The synthesis of semipermeable membrane microcapsules using in situ cyanoacrylate ester polymerization

5‘ Taylor&Francis J. MlCROENCAPSL1I.A’l‘lON, 2002, voe. 19, N0. 6, 699—724 gilealthsclences The synthesis of semipermeable membrane microcapsules using in situ cyanoacrylate ester polymerization R. HULL and R. P. CHANIBERS* Chemical Engineering Department, Auburn University, AL 368-$9, USA (Rereitied 9 August 1999; accepted 23 November 2000) A new material for the microencapsulation of bioiogical systems was discovered and characterized with regards to the eflects of reaction conditions on product yield. By using poly(cyanoacry!ate ester), membrane microcapsules were produced with suflicient strength and porosity to be effective in a process environment for the immobiiization and protection _of encapsulated material. After synthesizing numerous monomeric cyanoacrylate esters, the n-butyi derivative was discovered to give the best results with regards to microcapsule formation. Microcapsules were prepared by a dropiet technique in which an aqueous solution is sprayed into an organic solvent containing the cyanoacrylate ester monomer. By pre-treating the cyanoacrylate ester monomer with an anion exchange resin (Amberiyst A-21, Rohm and Haas, Philadelphia, USA), a significant reduction in the amount of acidic impurities which can adversel_v affect results was achieved. The use of polyvinyIpyrrolidone as a poiymerization initiator gave the best results of a variety of polymeric and n0n—p0iyn'1eric initiators investigated. Successful encapsulations were achieved using a solvent mixture of 60% (v/v) iso-octane/40“/u triehloroethyiene, 0.1% (w/v) polyvinyl- pyrroiidone initiator, pH 6.5 aqueous encapsulation solution, and 5% (vfv) methylcyanoacrylate/A-2I treated n-butylcyanoacryiate (added separately to solvent) made to a 4% (v/v) solution in solvent. Ester monomers were synthesized and used to prepare polymer membranes. Key-zvords: Microencapsulation, semipermeable membrane, biological immo- bilization, polymerization. Materials and methods Ail chemicals used in the synthesis of cyanoacrylate monomers were of reagent grade or better. Cyanoacetic acid and paraformaldehyde were obtained from Fisher Scientific (Pittsburgh, USA); piperidine, tricresyi phosphate, and hydro- quinone were obtained from Fisher Scientific (Pittsburgh, USA); and phosphorus pentoxide was obtained from Aiclrich Chemicai Company (lvlilwaukee, USA). In the formation of microcapsules, solvents used were of reagent grade or better and obtained from Fisher Scientific (Pittsburgh, USA), except decahydro- napthalene which was obtained from Aldrich Chemical Company (ivlilwatlkee, USA). Isobutyl cyanoacrylate, Tween 20 (polyoxyethylene sorbitan monolaurate), and polyvinylpyrrolidone (l\'I\V: 4-0000) were obtained from Fisher Scientific (Pittsburgh, USA). l\’lethoxypropyl cyanoacrylate was provided by Permabond Corporation (Bridgewater, NJ, USA). Amberlyst A-21 ion-exchange resin and *To whom correspondence should be addressed: jfana-'m:l' ofi1Iicraaaira,r:ru!mim1 ISSN 0}.‘65~20~i8 print,'lSSN H6-P5246 online it] 2002 Taylor 3' Francis Ltd hup:/,:‘www.tandf.co.u};/juurnals DOT: ]0.l0S0j02652040]10055225 700 R. 3’. Hall and R. P. Chambers poly (4--vinyipyridine) were purchased from Aidrich Chemical Company (1\*Iil- waukee, USA). Additional ion-exchange resins, A-1? and Dowex I-X8, were obtained from J. T. Baker Chemical Company (Phillipsburg, NJ, USA). Cyairoacrylnre ester sym‘h.esis Cyanoacrylate esters were prepared by the method outlined by Ieremias (1954). Cyanoacetate ester required by this method was synthesized by an acid catalyzed condensation reaction between cyanoacetic acid and the alcohol providing the ester side group. The reaction can be summarized as: CN CN I H’ 1 HC-COOH + RDH ¢'> I-IC - COOR + H20 To this end, 125 g cyanoacetic acid was placed in a I 1 reaction flask possessing a thermometer well. Approximately 250ml of alcohol (enough to dissolve all of the cyanoacetie acid) was added. VVhen ail of the cyanoacetic acid was dissolved, 5 ml concentrated sulphuric acid was added to the reaction mixture, and the flask connected to a totai reflux condenser. The mixture was heated and a vacuum applied until constant boiling was obtained. The vacuum was adjusted to give a boiling temperature of "’90°C. The mixture was allowed to reflux for more than ‘H1, at which time it was cooled and neutralized with saturated potassium carbonate. After filtering to remove precipitated sulphates, the reaction solution was allowed to stand overnight in a 500 ml separatory funnel. The formed organic phase was collected and vacuum distilled repeatedly to achieve a high purity product; the aqueous phase was discarded. Boiling points, densities, and yields of the synthesized cyanoacetate esters are given in table 1. Cyanoacrylate esters were prepared by a condensation reaction between for- maldehyde and the respective cyanoacetate ester. The cyanoacrylate ester mono- mers prepared in this manner polymerized spontaneously; therefore, a monomer was obtained by changing the reaction conditions to achieve depoiyinerization. The reaction can be summarized as: CN CN l l HC-COOR + I-IgCO qmmp. H2C=C-COOR + H20 CN CN ‘ I l nH;C=C - COOR "F? [ - CH; -C- ],, spent. | COOR CN CN 1 high T l [- CH; - C-].. ——-—-+ nH;C=C - COOR I vacuum COOR .S'y7rthes:'s of senn'pe1'mmble menzbrama nn'crompsm’es 701 Table 1. Boiling points, densities, and yields of various cyanoacetate esters. Ester side group Tb (pi, vacuum) Density (g/ml) Yieid (g) n-propyi 100°C (7-40mmHg) 1.01 70.6 lso-propyl 113 °c (740 mmHg) 1.01 97.1 Aliyi 113 “c (740 mmng) 1.05 64.4 n«buty1 E08 °c (750 rnmi-ig} 0.99 89.5 Iso—amyl 114- DC (745 mmHg) 0.97 1 16.8 n—he.\'yi 139 “C (740 mmHg) 0.97 78.9 Cyelohexyi 144 CC (750 mmHg) 1.05 84.0 n—octyl 161 0C (740 mmllg) 0.94 146.6 2-ethyl hexyl 105 °c (730mmHg) 0.94 72.5 n-decyl 196°C (740mmHg) 0.92 133.5 Benzyl 185 °c (740mmHg) 1.10 115.3 The amount of paraformaldehyde necessary to give “L5 X the moles of ester available was weighed and added to a three-neck 1 1 reaction flask to which a Dean- Stark trap and totai reflux condenser, a thermometer, and a 500 mi separatory funnel were attached. i\'Ietha11ol (lS0ml) was added through the Dean-Stark trap, and 0.5 ml piperidine to the reaction Flask. Heat was appiied until reflux was achieved. The cyanoacetate ester prepared by the aforementioned procedure was placed in the separatory funnel and added dropwise to the reaction mixture after reflux was achieved. The addition rate of the ester and the heating rate to the apparatus were adjusted to maintain a constant boiling condition in the reaction flask. After the addition was completed, 15ml tricresyl phosphate was added to the reaction flask, and the pH of the resulting mixture was tested to insure the solution was basic. If the solution was not basic (“pH 8), then piperidine was added sparingly until a proper pH was achieved. Since the polymerization of cyanoacry- late esters is anionic, a basic reaction solution insures a favourable depolymeriza- tion condition by stabilizing the anionic Species formed when the terminal group is