Methoxy-polyethoxy side-chain silastomers as materials controlling drug delivery by diffusion flux

Journal of Controlled Release 54 (1998) 273–282

Methoxy-polyethoxy side-chain silastomers as materials controllingdrug delivery by diffusion flux

*Helmut Loth , Ulrich Foltin¨ ¨Fachrichtung Biopharmazie und Pharmazeutische Technologie der Universitat des Saarlandes, Im Stadtwald, D-66123 Saarbrucken,

Germany

Received 23 January 1997; accepted 14 November 1997

Abstract

The density of a diffusion medium and the solubility of the diffusant in this material are predominant parameters whichcontrol the diffusion flux. Side-chain silastomers can be structurally modified in such a manner that density and polarity arewidely varied and adjustable to distinct conditions. The synthesis of cross-linked methoxy-polyethoxy side-chainpolysiloxanes is performed by reaction of a,v-bis-(trimethyl-silyloxy)-poly-(methyl-hydrogen-siloxane), a,v-divinyl-poly-(dimethyl-siloxane), and 4-propenyloxy-(49-methoxy-polyethoxy)-biphenyl. Silastomer foils were obtained with side-chainshaving up to 9 ethoxy groups. Various silastomer types were characterized by DSC-measurements, polarization microscopy,uptake of water and salicylic acid (as model drug), and by permeation measurements. The polarity of the polymers dependson their contents of ethoxy groups influencing the uptake of polar substances. Polymers with penta-ethoxy and hepta-ethoxygroups at the side-chains take up water under swelling. The solubility and the distribution coefficients of salicylic acid arelinearly correlated to the weight fractions of the methoxy-polyethoxy groups in all silastomer types. The temperaturedependence of the distribution coefficients of the penta-ethoxy and hepta-ethoxy polymer types shows deviations from theArrhenius equation. As the side-chains occupy considerable volumes, the density of the molecular packing within thecross-linked polysiloxane matrices is high. The arrangement of the side-chain domains, therefore, depends on the chainlength of the cross-linker; the diffusion coefficients are influenced by this parameter. Evidence for the existence of alyotropic liquid-crystalline phase was observed for one type of these polymers only. Methoxy-polyethoxy side-chainsilastomers as membranes or matrices are suited for controlled drug delivery. Drug liberation rates and swelling by uptake ofwater can be widely altered by variations of the side-chain structures. 1998 Elsevier Science B.V. All rights reserved.

Keywords: Methoxy-polyethoxy side-chain silastomers; Variable polarity; Variable density of the matrix; Lyotropic liquidcrystallinity; Drug solubility; Drug permeation; Drug delivery

1. Introduction or matrices [1,2] even for the delivery of macro-molecules [3]. Variation of the release rates is

Control of Fickian transport for drug release is achieved by changing type and structure of thewidely performed by the use of polymer membranes macromolecules in several ways influencing not only

the density of the matrix but also the solubility of the* migrating substance in the polymer. Furthermore,Corresponding author. Tel.: 149 681 3022278; fax: 149 681

3024677. this purpose can be attained by admixing suited

0168-3659/98/$ – see front matter 1998 Elsevier Science B.V. All rights reserved.PI I : S0168-3659( 97 )00218-6

274 H. Loth, U. Foltin / Journal of Controlled Release 54 (1998) 273 –282

additives [4–7]. These additives, however, may be and proportions of cross-linkers and side-chains. Inreleased after application changing the drug delivery order to obtain thin and regular foils, the reactionrate and cause unwanted side-effects in the organism. proceeded for 8 h at 608C under nitrogen in a vesselSilicone-based devices have been reported as useful (diameter: 11 cm) according to [11]. The resultingadhesives for transdermal drug delivery [8]. Side- products were obtained in the swollen state havingchain silastomers offer many-fold structural var- taken up toluene. In order to remove this solvent,iabilities. They can be modified by altering the chain they were carefully treated for some days withlength and the degree of cross-linking as well as by methanol–toluene mixtures which were changedchanging the type, structure, size, and content of the under reduction of the toluene content from time toside-chains [9–11]. In addition, the physical state or time, until pure methanol was used as eluent. Finally,modifications of the side-chain domains are impor- the foil was freed from residual solvent at 608C andtant parameters controlling the diffusivity in thermot- 0.1 mbar. The thickness of synthesized membranesropic and lyotropic liquid-crystalline (lc) media was between 250–350 mm.showing particular temperature dependence or swell-ing by uptake of water. Temperature-dependent 2.1.2. Syntheses of side-chains [4-propenyloxy-(49-pulsatile drug release from methacrylate-based poly- methoxy-polyethoxy)-biphenyl]mers has been reported [12]. Whereas we have 4,49-Dihydroxy-biphenyl: Bayer AG, D-Lever-already investigated these aspects with thermotropic kusen, Germany.lc side-chain silastomers, in order to look for special 3-Bromo-propene(1): E. Merck, D-Darmstadt,effects useful for diffusion flux control [9–11], the Germany.present paper is concerned with lyotropic polymers 4-Propenyloxy-49-hydroxy-biphenyl: 4,49-of analogous structures. Dihydroxy-biphenyl (1 mol) and finely-pulverized

