatin was used, perhaps owing to the fact that alkaline-processed gelatin gave a pH of 7.5, promoting the formation of the non-ionized form of dibucaine (pA"a 1.6 and 8.3) and thus its greater uptake by the solvent as a result of the greater o/w partition coefficient. When the pH of the aqueous phase was raised to 8.6, a greater fraction of the dibucaine was in the non-ionized form and a greater incorporation of drug was observed.
The use of hydrophilic colloids to stabilize the emulsion may have an effect on the shape or particle size of the final product. For example, the use of methylcellulose 400 and polyvinyl alcohol as a stabilizer for the external aqueous phase resulted in oval-shaped microcapsules; most of the products were, however, spherical (Cavalier et al., 1986). In other experiments microsphere size was dependent on the type and concentration of the emulsifying agent; for example, microsphere size increased with increasing polyvinyl alcohol concentration (Benita et al. 1984). Other researchers have obtained smaller microspheres with a 1 °7o sodium alginate solution which had a higher viscosity than a 1 or 2% gelatin solution (Wakiyama et al., 1981; Kojima etal., 1984). In some cases high concentrations of gelatin decreased aggregation (Wakiyama etal., 1982).
Solvent evaporation. Solvent evaporation may be accomplished by stirring the emulsion in an apparatus where the surface is exposed to air. Forced air or nitrogen may be used to promote a more rapid evaporation rate. Heat and reduced pressure may also be used but should be controlled at such a rate that microcapsules with a smooth surface are obtained if desired. Heat and low pressure may cause foaming of the emulsion system which should be avoided, particularly at the early stages of phase separation (Fong, 1988).
Stirring rate. After the emulsion containing the particles with appropriate size has been formed and before evaporation has begun, the stirring rate should be such that there is minimum aggregation of the droplets until the microspheres are hardened. The main factors that control the particle size are speed, equipment and the concentration of the polymer in the dispersed phase and the concentration of the hydrophilic polymer or surfactant in the aqueous phase. Particle size tends to decrease and the size range is narrowed as the mixing speed increases (Benita etal., 1984). Stirring speeds of 800-1600 r.p.m. have been used when either gelatin or polyvinyl alcohol have been used as the emulsifier. At 900 r.p.m., small holes were observed in the microspheres, while at 500r.p.m. they were absent (Nozawa and Higashide, 1978).
Reactor design. The use of baffles minimizes the vortex which can lead to microsphere aggregation and, in addition, droplet breakup occurs with the result that the average size of the microcapsules is decreased and the microsphere yield is increased (Bodmeier and McGinity, 1987c).
Evaporation or removal with an immiscible organic liquid, a non-solvent for the polymer
Core. The core should have a minimal solubility in the immiscible organic liquid in order that most of the core is encapsulated. Tartrazine has been encapsulated with cellulose acetate trimellitate (Sanghvi and Nairn, 1991) when evaporation of the polymer solvent is not permitted. When evaporation is allowed to proceed, several pharmaceuticals have been encapsulated: tetracycline, loperamide, metoclopramide, hydrochloride and also biological material (Maharaj etal., 1984) and drug-resin complexes (Sprockel and Price, 1990).
Wall polymers. The polymers used in this process should not dissolve in the suspending medium, for example, ethylcellulose dissolved in acetone was dispersed in a non-volatile hydrocarbon liquid (Dispersol 81515) (Yoshida, 1972). Another example is cellulose acetate phthalate dissolved in a mixture of acetone and ethanol, 95%, and then added to mineral oil (Beyger and Nairn, 1986). The size of the microcapsules increased from an average diameter of 140 ftm to 295 ¿¿m when the concentration of the polymer, Eudragit RS was increased two and a half times. The reasons for this increase in size are attributed to the increase in viscosity of the dispersed phase as a result of higher polymer concentration and an increase of polymer inside the droplets affecting a larger volume (Pongpaibul etal., 1984).