removed. Heating rate \vas increased and methanol slowly siphoned ofl‘ the Dean- Stark trap. When the reaction temperature reached 80 DC and "100 ml ofmethanol collected, l65 mL benzene was slowly added to the reaction mixture. Azeotropic distiliation was continued until most of the water formed by the reaction and methanol were removed from the system by siphoning the Dean-Stark trap. Volumes of distillate collected were monitored to insure complete removal of a methanoi and water. Upon complete removal of the methanoi and water, the apparatus was converted to a long-path vacuum distillation unit by removal of the Dean-Stark trap/totai reflux condenser and the thermometer and connecting a gas addition tube and Clasien distillation head with a 300 mm \Vest condenser. After siightly cooling the reaction mixture, 5.6g phosphorus pentoxide and 4.2g hydroquinone were added and sulphur dioxide bled through the system. Benzene used in the azeotropic distiiiation was distilled until most of the volume added was collected. The apparatus was then converted to a short-path distillation unit with the connection of the vacuum adapter directly to the Claisen head. A long neck, 100 ml receiver, half—filled with glass beads and into which small amounts of phosphorus pcntoxide and hydroquinone were added, was connected to the vacuum adapter and chilled in an iced water bath. High vacuum (< 1-2 mm Hg) and heat was applied to the system until depolymerization of the cyanoacrylate 702 R. j’. Hell and R. P. Chambers Table 2. Depolymerization temperatures, boiling points, and yields of cyanoacrylate esters. Ester side group Tdp (pdp vacuum) T1, (pi, vacuum) Yield (ml) n-propyl 95 5C (760 mmHg) 85 DC (750 mmHg) 32 Iso-propyl 100°C (760rnmHg) 82°C (rsommng) 50 A113-1 150°C (760 mmHg) 94°C (750mmHg) 17 Iso—a1nyl +170:c (700 mmHg) 90:C (760 mmHg) 62b n-hexyl 175 C (760 mmHg)" 95 C (760mml-lg) 3 Cycloliexyl 175°C (760 mmHg) 92°C (760mmHg) 105 11-octyl (poor yield)” 129°C (700 mmHg) 5 2-ethyl hexyl (poor yield) 122 “C (7(:Umm1-lg) 4-1' n-decyl (poor yield)“ 152°C (760mrnHg) 4-D Benzyl (not successful) " Excessive foaming at high vacuum. 1’ Significant cyanoacetate ester contamination. Table 3. Retention times for various cyanoacrylate and cyanoacetate esters. Ester side group 1', (min), acrylate t, (min), acetate Iso-propyl 4.08 5.38 n-propyl 5.19 - Ally] 4.88 6.95 Is0~buty1 5.44 — n-butyl 6.51 —- lflethoxypropyl 7.39 — 1so—amyl 6.81 7.40 n—hexyl 7.11 8.70 Cyclohexyl 9.29 10.77 Propoxypropyl 8.03 ~A n-octyl 9.96 10.51 2-ethyl hexyl 7.83 9.37 n—decyl 10.88 12.75 ester polymer occurred. The monomer vapour was condensed and collected. Repeated Vacuum clistillations using sulphur dioxide bleed and phosphorus pent- oxide/hydroquinone additions to inhibit spontaneous polymerizations were per- formed until a high purity product was achieved. Depolymerization temperatures, boiling points, and yields are given in table 2. Purity was determined by gas chromatograpliy using a Varian 1\*Iodel 3700 Gas Chromatograph equipped with a 30111, 0.75 mm ID wide-bore capillary column of 1.0 micron film thickness Supelcowax 10. Injector temperature was set at 200 DC and flame ionization detector temperature was at 230 DC. Detector sensitivity was set at 10-9. Satisfactory separations were achieved using a temperature program starting at 120 DC and ending at 220 0C with a 10°C/min heating rate. One microliter of 10% solution of monomer in methylene chloride was used for analysis. Retention times for cyanoacetate and eyanoacrylate esters are given in table 3. .S'_vnt1zesr's of senzipermeable membrane r.m'cro.-rapszrles 703 1laiicroca,bs1t1e Formation Nlicrocapsules were formed primarily by the drop technique in which aqueous droplets were dispersed by syringe and needie into an organic solvent containing the cyanoacrylate monomer (see figure 1). The size of the microcapsules produced depended on the size of the needle used for dispersing and the polymerization rate of the particular reaction conditions. If a reaction condition produced stable microcapsules, then they were usually 1.0 mm in diameter. Depending on the experiment, the organic solvent was either pure hydrocarbon or a mixture of hydrocarbon and halogenated hydrocarbon. A satisfactory solvent system for the formation of n-butyl cyanoacrylate polymer membrane microcap- sules was found to be a mixture of 60% iso-octane and 40% trichloroethylene, although other solvent systems were tried. Solvent mixtures with densities approximately equal to 1.0 were preferred, since these were isopycnic with the aqueous droplets at the top or bottom of the test tube prevented, which could lead to a lack of uniformity in the forming microcapsule membrane. Various com- pounds were tested for their ability to initiate cyanoacrylate ester polymerizations. A 1% (V/v) solution of butyiamine in distilled water was chosen as a reference initiator. I\'Iore than likely, the cyanoacrylate ester monomer was contaminated with impurities such as cyanoacetate ester and alcohols left over from synthesis, no E Aqueous droplet Aqueous phase .«:< - polymer m — monomer I — initiator Interface Organic phase Figure 1. Schematic of the process used to produce po1y(cyanoacyiate ester) membrane microcapsules. Organic phase contained the cyanoacrylate ester monomer. The aqueous phase contained initiator and the material to be encapsulated. 704 R. j’. Ho]! and R. P. Clmmbers matter what the source. To further purify these salnples before their use in encapsulation, the monomers were treated with an ion-exchange resin to remove acidic impurities. A 31111 disposable syringe fitted with a 25mm Acrodis CR disposable filter assembly, PTFE, 0.45t1tn (Gelman Sciences lnc., Ann Arbor, MI, USA), was filled with Amberlyst A-21 ion—exchange resin. The amount of ion-exchange resin used in each purification depended on the viscosity and the degree of contamination of each monomer sample. Approximately 2ml of resin was used to treat lml of n-butyl cyanoacrylate and significantly improved the quality of the sample. The monomer must be used immediately after treatment for it polymerized rather quickly. In a typical encapsulation experiment, i00t1l of treated cyanoacrylate ester monomer were added to Sml of solvent. The monomer/solvent solution was immediately used to encapsulate 0.5ml of aqueous initiator solution dispersed into the organic phase by a 3 nil disposable syringe with a 21 gauge, 900 bevel cut needle. The mixture was allowed to react for 10min after which the capsules (if formed) were collected on a 6011111 mesh nylon filtration material (Spectramesh, Spectrum 1\'Iedical Industries, lnc., Houston, USA) and washed on the filter. Nlicrocapsule washing was performed in three stages; the first wash was with cyelohexane to quench the reaction, the second was a 5% Tween 20 solution to remove organics, and the third wash was with distilled water. “lashed micro- capsules were collected and transferred onto a Petri dish cover to dry before weighing and analysis. Samples were weighed on a l\'lettler HSIAR analytical microbalance. Considering the reactivity of the cyanoacrylate ester monomers, 100% conversion of monomer to