sodium hydroxide (1.05 mol) were refluxed inmethanol (500 ml) for 30 min; ground sodium iodide

2. Materials and methods (a pinch) and 3-bromo-propene(1) (1 mol in smallportions during several hours) were added, and the

2.1. Materials mixture was refluxed for 24 h. After evaporation ofthe solvent, the residue was repeatedly crystallized

2.1.1. Synthesis of side-chain silastomer foils from sodium hydroxide solution (10% in water); thenthe hot solution (908C) was filtered and neutralized

Main-chain: a,v-Bis-(trimethyl-silyloxy)-poly- by diluted hydrochloric acid. The sediment was(methyl-hydrogen-siloxane) with 50–70 methyl-hy- crystallized from ethanol giving 4-propenyloxy-49-drogen-siloxane units (60 on average, average mo- hydroxy-biphenyl (mp. 166–1678C; literature: 1678Clecular weight: 3770), Wacker-Chemie GmbH, D- [15]).Burghausen, Germany. 4-Propenyloxy-(49-methoxy-polyethoxy)-bi-

Cross-linker: a,v-Divinyl-poly-(dimethyl-silox- phenyls: 4-Propenyloxy-49-hydroxy-biphenyl (1ane), Wacker-Chemie GmbH, D-Burghausen, mol) and finely-pulverized sodium hydroxide (0.11Germany; z510 (average molecular weight: 928), mol) were refluxed in 2-butanone (100 ml) for 15z525 (average molecular weight: 2040). min, and then ground sodium iodide (a pinch) and

The syntheses of the silastomer foils were exe- methoxy-ethylene chloride or one of the methoxy-cuted according to Finkelmann et al. [13] and in polyethoxy chloride compounds (0.1 mol; see Sec-analogy with a method described earlier [11]. Stoi- tion 2.1.3) were added. The mixture was refluxed forchiometrical portions of the reactants (main-chain, 8 h. After filtration, 200 ml of 2-butanone and watercross-linker and side-chain giving together about 1.5 (until saturation) were added at room temperature.g of silastomer foil) were dissolved in 5 ml of The solution was purified by repeated shaking withtoluene, and 50 ppm dicyclopentadienyl platinum water. Then, the organic solvent was removed, andchloride synthesized pursuant to [14] were added. the residue was separated by column chromatog-Diverse polymers were obtained by varying the types raphy (silica gel, diethylether). After initial elution of

H. Loth, U. Foltin / Journal of Controlled Release 54 (1998) 273 –282 275

nonreacted 4-propenyloxy-49-hydroxy-biphenyl, 4- and tetrahydrofurane (150 ml). Synthesis and purifi-propenyloxy-(49-methoxy-polyethoxy)-biphenyl was cation were performed as described above (seeextracted by an acetone–diethyl ether mixture (3:2). triethylene glycol monomethyl ether chloride). Tetra-

23The reaction products were crystallized from ethanol: ethylene glycol dichloride: b.p. 1208C (at 10204-Propenyloxy-(49-methoxy-ethoxy)-biphenyl: mbar), n 51.4639.D

m.p. 139.68C. Heptaethylene glycol monomethyl ether chloride:4-Propenyloxy-(49-methoxy-triethoxy)-biphenyl: Sodium hydroxide (0.5 mol) and pulverized sodium

m.p. 868C. iodide (a pinch) were dissolved in triethylene glycol4-Propenyloxy-(49-methoxy-pentaethoxy)-bi- monomethyl ether (1 mol); tetraethylene glycol

phenyl: m.p. 62.5–63.58C. dichloride (0.5 mol) was added dropwise (tempera-4-Propenyloxy-(49-methoxy-heptaethoxy)-bi- ture ,608C). The mixture was stirred at 1008C for 8

phenyl: m.p. 46.8–48.58C. h and poured into iced water afterwards. Further4-Propenyloxy-(49-methoxy-nonaethoxy)-bi- purification was performed as described above. Hep-

phenyl: m.p. 408C (literature: 408C [15]). taethylene glycol monomethyl ether chloride: b.p.23 201328C (at 10 mbar), n 51.4551.D