Polymer solvents. In order to avoid evaporation, the use of polymer solvent should be selected so that it has some solubility in the immiscible organic liquid, thus avoiding the use of temperature and the destruction of heat labile drugs. The removal of acetone, which has limited solubility in mineral oil, effects the phase separation of the polymer cellulose acetate trimellitate and subsequently microcapsules are formed (Sanghvi and Nairn, 1991, 1992).
Surfactants. The use of surfactants with low HLB values increases the region of the phase diagram where microcapsules were formed. The use of these surfactants tends to give a smooth surface on the microcapsules and at 3% concentration gives smaller microcapsules. Surfactants with a higher hydrophilic lipophilic balance value decrease the region on the triangular phase where microcapsules could be produced (Sanghvi and Nairn, 1991).
Immiscible organic liquid. Mineral oil is used extensively in this process of phase separation. It has been used in preparation of microcapsules with the following polymers: polymethyl methacrylate (Sprockel and Price, 1990), cellulose acetate phthalate (Beyger and Nairn, 1986), and ethylcellulose (Kaeser-Liard etal., 1984).
This section is concerned with some physical and chemical parameters, mechanisms and theories and/or experimental evidence which support the various concepts for coacervation-phase separation and deposition of the coacervate onto the core. As a result, this section is split into three parts:
1. A single wall-forming polymer soluble in water
2. Two wall-forming polymers soluble in water
3. A single wall-forming polymer soluble in an organic liquid
A single wall-forming polymer soluble in water
In order to prepare a satisfactory solution of a water-soluble polymer, it is necessary to disperse the polymer in water and allow it to become fully hydrated, perhaps using appropriate temperature conditions, addition of a small amount of non-solvent and/or mechanical means.
Coacervate-phase separation is induced by a number of techniques such as addition of a water miscible solvent, the non-solvent, which is not a solvent for the polymer, or a salt that binds a considerable amount of water thus removing it from the polymer. This process can best be illustrated by three component phase diagrams. The components usually are polymer, solvent and the agent which effects coacervation such as salt or ethanol. An example is given in Fig. 2 for gelatin, water and ethanol (Nixon etal., 1966). It can be seen that a dilute solution of gelatin in water, say 10%, will form a single phase. As ethanol is added, the two-phase region is encountered and the polymer-rich phase, the coacervate, is produced which encapsulates the core material if present. The polymer-poor phase functions as the medium to suspend both the coacervate and core which is being encapsulated. At pH values distant from the isoionic point, flocculation occurs in the presence of ethanol because the gelatin is fully stretched and is unable to entrap the occlusion liquid. The flocculation region is generally not satisfactory for encapsulation. The concentrations in the various regions in this system were studied by using refractive index and specific gravity (Khalil etal., 1968; Nixon etal., 1966). Rigidization of the coat may
Fig. 2 The composition of coacervate and corresponding equilibrium liquid. O Coacérvate; • equilibrium liquid; A total mixture. Reproduced with permission from Nixon eta/. (1966), /. Pharm. Pharmacol. 18, 409-416. The Royal Pharmaceutical Society of Great Britain, London.
Fig. 2 The composition of coacervate and corresponding equilibrium liquid. O Coacérvate; • equilibrium liquid; A total mixture. Reproduced with permission from Nixon eta/. (1966), /. Pharm. Pharmacol. 18, 409-416. The Royal Pharmaceutical Society of Great Britain, London.
be effected by temperature change, use of a cross-linking agent and/or the use of appropriate non-solvent for the coat.
Phase diagrams are also useful for preparing nanoparticles which are in the nanometre size range. For example, one method of preparing nanoparticles from gelatin or albumin is to desolvate the polymer with a salt which is highly hydrated, thus causing the coacervate to form. Then the protein is just resolvated with a small amount of water or isopropanol. In this procedure the phase diagrams are prepared using light scattering to measure the onset of coacervation and thus the appropriate conditions to achieve nanoencapsulation. The nanoparticles can be rigidized with a suitable cross-linking agent such as glutaraldehyde (Oppenheim, 1986). Thus, phase diagrams are useful for preparing nanoparticles which form in the precoa-cervation region, microcapsules or microspheres which separate in the coacervation region, and purification of the protein in the flocculation region when excess salt is added to the system (Oppenheim, 1986).