polymer is theoretically possible. Therefore, per cent theoretical yield was calculated from the following equation: % yie1d,=gig-“>< 100 {1} gm where gmxgrams of monomer used, gp:grams of polymer used, and % yieldt :percent theoretical yield. Results Cymzomtryhzrte synthesis in the synthesis of homologous cyanoacrylate ester monomers, lower members of the series were successfully synthesized. However, higher members of the series were diflicult to synthesize. Excessive foaming of the reaction mixture with the high homologues was evident when a high vacuum was applied to the system. Subsequently, yields of these components were very low and substantially con- taminated with their precursor cyanoacetate ester. Of the monomers that were successfully synthesized, gas chromatograms showed a broad peak corresponding the cyanoacrylate ester when high purity was achieved. Since these compounds are extremely reactive, this result is reasonable. flrffcrocapsule formamin In an attempt to further purify the cyanoacrylate ester monomers before encapsulation, they were eluted through columns of ion-exchange resin or known acid-scavenging polymers. Five packing materials were investigated: Synrlresfs of semipermeable membrane nu'cracapsules 705 DO\VEX X1-8, A-ll’, Amberlyte A-21, polyvinylpyrrolidone (1\'Iw 40000), and poly(4-vinylpyridine). Of the packing materials tested, A-»lP and Amberlyte A-21 had a significant effect of purifying the monomers, as indicated by the short time periods necessary to totally solidify a monomer sample after treatment. Untreated monomer did not solidify within any appreciable time period. lVIost reaction conditions did not produce intact microcapsules; only those conditions which successfully produced microcapsules were recorded. Depending on the monomer used in the encapsulation, microcapsule formation was found to be critically dependent on the dielectric constant of the solvent and on the type of initiator used. Only conditions which enhanced both initiation and propagation produced intact microcapsules. To successfully encapsulate an aqueous droplet, enough polymer chains must form at the interface and intertwine to form a stable, continuous membrane. This can either occur with many short polymer chains or with fewer long polymer chains. Washing the reaction products on the large mesh filter provided a qualitative determination of microcapsule strength. Only those microcapsulcs with suflicient integrity to remain intact through the cyclohexane, detergent, and water washes were further analysed. Furthermore, the surviving rnicrocapsules were weighed before these analyses to give a measure of how favourable each reaction condition was to the formation of polymer. These results were converted into per cent theoretical yield by assuming 100% conversion of monomer to polymer. l\'Iost experiments were repeated twice; error bars depicted in the figures signify one sample standard deviation around the mean. Of the many organic solvent systems explored to dissolve the cyanoacrylate ester monomers, three systems were found to give the most favourable responses in terms of yield of microcapsules, membrane integrity, and membrane polymer weight. Figures 2 and 3 show that for decahydronapthalene/carbon tetrachloride and clecahydronapthalene/trichloroethylene, a maximum in per cent theoretical yield is achieved at an approximate solvent ratio of 85 : 15 and 95 : 5, respectively, for each system. This corresponds to a solvent mixture density of “'1. The dielectric constants for these two solvent mixtures are similar, with a value of 2.184 for the decahydronapthalene/carbon tetrachloride system and a value of 2.235 for the decahydronapthalene/trichloroethylene system (both determined from simple mixing rules). The critical dependence of yield on solvent mixture dielectric constant is displayed in figure 4. For six different solvent mixtures, all standardized to a solvent mixture density of "1, the maximum per cent theoretical yield occurs in the region of dielectric constants of 2.05-2.20. In figure 5, different cyanoacrylate ester monomers exhibit different dielectric constant optirnums depending on the nature of the ester side group. For more hydrophobic ester side groups, the dielectric constant optimum is shifted to lower numbers, while more hyrophilic ester side groups exhibit a high value dielectric constant optimum. For example, isoamylcyanoacylate (the most hydrophobic of monomers tested) exhibited a per cent theoretical yield maximum at 100% decahydronapthalene (dielectric constant: 2.174), while n-propyl and isobutylcyanoacrylate had per cent theoretical yield maximums at 85% (v/v) decahydronaphthalene/15% carbon tetrachloride (dielectric constant: 2.184). l\'Iethoxypropylcyanoacrylate (the most hydrophilic monomer by nature of the methoxy group) exhibited an optimum in 50% (v/v) decahydronaphthalene/50% trichloroethylene (dielectric constant: 2.787). 700 R. 3'. H011 arid R. P. Cl'mmbers 4-0 5% Theoretical yield 20 so 10 '3 r‘|IIII|t‘l'T*IfiI“Ir1-[‘T‘1II]11—r—I—| 2.10 2.20 2.30 2+0 2.50 2.60 2.70 Dielectric constant Figure 2. Per cent theoretical yields of polymer for deeallytlronaphtllalenefearbon tetra- chloride solvent. Changes in solvent dielectric constant achieved by varyirlg ratio of hydrocarbon to halogenated hydrocarbon. Both results use LO mu butylamine as the initiator and 2% (V/V) solution in solvent of A-21 treated n-butylcyanoacrylate. 8 4-0 55 95 Theoretical yield 25 so 20 .‘9 2.17 2.15 us 2.2:) Dielectric constant Figure 3. Per cent theoretical yields of polymer for cleeahydronaphthalene/triehloroethy- lene soivent. Changes in solvent dielectric constant achieved by varying ratio of hydrocarbon to halogenated hydrocarbon. Both results use i.DmM butylamine as the initiator and 2% (V/V) solution in soivent of A-21 treated n-butyleyanoacrylate. Synz‘hesi's of seiizipermeable membrane rm'cr'0capsztles 707 60 56 Theoretical yield 20 an ‘E0 2.90 2.10 no 130 2.4a 2.50 2.50 Dielectric constant Figure 4. Per cent theoretical yield of polymer for various dielectric constant solvents. Changes in soivent dielectric constant achieved by varying ratio of hydrocarbon to halogenated hydrocarbon. Solvents were made to a density of 1.0. All results use 1.0 mkl butylamine and a 2% (v,/\') solution in solvent of A-2] treated n-butylcyano« acrylate. Dielectric constants: iso-octane/carbon tetrachloride (66:34):‘‘ 2.041; n~ decane/‘carbon tetrachloride (69:3l):‘ 2.068; cyclohexane/carbon tetrachloride (73:27)=‘ 2.081; decahydronaphthalene/carbon tetrachloride (85:1S):“‘ 2.184; decahydronaphthalene/trichloroethylene (82:'l8):" 2.395; and n-ciecane/trichlor— oethylene (63:37):" 2.512. The efiect of initiator concentration on per cent theoretical yield is shown in figure 6. As the concentration of butyiamine initiator is increased, the yields of intact microcapsules increases to a maximum at ""0.5 mM. After this point, the per cent theoretical yield decreases as the concentration increases. Since cyanoacrylate ester monomers polymerize anionically, the pH of the aqueous encapsulation solution might have a significant eflect on the resulting microcapsules. Figures 7-10 