2.1.3. Methoxy-polyethoxy chloride compounds Heptaethylene glycol monomethyl ether: Tri-Tetraethylene glycol, triethylene glycol mono- ethylene glycol monomethyl ether chloride (0.5 mol)

methyl ether, methoxy-ethylene chloride, and bis(2- was added dropwise to a mixture of tetraethylenechloroethyl) ether: E. Merck, D-Darmstadt, glycol (1 mol), sodium hydroxide (0.5 mol), andGermany. sodium iodide (a pinch). The solution was stirred at

Triethylene glycol monomethyl ether chloride: 1008C for 72 h. The purification was performed asThionylchloride (1.1 mol) was added dropwise to a described above (see triethylene glycol monomethylmixture of triethylene glycol monomethyl ether (1 ether chloride). Heptaethylene glycol monomethyl

23 20mol), pyridine (1.1 mol), tetrahydrofurane (150 ml), ether: b.p. 1408C (at 10 mbar), n 51.4571.D

and dimethyl formamide (some drops). The solution Nonaethylene glycol monomethyl ether chloride:was intensely stirred and the temperature was not Heptaethylene glycol monomethyl ether (1 mol),allowed to exceed 808C. After stirring for 5 h, the bis(2-chloro-ethyl) ether (0.5 mol), sodium hydrox-mixture was cooled to room temperature; then it was ide (0.5 mol) and sodium iodide (a pinch) wereadded to iced water which was repeatedly extracted treated as described above. Nonaethylene glycol

23with diethyl ether afterwards. The ether phase was monomethyl ether chloride: b.p. 1898C (at 1020 20neutralized by sodium hydrogen carbonate and dried mbar), n 51.4592 (literature: n 51.4610 [15]).D D

by sodium sulfate. After evaporating the ether, theresidue was purified by fractional distillation. Tri- 2.1.4. Other substancesethylene glycol monomethyl ether chloride: b.p. Silgel 600: Wacker-Chemie GmbH, D-Burg-

20608C (at 16 mbar), n 51.4414.D hausen, Germany.Pentaethylene glycol monomethyl ether chloride:

Sodium hydroxide (1 mol) and pulverized sodium2.2. Methods

iodide (a pinch) were dissolved in triethylene glycolmonomethyl ether (1 mol). Bis(2-chloro-ethyl) ether

2.2.1. Thickness of the silastomer foils(1 mol) was added dropwise (temperature ,508C).Thickness meter, model 5041 type VRZ 181 withThe mixture was stirred at 1008C for 8 h and poured

tactile probe MT 10 B, Heidenhain, D-Traunreut,into iced water afterwards. Further purification wasGermany; accuracy 61 mm.performed as described above. Pentaethylene glycol

23monomethyl ether chloride: b.p. 1158C (at 1020mbar), n 51.4510. 2.2.2. Thermal studiesD

Tetraethylene glycol dichloride: Thionylchloride Thermal Analyzer, model 990 with DSC cell 910,(1.1 mol) was added dropwise to a mixture of Dupont de Neymours, D-Bad Nauheim, Germany.tetraethylene glycol (0.5 mol), pyridine (1.1 mol), Polarization microscope with heating stage, Ortho-


lux II Pol-BK, Ernst Leitz GmbH, D-Wetzlar,Germany.

2.2.3. Permeability measurementsPermeability measurements were performed in a

two-compartment diffusion cell with mounted silas-tomer membranes as described in [11]; diffusant:salicylic acid; donor phase pH 2.9, receptor phase pH7.4; temperature range: 20–808C.