Khalil etal. (1968) investigated the role of pH in the coacervation of gelatin. Since gelatin exists as a randomly coiled configuration in solution, the shape of these coils is influenced, by the ionization of the acidic and basic groups. A stretched configuration is predominant when the gelatin is mainly in the anionic or cationic form. A random coiled structure is favoured at the isoionic point as a result of inter- and intra-molecular attractive forces. The role of pH as it affects coacervation was explained by two factors which appear to influence coacervation of polyelectrolyte systems, namely inter- and intra-molecular attractive coulombic forces, and hydration. The authors postulate that the first of these effects favours phase separation and formation of floccules while the second promotes redispersion of the molecular species. A proper balance between these factors promotes the formation of a colloid-rich isotropic liquid phase which is the coacervate. At the isoionic point there is a balance between the attractive forces of the oppositely charged sites and the hydration effect. The authors found that at the isoionic point, a coacervate was readily obtained when ethanol was added. As pH values move away from the isoionic point, attractive forces decrease and hydration of the gelatin increases; both of these tend to prevent coacervation. At pH values considerably different from the isoionic point, flocculation occurs upon addition of ethanol as the molecule is stretched and cannot entrap sufficient water. At intermediate pH values, a viscous gel is formed. There is more flexibility in the gelatin chain but insufficient water is entrapped to form a coacervate.
When sodium sulfate is used as the coacervating agent, the ions in solution shield the charges on the gelatin material and thus the forces of repulsion are modified by the added salt. At pH values on the acidic side of the isoionic point of gelatin the sulfate ions are associated with the positively charged groups and coacervation proceeds satisfactorily.
Madan (1978) has suggested that deposition of the polymer onto a core may take place as a result of:
1. Molecular interaction between the colloidal macromolecular particles,
2. Coacervate droplets may coalesce about the core particles,
3. Single droplets may encompass one, or a group of core particles.
An examination of the surface characteristics of microcapsules of gelatin, prepared by coacervation using ethanol or Na2S04, showed that the dried microcapsules had no cracks or fissures. The smoother surfaces of the microcapsules produced using ethanol showed marked surface folding, attributed to vacuole formation in the coacervate droplet, which increased with time, allowing for the formation of the coacervate coat. As a result of these observations, Nixon and Matthews (1976) proposed that microcapsules prepared by coacervation resulted from the merger of several smaller microcapsules.
Several microencapsulation procedures involve the encapsulation of immiscible liquids which may, or may not, contain a drug dissolved, dispersed or emulsified in the liquid. In order for the process of microencapsulation to proceed properly, the coacervate must engulf the liquid drop and then be hardened. In this system there are three immiscible phases present - the liquid core, the coacervate and the polymer-poor phase. Torza and Mason (1970), both theoretically and experimentally, investigated the interfacial phenomenon of systems that contain three liquids which are immiscible and indicated the spreading coefficient necessary for a coacervate droplet (liquid 3) to surround the core liquid (liquid 1) when both are in an immiscible continuous phase, the polymer-poor phase (liquid 2). The three spreading coefficients for the three-phase system are:
In order to consider the process it is assumed that <t12> a23 thus S¡ < 0, and thus only three possible sets of values of S exist. These correspond to complete engulfing, partial engulfing and non-engulfing of the liquid core. Complete engulfing occurs under the condition that Sj <0, S2 < 0, S3 > 0. As the interfacial tension between the polymer-rich phase, that is the coacervate, and the polymer-poor phase is low and if the interfacial tension between the core and the continuous medium is higher than between the core and the coacervate-rich phase, then the above requirements are satisfied and engulfing will occur. Torza and Mason (1970) conducted a number of experiments using different liquids and found that of 20 systems studied, only five did not correspond with the theory. The method of engulfing was determined with a high-speed movie camera and was shown to involve two competitive processes - spreading and penetration. This theory may apply to any coacervate system involving liquids.