show this point to be true. in figures 7 and 8, percent theoretical yield of intact rnicrocapsules increased with increasing pH for both initiator enhanced and unenhanced polymerizations, but the per cent theoretical yield appeared to increase at a siower rate between pH 6.5-8.0. Yields of polymer for initiator enhanced polymerizations did not Vary sig- nificantly, despite orders of magnitude variations in initiator concentration, as shown in figures 9 and 10. Also, figures 9 and 10 show that difierent pHs produce the same if not similar afiects on polymer yield through widely varying initiator concentrations. The yield of polymer is decreased at lower pH; however, as the initiator concentration was increased, per cent theoreticai yield dependence exhibited a complex behaviour, with a minimum and maximum in the yield occurring for both pHs tested. l\’Iaximums in yields for both pHs tested occurred 708 R. 3'. Hall and R. P. Cl'Iambers 55 Theoretical yield 19 20 1.80 2.00 no 2.40 2.60 2.5:: Dielectric constant Figure 5. Per cent theoretical yields of polymer for various dielectric constant solvents. Changes in solvent dielectric constant achieved by varying ratio of hydrocarbon to halogenated hydrocarbon. Various cyanoacrylate ester monomers were used in a 2% (V/V) solution in solvent. All results used 1.0 nm! butylaminc. Methoxypropylcyano- acrylate (——); Iso—amylcyanoaerylate (— -—); n—propylcyanoacrylate (W m); and isO~ butylcyanoacrylate (- - -). in the region of 0.01 and U.1m.\I butylamine, while minimums occurred at “I .O1n.\1 butylamine. W'ith the hope of changing the characteristics of poly(cyanoacrylate ester) membrane microcapsules, various mixtures of cyanoacrylate ester monomer were prepared. Since each monomer exhibits its own peculiarities, mixing them together might produce microcapsules possessing advantageous characteristics of both. To this end, n-butylcyanoacrylate monomer was mixed with cyclohexy], ally], and methoxypropylcyanoacrylate. Figure 11 shows that the per cent theor- etical yield of intact microcapsuies decreased when any of the monomers was mixed with n-butylcyanoacryiate with the most pronounced effect occurring for allyi/n-butylcyanoacrylate mixtures. Using a polymeric initiator changes the dependence of yield on initiator concentration. By using polyvinyipyrrolidone as the initiator instead of butyla- mine, figure 12 shows that the yield increases with increasing initiator concentra- tion as opposeci to a per cent theoretical yield maximum occurring when butylamine was used (figure 6). As with the previous mixed monomer experiments using cyclohexyl, allyl and methoxyproplcyanoaciylate, methylcyanoacrylate was also used in this capacity. Figures 13-15 show the results of these experiments using difierent solvent mixtures, monomer ratios, and total monomer concentrations in solvent mixture. Syntlresis of senzipermeable membrane Irzicrocapsxrles 709 8 70 60 50 40 $8 Theoretical yield 20 so 10 0.01 0.1 1 19 mo Concentration (mM) Figure 6. Per cent theoretical yields of polymer for various concentrations of butylamine initiator. Solvent was decahydronaphthalene/carbon tetrachloride (85:15) and a 2% (vfv) solution in soivent of A-21 treated n-butylcyanoacrylate was used. 8 30 95 Theoretical yield in 20 5.50 6.00 9.5-: 1.00 no 3.00 3.59 pH Figure 7. Per cent theoretical yields of polymer at various aqueous encapsulation solution pHs. Aqueous encapsulation soiution buffered with 50 mm phosphate. No initiator was used in this result. Ali results used a 2% (vjv) solution in solvent of A-21 treated n-butylcyanoacrylate and ciecahycironaphtiialene/carbon tetrachloride (85:15). 710 R. 3'. H011 and R. P. Cltambers 30 53 Theoretical yield 1!} 20 5.50 am 5.50 1.99 1.5a am 5.50 pH Figure 8. Per cent theoretical yields of polymer at various aqueous encapsulation solution pl-Is. Aqueous encapsulation solution buffered with 50mm phosphate. 0.201111! butyl-amine was used as the initiator. Ali results used a 2% (V/v) solution in solvent ofA-21 treated n-butyicyanoacrylate and decahydronaplithaleilefcarbon tetrachloride (85:15). 30 98 Theoretical yield 10 2o 3 I I n s 0.001 om 0.1 : Concentration of initiator (mm) Figure 9. Per cent theoretical yields of polymer at various concentrations of butylamine initiator. Aqueous encapsuiation solution buffered with 50mM phosphate and at pH 6.5. All results used a 2% (V/V) solution in solvent of A-21 treated n-butylc_\'anoacry- late and clecahydronaphllialene/carbon tetrachloride (85:15). Synthesis of sennpermeable membrrme microcapsules 711 8 JO 5 Theoretical yield 10 20 I I I I I 0.001 cm 9.1 _ _ ‘ 1 10 100 Cnncentrohon of smtlotor (mlvl) Figure 10. Per cent theoretical yields of polymer at various concentrations of butylamine initiator. Aqueous encapsulation solution buffered with 50 mu phosphate and at pH 8.0. All results used a 2% (v/V) solution in solvent of A-21 treated n-butylcyenoacry- late and decahydronaphthalene/carbon tetrachloride (85:15). as Theoretical yield 1a 20 o.oo one o.+o o.so 0.30 ma Fraction of n-butyl monomer Figure 11. Per cent theoretical yields of polymer for various mixtures of cyanoacrylate ester monomers. All results used optimum solvent types and compositions. Cyano» l acrylate ester mixtures were A-21 treated and used in a 2% (vfv) solution in solvent. i.0mM butylamine was used in all results. Cyclohexyl/n»butylcyanoacrylate ( ); Allylmvbutylcyanoacrylate (— —); and Methoxypropyl/r1—butylcyanoacrylate (- - - -). 712 R. j’. Ho]! and R. P. Cfmmbers 70 60 50 40 as Theoretical yieid 20 so ‘IO OJJOGOI CKDCOI I 0.00] I 1101 .1 I 1 Concentration of lnltiator €95 w/v) a a {O Figure 12. Per cent theoretical yields of polymer at various concentrations of polyvinyl- pyrrolidone (MW : 40 000) initiator. Aqueous encapsulation solution was at 6.5 and bufiered with 50m“ phosphate. All results used A-21 treated n—butyleyanoacrylate and decahydronaphthalenefcarbon tetrachloride (85:15). 0.100 0.080 0.000 ‘field (g of polymer) u.o2o 0.040 l 8 C) '5o.ou 0.20 0.40 use man 1.09 Concentration of initiotor (3% w/V) Figure 13. Yields of polymer at various concentrations of polyvinylpyrrolidone (l\‘I“- = 40 G00} initiator. Aqueous encapsulation solution was at pH 6.5 and buffered with 501n.\1 phosphate. Results shown are for 10% (V/v) methylcyanoacrylate in n- butylcytznoacrylate to make a 2% (V/V) total monomer in solvent solution. Two solvent mixtures were used: decal)ydronaphthalene/carbon tetrachloride (85:15) and decah_v— dronaphthalene/tricliloroethylene (82:18). Decahydronaphthalenefcarbon tetrachlor- ide ( ); and decahydronaphthalene,"t1-ichloroethylene (— — -). S_wrt/resis of sernipernreable membrane mthracapsrties 713 ‘field (g of polymer) 0.100 azoo moo I‘ \ \ \ \ \ I’ 650.00 0.20 0.40 9.50 0.50 1.00 Concentration of initiator (58 w/V) Figure 14-. Yields of polymer at various concentrations of poivinylpyrrolidone (MW 3 40 000) initiator. Aqueous encapsulation solution was at pH 6.5 and buflered with 50 mm phosphate. 5% (vfv) methylcyanoacrylate in n-butylcyanoacrylate and 60% (vfv) iso—octane/40% trichloroethylene were used in all results. Results are shown for various concentrations of monomer solvent. 1 “/0 total monomer in solvent (— —}; 2% total monomer in solvent (—- M); ti"/u total monomer in solvent (WM); and 8% total monomer in solvent (- - - -). Changing the solvent mixture from clecahydronapthalene/carbon tetrachloride to decahytlronapthalene/trichloroethylene increased the yield ofintact microcapsules. A solvent mixture of iso-octane/triehloroethylene produced yield results inter- mediate to the other two solvent mixtures. It as this solvent mixture which was used for the remaining experiments because of better results obtained from gel permeation chromatography and scanning electron microscopy analyses. Figures 14- and 15 show that by increasing the total monomer concentration in the solvent mixture, the yield of intact rnicrocapsules increases for both 5% methyicvan0a- crylate and 10% Inethylcyanoacrylate. For the experiments recorded, the yield of microcapsules also increased with increasing concentration of initiator; however, yield of microcapsules became consistent after 0.1% (w/v) polyvinylpyrrolidone. Additionaliy, 20% methylcyanoacrylate mixed monomer solutions did not produce successful results with any of the reaction conditions formulated in this study; also 10% methylcyanoacrylate mixed monomer solutions in a 1% total monomer solution in solvent did not produce successful results. Plotting the results of the previous experiments on a per cent theoretical yield basis gave additional information. Figure 16 shows the effect of changing polyvinylpyrrolidone initiator concentration in 85% (v/v) decahydronaphthalenef 15% carbon tetrachloride and 82% (vfv) decahydronaphthalene/18% trichloroethy- lene. By using the solvent mixture which more favourably enhanced initiation (tlecahydronapthalene/trichloroethylene), the rate of increasing per cent theor- 714 R. j’. H011 and R. P. Clmmbers 0.400 r a ___,——*" f“‘\ .—’—’ s ,»~’ E ,’ _>.~ ,/ O 3:, £18 / wwq / on I O1 ‘_J E 2 5- 0.100 .00 0.20 0.1.4: o.eo 0.59 1.00 Concentration of initiator (95 w/v) Figure 15. Yields of polymer at , various concentrations of polyvinylpyrrolidone (i\’I..- = 4-0 000} initiator. Aqueous encapsulation solution was at pH 6.5 and buffered with SUmM phosphate. 10% (v/v) methylcyanoacrylate in n-butylcyanoacrylate and 60% {v,"v) iso—oCtane/40% trichioroerhylene were used in all results. Results are shown for various concentrations of monomer solvent. 2% total monomer in solvent (— —}; 4% total monomer in solvent ( ); and 8% total monomer in solvent {- — — -). etical yield with respect to initiator concentration was shown to be greater than the rate for the solvent mixture, which was not favourable to initiation (decahydro— napthalene,icarbon tetrachloride). Figures 17 and 18 Show the effect of changing initiator concentration in 60% (v/ v) iso-octane/40% trichloroethylene with 5% (v/v) and 10% (v/v) methylcyanoa- crylate in n-butylcyanoacrylate, respectively. Also, the total amount of monomer added to solvent was varied from I% (V/v) to 8% (v/v). For the 5% (v/V) methylcyanoacrylate experiments (figure 17), the per cent theoretical yield was higher for 1% total monomer than for 2% total monomer. After this value, increasing amounts of total monomer in solvent increased the per cent theoretical yield. Also, rates of increase for per cent theoretical yield with respect to initiator concentration were generally higher for the higher percentages of total monomer in solvent. As mentioned above, for lO% methylcyanoacrylate, microcapsules were not successfully procluced using 1% total monomer in solvent. However, trends similar to those for the 5% methyl derivative experiments with regards to rates of increase for per cent theoretical yield with respect to initiator concentration and to per cent theoretical yield values for diflerent total amounts of monomer in solvent were observed. First, per cent theoretical yields were generally higher using the 5% methyl derivative mixtures as opposed to the 10% mixtures in corresponding experiments. Also, the 5% experiments produced rates of per cent theoretical yield increases higher than for the 10% experiments within similar reaction conditions. i i l l i l l l 4 E T I Syntizesis of semipeiwzeable membrmze mfcrompsuies 715 BO 60 95 Theoretical yield 20 +0 \ \ I I 1 a 3 0.11001 o.oo1 _o.o1 _ __ 9.1 1 Concentration of Initiator (5 w/v) 10 Figure 16. Per cent theoretical yields of polymer at various concentrations of polyvinyl~ pyrrolidone (i\'I\\' 2 40000) initiator. Aqueous encapsulation solution was at pH 6.5 and buffered with 50m.\l phosphate. 10% (v/v) methylcyanoacrylate in n-butylcya- noacrylate to make a 2% (v/V) total monomer in solvent solution. Two solvent mixtures were used: decahydronaphthalenejcarbon tetrachloride (85:15) and decah_v- dronaphthalene/trichloroethylene (82:18). Decahydronaphthalene,/carbon tetrachlor- ide ( ); and decahydronaphthalene/trichloroethylene (— — —). Discussion To successfully microencapsulate biological material, a method of forming strong, porous rnembrane must be devised. The use of poly(cyanoacrylate esters) has been shown in this study to satisfy these requirements. However, the high reactivity of these monomers allows many reaction variables to have significant influence on the polymerization reaction and the resulting polymers. Several of these variables have been determined in this study and their influences on the polyrnerizations of cyanoacrylate esters elucidated. As a result, stable, porous microcapsuies utilizing cyanoacrylate ester polymer membranes were able to be synthesized. The reaction variables discovered to have significant influence on cyanoacrylate ester anionic polymerization are: solvent dielectric constant and type, initiator type and concentration, monomer type and concentration, pH of the encapsulated solution, surface tension between the aqueous encapsulation solution and the solvent, proper technique and the presence of acidic compounds (usually impurities). As with other studies reported in the literature involving cyanoacrylate esters, difficulties were initially encountered with the contamination of cyanoacrylate ester monomer stocks with acidic impurities (usually carried over from synthesis). 716 R. J. Hall and R. P. Chambers 100 60 60 96 Theoretical yield 20 40 0.0001 0.005 0.01 O I . 1D Concentration of initiator (is W/V)‘ Figure 17. Per cent theoretical yields of polymer at various concentrations of polyvinyl- pyrrolidone (M“- : 40 000) initiator. Aqueous encapsulation solution was at pH 6.5 and buffered with S0mM phosphate. 5% (v/v) methylcyanoacrylate in n—butylcyano— acrylate and 60% (vfv) iso—octanef*l-0% trichloroethylene were used in all results. Results are shown for various concentrations of monomer in solvent. 