In a study of the microencapsulation of oil droplets with or without a drug, clofibrate or chlormethiazole, using gelatin with an isoelectric point of 4.85 in the presence and absence of a surface-active agent, Siddiqui and Taylor (1983) were able to relate the ionic charge on the coacervate, which was negative, and the spreading coefficient of the liquid substrate. The surface active agents cetrimide, sodium lauryl sulfate or a double salt hexadecyltrimethylammonium lauryl sulfate were used in the process. Spreading coefficients calculated from interfacial tension values indicate that the coacervate should spread more easily in the presence of the double salt, and less so in the presence of either of the other two surfactants. It was noted that conditions for measuring the spreading coefficient and microencapsulation were not identical. Measurements of the charge size on the dispersed particles showed that the oil droplets and coacervate droplets can be expected to have opposite charges, except in the presence of sodium lauryl sulfate. Microencapsulation was satisfactory with both the double salt and cetrimide, but not with sodium lauryl sulfate. The authors suggest that some cetrimide may enter the coacervate phase and subsequently microcapsules tend to agglomerate. Addition of sodium lauryl sulfate at this stage tended to prevent agglomeration. The authors suggest that the use of the double salt enhances the attachment of the coacervate to the oil droplet surface. In fact, this double salt produces the smoothest microcapsule and permits a slower release of the drug from the oil droplets.
By electrophoresis and adsorption studies of gelatin and various core materials using six different kinds of coacervating agents, Okada etal. (1985a) were able to show that suitable encapsulation by gelatin is affected by the affinity between the core material and the coacervate phase. If a large amount of gelatin is adsorbed prior to coacervation, then encapsulation is successful.
In a subsequent paper Okada etal. (1985b) showed that carboquone could not be encapsulated with gelatin unless methanol or sodium sulfate solution was used as the coacervate inducing agent. If, however, the drug is recrystallized from a solution of an ionic polymer and if the pH of the solution is appropriate, then the drug can be encapsulated using many coacervating inducing agents. It was shown that the electrostatic attraction found between the gelatin and the polymer attached to the drug has an important role.
Instead of using alcohols of low molecular weight to cause coacervation, Jizomoto (1985) used polyethylene glycol or polyethylene oxide. Both these polymers have the same general structural formula, but the former name is usually used for polymers with molecular weights < 20 000 and the latter for those greater than tens of thousands. Polyethylene oxide or polyethylene glycol was added to aqueous gelatin 2.2% with stirring at 40° C at various pH values to influence phase separation. The phase diagram shows that the addition of a small quantity of polyethylene oxide or polyethylene glycol causes phase separation over a wider pH range and this phenomenon is largely dependent on molecular weight. A plot of the log of minimum concentration of polyethylene glycol or polyethylene oxide required to effect phase separation against the log of molecular weight at pH 8.7 gives a straight line. Elemental analysis of the coacervate is identical with that of gelatin; thus the phase separation induced by the addition of polyethylene oxide or polyethylene glycol is caused by an incompatibility. After reviewing the theory proposed by Bailey and Callard (1959) which suggests that water molecules are orientated with respect to polymer chain and the postulation of Kagemoto etal. (1967) that the parameter ascribed to the interaction between the oxygen atoms of the ether bonding in polyethylene oxide chain and the water molecules is proportional to a function of the molecular weight, Jizomoto suggests that the effect of polyethylene oxide or polyethylene glycol on phase separation in relation to molecular weight cannot be explained by dehydration. Blow and coworkers (1978) suggested that polyethylene oxide causes a decrease in 'free water' but in the present experiments, polyethylene oxide concentration to bring 'free water' to zero was not appreciably influenced by molecular weight. Therefore, the dependence of phase separation on molecular weight cannot be explained by the concept 'free water'. The author used a simplified equation of the chemical potential expressed in terms of molalities of two polymers, the solvent and the exclusion volumes as described by Edomond and Ogston (1968). Calculations were made to estimate the minimum concentration of the polymers required to cause phase separation. The author concludes that the excluded volume should make the main contribution to the induction of phase separation.
Was this article helpful?