1% total monomer in solvent (— -——); 2% totai monomer in solvent (- —); 4% totai monomer in solvent ( ); and 8% total monomer in solvent (- - - -). This problem could have proved disastrous to consistent results unless a technique could be devised to substantially clean—up the stocks prior to use. Since the monomers are extremely reactive, the technique would have to be fast and simple to be usefui. The discovery that anionic ion-exchange resins could be used to significantly improve the quality of the monomer stocks solved this problem. Also, by using disposable syringes and small quantities of resin, the method proved both quick and eflicient at preparing the monomer stocks for use. Reaction conditions generally found to produce microcapsules include a solvent mixture with a dielectric constant in the range of "’2.l-2.2 for polymer- izations using n-butylcyanoaciylate, buffered encapsulation solution at approxi- mately a neutral pH, and a moderately nucleophilic initiator (pKa 2 9"l0). Initiation is favoured by high concentrations of initiators, high dielectric constant solvents, species (salts) which stabilize the betaine, high nucleophilic strength initiators, and initiation complexes devoid ofsteric hindrances. Conditions which favoured propagation allowed the growing polymer chains to intertwinc before being terminated, producing intact membranes. Propagation is favoured by low initiator concentrations, iow dielectric constant soivents, conditions which rnainw tain the electronegativity of the propagating chain (low salt or high concentration of negative ions associated with the activated endgroup), and samples substantially free of acid impurities. Figure 19 shows the influence of both initiation and Synfizesis of semipermeable membrane nucrocapszrles 717 60 60 95 Theoretical yield 20 -so I I 0.003! 0.00! I 0.01 I I 0.1 1 Concentration of Initiator (58 w/v) ID Figure 18. Per cent theoretical yields of polymer at various concentrations of polyvinyl- pyrrolidone (l\'l\\' : 40 O00) initiator. Aqueous encapsulation solution was at pH 6.5 and bufliered with 50mm phosphate. 10% (v/v) methylcyanoacrylate in n—l)utyl~ cyanoacrylate and 60% (vfv) iso-octane/40% trichloroethylene were used in all results. Results are shown for various concentrations of monomer in solvent. 2% total monomer in solvent (— -~); 4% total monomer in solvent (———); and 8% total monomer in solvent (~ — - -). propagation, as affected by a hypothetical reaction property on membrane quality such as yield or thickness. For example, as the solvent dielectric constant is increased, initiation is favoured at the expense of propagation, producing low molecular weight polymers which could not sufficiently intertwine to produce a stable membrane and, consequently, a mierocapsule. Unfortunately, conditions which favour initiation over propagation are more easily achieved in anionic polymerization than conditions which favour propagation. Since the cyanoacrylate ester monomels are extremely reactive and can be initiated by a variety of compounds, lo\v concentrations of initiator are hard to achieve. h'Ionomer poly- mers are not soluble in solvents of low dielectric constant which would favour propagation. The eflect of solvent dielectric constant was the first reaction variable to be studied. The dielectric constant (or relative permittivity) of a material is the ratio of the capacitance which occurs when the material is placed between plates of a parallel-plate capacitor to the capacitance which occurs in the same capacitor when the material is removed and replaced with a vacuum. By inserting the material between the capacitor plates, a reduction in the potential difference is observed and caused by the induction of negative charges in the material adjacent to the positive plate and the induction of positive charges adjacent to the negative plate (Sears .91 mi. 1977). The analogy can be extended to solution electrochemistry in much the 718 R. j‘. Ho]! and R. P. Chambers propagation Enltlaflon Membrane quali Reaction property Figure 19. The effects of initiation and propagation singly and combined on a membrane quality as affected by a hypothetical reaction property (solvent dielectric constant, initiator concentration, etc.}. same manner (Atkins I978). The Coulombic potential as a function of distance from an ion in a vacuum can be expressed as ¢I,'{r) : lqi/‘hreolll/rl where 59 :vacuum permittivity. “lhen the ion is in solution, the solvent decreases the strength of the potential by the induction of opposite charge in solvent molecules adjacent to the ion analogous to the parallel-plate capacitor. Therefore, the expression for the Coulomhic potential is modified to qbslrl = lqi/4-7rE{JI{,ll1/3‘) where K} 2 dielectric constant. The reduction of the Coulombic potential is of great importance to solvents. A solvent with a high dielectric constant, such as water lK, : 78.54-l, can reduce the potential to small values at very short distances from the ion, as opposed to a solvent with a low dielectric constant, such as benzene (K, : 2.274). Therefore, ions in water interact weakly and do not tend to associate with each other, whilst ions in benzene tend to be insoluble. Another electrical property of molecules which has importance with regards to solvents is the dipole moment. Dipole moments can be either permanent or induced by an electric field depending on the constituents of the molecule. Non- poiar molecules, such as benzene, do not possess a permanent dipole moment but one can be induced by an electric field, polarizing the molecule. Polar molecules, such as water, possess a permanent dipoie moment. However, both the permanent Synthesis of semipermeable membrmre mfcrompsules 719 and induced dipole moments of a molecule can be related to its dielectric constant through the Debye equation (Atkins 1978), which is expressed as Niog "i',[z2/35017871) 3 3(Kr ~ 1}/(K. + 2) where, (x Zpolarizability (dipole moment inductibility), p Zpermanent dipole moment, It :Boltzmann constant, T“—"'temperature (K), and Nlznumber of molecules/unit volume. Therefore, the dielectric constant can give an indication of the charge char» acteristics of a given solvent. The value describes how easily charge can be separated on the solvent molecule, with low values indicating that charge separa- tion is not favoured with non-polar, hydrocarbon solvents as opposed to high value solvents such as water. Table 4 gives the dielectric constant, polarizability, and dipole moment for the important compounds used in this study. Initial experiments with pure hydrocarbon or halogenated hydrocarbon were not very successful (results not reported) and emphasized the need for an organic phase which had a density approximately equal to the aqueous phase being encapsulated. Surveys of available organic solvents discovered that most solvents with densities close to 1.0 had too high dielectric constants to be useful. Sub- sequent experiments bore this fact out (results not reported). Therefore, mixed solvents containing hydrocarbons and halogenated hydrocarbons with densities adjusted to 1.0 were decided to give the best solvent conditions possible. However, when these solvent mixtures were formulated, the effects of dielectric constants were noticed. As reported in figures 24, the dielectric constant of the solvent had a significant influence on the yield of successfully produced microcapsules. Even amongst solvent mixtures with the same density, dielectric constant effects produced a 3-fold increase in yield from the lowest solvent mixtu re tried to the highest. That different cyanoacrylate ester monomers seem to have different dielectric constant optirnurns depending on the monomer’s hydrophobicity, as shown in figure 5, adds credence to the results that the dielectric constant plays an important role in cyanoacrylate ester polymerization. Since the cyanoacrylate esters poly- merize anionically, a stable initiation complex must be formed before a propagat- ing polymer chain is realized. The initiation complex for this type of polymerization is a betaine; therefore, a sufficiently charged molecular environ- ment must exist to stabilize the newly formed betaine (see figure 20). A solvent mixture of such a dielectric constant which optimizes this stability will enhance the Table 4-. Dielectic constants, polarizabilities, and dipole moments for compounds used in this study. Com pound K, (\(cm3 )0 ,u(D) Iso-octane 1.940 196 — n-decane 1.991 241 — Cyclohexane 2.023 137 — Decahydronaphthalene 2.174 216 -- Carbon tetrachloride 2.238 I40 -- Trichioroethylene 3.4 197 1.43 \Vater 73.54 13.6 1.85 "Polarizability calculated from Debye equation. 720 R. j’. H011 and R. F. Chambers ©© initiation complex in a low dielectric constant solvent - charge separation not enhanced — bond strain increased » initiation complex teas stable QED c;T_»> 43-‘) & C;—S_9 iii i CN 33' . 1 H3C—(CH2)3—NH2“CH2*E3“‘~~—% 1 E coca cm i Gig (:19 €33 initiation complex in a high dielectric constant solvent -- charge separation enhanced - bond strain decreased - initiation complex more stable Figure 20. A schematic of the diflerent eflects of low and high dielectric constant solvent on the stabilization of a buty[amine/cyanoacrylate ester initiation complex. number of successful initiations. However, some solvent mixtures of similar dielectric constant did not have the same effect on yield (the results from cyclohexane/carbon tetrachloride in figure 4). Therefore, the components of the solvent mixture itself may play a role in stabiiizing the initiation complex. Another example of this result can be seen in figures 3 and 4-, in which trichloroethylene solvents had the ability to give fair yields and membrane characteristics at dielectric constants higher than those which gave optima for carbon tetrachloride solvents. This effect may be attributable to the similarity in double-bond character between the unsaturated trichloroethylene and the vinyl group on the cyanoacry- late ester monomer. This brings up the subject of the influence of initiator type and concentration on the cyanoacrylate ester polymerization. Preliminary studies indicated that yield was roughly correlated with nucleophilicity as indicated by a compound‘s pKa. Therefore, piperidine (pK3 : H.123) was a much more potent indicator than benzylamine lpKa : 9.33) (results not reported). Butylamine (pKa = 10.77) was chosen as a reference initiator in the beginning of the study because of its intermediary pKa among bases (primarily amines) surveyed as being possible successful initiators. As figure 6 shows, increasing the concentration of this initiator increased the yield of polymer, as would be expected. Anomaiics present in the results at low initiator concentration can be attributed to the fragility of the resulting rnicrocapsules. This effect can also be seen in the previous results discussed above concerning dielectric constants. Synflresis of semfperrrleable membrmie rrzicrocapsttles 721 Since cyanoacrylate esters polymerize by an anionic reaction mechanism, the pH of the aqueous encapsulation affects microcapsule membrane formation. in the previous experiments discussed above, the butylamine initiator was dissolved in distilled water. In subsequent experiments, a phosphate bu Her was used at various pHs to establish the dependence of the polymerization of pH. The results indicate that pH does play an important role in cyanoacrylate ester polymerization. As pH is raised, the yield of polymer increased (figures 7 and 8). This goes with expected results of anionic polymerization as, at higher pH, more anions are present to initiate the reaction. An interesting result, however, was that the yield with initiator was comparable to the yield without initiator. Another interesting result was that the presence of initiator made the yield somewhat consistent despite changes in pH. This result is consistent with the fact that the concentration of viable butylamine initiator (anions) will increase with increasing pH, since the pK.., of the amine is 10.77. The results without initiator seem to indicate that an unknown species, not affected by pH above 6.0, at a constant concentration between each experimental run is responsible for the consistency of the results. The species could possibly be residual cyanoacetic acid (pK.,, 2 2.47) or its ester; Costa at :71. (1983) mentioned that weak acids can also act as indicators if deprotonated. Or, it could be a component bled from the ion-exchange resin used in the treatment of monomer before each encapsulation experiment. Of course, it is conceivable that any impurity with basic tendencies could be responsible for this result. VVhen the concentration of butylamine was varied in buffered solution, com- plex behaviour was exhibited. The yield of polymer went through a maximum at low concentration and then a minimum at higher concentrations (figures 9 and l0). This result was displayed at both pH 6.5 and 8.0, although the maximum and minimum were displayed to higher concentration values for the lower pH. Also, yields were consistently higher for the higher pH. Apparently, a concerted effort is being displayed between the influence of pH and concentration on the resulting polymerization. This would indicate either that the initiator was more effectively utilized at the low concentrations as opposed to the higher ones or that at low concentrations more initiator was present at the interface to participate in polymerizations. The First mechanism implies a disruption in the usual electro- chemistry of butylamine, possible caused by the presence of competing or impeding ions. The second mechanism could be thought of as a change in the surface tension properties of the organic/aqueous interface caused by differences in the packing of interfacial ions at different concentrations of initiator (an efiect postulated by Florence er ml. 1976). By mixing different cyanoacrylate ester monomers together before polymeriza- tion, copolymers of the different monomers could be prepared. Since the different monomers possessed unique properties by themselves and when complexed with others, the resulting copolymers were hoped to possess properties indicative of each monomer. By manipulating the amount of each monomer, copolymers could be optimized in certain properties deemed important to particular applications. Four cyanoacrylate ester monomers mixed separately or in combination with n- butylcyanoacrylate were used in this capacity. i\’Iethylcyanoacrylate was chosen for its rapid degradation rate (Leonard er al. 1996a, 1996b, 1967, Vezin and Florence 1978, 1980), as opposed to mbutylcyanoacaylate in hopes of etching large pores through the microcapsule membrane. Since its double-bond had the capacity to 722 R. j’. Hair’ and R. P. Chambers cross—linlt (Cheung at at’. E985), allylcyanoacrylate was used with the hope of strengthening the microcapsule membrane. Cyclohexylcyanoacrylate, because of its large ester side group, was employed to inhibit the formation of crystallinity in the membrane polymer, thus reducing the possibility of membrane brittleness and increasing membrane flexibility. Talks with a representative of Permabond Cor- poration (Bridgexvater, N], USA) brought attention to the fact that the methox— ypropyl derivative displayed the most flexibility of the cyanoacrylate ester polymers tested by this company. Therefore, complexing this derivative with n- butylcyanoacrylate was hoped to increase the flexibility of the resulting copoiymer. Results indicated that some mixing of the monomer properties could be achieved in the formed copolymer; however, the addition of any monomer did reduce the yield of polymer compared to straight n-butylcyanoaciylate polymerization. The addition of the methyl derivative did result in the formation of ores through the membrane after a 24 h degradation in pl'I 8.0 buffer when iso-octane/trichloiu octhylene was used as the solvent mixture. The other solvent mixtures used in this capacity: decahydronapthalene/carbon tetrachloride and decahydronaphthalenef trichloroethylene were unsuccessful at consistently producing porous membranes. Allylcyanoacrylate had the most drastic eflect on yield, with no successful encapsulations occurring above an additional concentration of 10% of the total monomer. Results with cyclohexyl and methoxypropylcyanoacryiate were incon- elusive. Wihen it was discovered that butylamine produced extremely brittle micro- capsules, polyvinylpyrrolidone was tried as an initiator. The result using poly- vinylpyrrolidone as the initiator was that yield was enhanced by increasing concentrations of the initiator and the resulting microcapsuies were less brittle. Increasing yield was a similar result obtained using increasing concentrations of butylamine. Since various techniques can be used to add combinations of difierent cyan- oacrylate ester monomers together in the solvent mixture and since these monomers are known to be highly reactive, a series of experiments were performed to determine if the manner in which monomers were added to solvent could eflect the resulting poiymerizations. Results indicated this to be true. The best method of adding cyanoacrylate ester monomers was found to be to add them separately to the solvent, pre-treating them with ion exchange resin, as discussed above. The final series involved determining the optimum reaction conditions to form thin, strong, porous, polymer membranes for microencapsulation based on the characteristics of the polymerization determined from the previous experiments. From these experiments, satisfactory results could be obtained using 60% (v/v) iso- octane/40% trichloroethylene as the solvent mixture, 0.0l~0.l0"/u (w/v) polyvinyl- pyrrolidone as the initiator (and, possibly, propagation enhancer), and 5% (V/v) methylcyanoacrylate/95% A-21 treated n-butyicyanoacrylate monomer mixture added separately to solvent to give a 4% (v/v) solution. Increasing the total monomer concentration in solvent to 8% substantially increased the yield of microcapsules, while 1% and 2% concentrations did not produce a strong enough membrane for practical applications. The addition of a small quantity of allylcya- noacrylate to the monomer mixture (2% of total monomer) was also found to give satisfactory results. Both formulations were used in subsequent encapsulation experiments. .S'yntlzesr's of semipermeable membrane Jmhocapsules 723 Conclusions iVIany discoveries have been made in the course of this investigation. First, higher homologues of the cyanoacrylate ester series are difiicuit to synthesize. Secondly, the treatment of cyanoacryiate ester monomer stock with ion exchange resin can significantly reduce the amount of acidic impurities, which can adversely effect the consistency of results and the polymerization itself. The yield of polymer was found not to correlate accurately with the production of microcapsules possessing optimum encapsulation properties. The best micro- capsules were produced at reaction conditions, which did not give the highest yields. Polyvinylpyrrolidone was found to act as an eflective poiymerization initiator. By the use of this polymer, good quality membrane polymers could be obtained. By the successful manipulation of reaction conditions which include solvent mixture components and dielectric constant, initiator type and concentration, control of aqueous encapsulation solution pH, proper technique and monomer stocks sufliciently pure of acidic impurities, poly(cyanoacrylate ester) membrane microcapsuies can be produced. In conclusion, cyanoacrylate esters can be used to synthesize microcapsules quickiy and efiectively. Also, because of the nature of the polymerization, they should be able to be used to encapsulate biological material including live cells within porous, strong microcapsules without significantly denaturing the material or killing the cells being encapsulated. This discovery opens the door to many potential applications in biotechnology, which hitherto could not be successfuiiy implemented. References ATKINS, P. \V., 1978, Physical C}'1e1m'.r1‘ry (San Francisco: \V. H. Freeman and Company), pp. 31+—338. CHEUNG, K., GLITE{RIE, ]., OTTERBURN, M., and RODNEY, I., 1985, The dynamic mechanical properties of poly alkyl-2—cyanoacryiates. Poiymer Preprints, 26, 224-225. Cosn, G., CRUNIN, J., PEPPER, D., and LooNAN, C., 1983, Termination and transfer by acids in the pyridine-initiated polymerization of butyl cyanoacryiate. European Polymer jazirrral, 19, 939-945. FLUREi\'CE. Au HAQ, M., and ]oHNsoN, }., I976, Interfacial properties of polymethylw cyanoacrylate and poIybutyl-cyanoacrylate. joaarrial of Pharmacy rmdPlmrnraco1ogy, 28, 539-543. LEONARD, F., COLLINS, J., and PORTER, H., l966a, Interfacial polymerization of n-akyl- cyanoacrylate homoiogs. jomvml of Applied Polymer Science, 10, 1617-1623. LEONARD, F., KUI.KAItNl, R., BRANDES, G., NsLsoN, 1., and CAMERON, ]., 1966b, Synthesis and degradation of poly(alkyl-cytmoacrylates). _7ourm'u' of Applied Polymer Science, ll}, 259-272. LEONARD, F., KLILI{Aa\*1, R., NELSON, J., and BRANDES, G., 1967, Tissue adhesives and liemostasis-inducing compounds: the alkyl cyanoacrylates. jbzmml of Bionzedira! 1h'ater'z'aIs Research, 1, 3-10. PEPPER, D., and RYAN, B., 1983, Initiation processes in polymerizations of alkyl cyanoacrylates by tertiary amines: inhibition by strong acids. lliaieronroierular C‘izem1'sfry, 184, 383w39-l-. Seams, F. W., ZEMANSKY, M. W., and YOUNG, H. D., 1977, Um'vers:'t_v Pizysics, 5th edn. (Reading, MA: Addison-VVesley Publishing Company), pp. 464-471. 724 Synflzesis of seimpermeable membrane micmcapsm'es VEZIN, \V., and FLURENCFI, A., 1978, In 'vin‘o degradation rates of biodegradable poly~n—aEkyl cyanoacrylates. jromuml of Pharmacy and Plmrnmcology, 30, SP. VEZIN, \V., and FLORE.\'(:F., A., 1980, In vitro heterogeneous degradation of pol_v(n—alk_vl— cyanoacrylates). _‘)’om'Jm{ of Biorrredfcd I'I'[afen'aIs Research, 14, 93-106.