2.2.4. Distribution coefficientsSmall pieces of the silastomers were equilibrated

Fig. 1. Representative structure unit of the methoxy-polyethoxywith aqueous salicylic acid solutions (18,24 mmol / l,side-chain silastomers.

pH 2.9) by shaking at constant temperatures (range20–808C). Subsequently, the silastomer pieces were

1. Length of the cross-linker chains z510 or 25rinsed with water and completely extracted with

dimethylsiloxane units.ethyl acetate. The salicylic acid in the eluents was

2. Number of ethoxy (EO) units at the side-chainquantitatively estimated by means of HPLC.

n51, 3, 5, 7 or 9 (it should be considered, that therelative weight portion of ethoxy units in poly-mers with the cross-linker z510 is considerably

3. Results and discussion higher than in comparable types with z525; seeTable 1).

3.1. Synthesis of methoxy-polyethoxy side-chain 3. Portion of the main-chain units substituted bysilastomers side-chains in %5s [stoichiometrically calculated

by means of the relationship x1y560 (number ofThe thermotropic lc-silastomers investigated ear- side-chains per main-chain1number of cross-

lier [10,11] are cross-linked polysiloxanes with 4- links per main-chain5number of methyl-siloxanemethoxyphenyl-(49-alkenyloxy-benzoate) groups as units per main-chain)].side-chains; their permeability for drugs is relativelylow, and they do not swell with the uptake of water. These variables determine the structures and prop-Increasing permeability is expected, if the polarity of erties of the silastomers. They are combined in athe side-chains is enhanced by substitution with code characterizing the polymer type in the follow-polyethoxy groups. The solubility of a polar diffus- ing manner:ant is presumed to increase in accordance with the

n 2 EOszresults of Feldstein et al. [16]. Furthermore, it isworthwhile to test, whether the incorporation of In principle, the membrane foils were synthesized bywater may alter the diffusivity through the matrix. the method described earlier [10,11]. In order toFor this reason, 4-propenyloxy-(49-methoxy-poly- obtain silastomers with mechanical properties appro-ethoxy)-biphenyl groups were used as side-chains for priate for use as membranes, the syntheses can bethe silastomers investigated in the following. The performed within limited ratios of the reactants only,polymer network (Fig. 1) was modified by varying otherwise, incomplete cross-linking occurs [17]. Ourthe length of the cross-linker chains and by the ratio experience showed that approximate rules referringof cross-linker to side-chains both of which are to the number n of the ethoxy units in the side-chainssubstituted at the main-chains. The different silas- are as follows: If n#3, the main-chain can betomer types, therefore, are characterized by the substituted by the biphenyl side-group up to aboutfollowing parameters: 90%; if n$5, the content of side-chains has to be

diminished with increasing n and with decreasing


Table 1List of the synthesized side-chain silastomers and the contents of polar groups in relation to the silastomer mass

Polymer type Weight Weight fraction Weight Molar fractionfraction of of the methoxy- fraction of of methoxy-

athe side-chain polyethoxy group polyethoxy units polyethoxy units

1-EO 70 10 0.496 0.131 0.077 0.3021-EO 70 25 0.351 0.093 0.054 0.2031-EO 50 25 0.199 0.053 0.031 0.1143-EO 80 10 0.657 0.288 0.233 0.5303-EO 70 10 0.563 0.247 0.200 0.4643-EO 70 25 0.414 0.181 0.147 0.3383-EO 50 10 0.387 0.170 0.137 0.3313-EO 50 25 0.245 0.107 0.087 0.2045-EO 70 10 0.615 0.335 0.294 0.5655-EO 70 25 0.466 0.254 0.223 0.4345-EO 60 10 0.527 0.287 0.252 0.4985-EO 60 25 0.370 0.202 0.177 0.3525-EO 50 10 0.439 0.239 0.210 0.4275-EO 50 25 0.287 0.156 0.137 0.2787-EO 50 25 0.324 0.200 0.182 0.3397-EO 40 10 0.391 0.242 0.220 0.4197-EO 30 10 0.298 0.184 0.168 0.3327-EO 30 25 0.175 0.108 0.098 0.1927-EO 20 25 0.111 0.069 0.062 0.1259-EO 30 25 0.197 0.133 0.123 0.229a Calculated by taking 1 monomer unit (e.g. ethoxy or siloxane) as ‘molar’ unit.

chain length z of the cross-linker. Taking intoTable 2account this limitation, the side-chain silastomersSolubility (c ), apparent distribution (k ) and permeability coeffi-o Dlisted in Table 1 were synthesized.cient (P) of salicylic acid in several side-chain silastomers at 208CThe syntheses with the 9-EO side-chain resulted inPolymer type Apparent Solubility Permeabilityfoils only, if the long cross-linker chain (z525) was

distribution c (%) coefficientoused and the content of the side chain was low 7 2 21coefficient P?10 (cm s )(s#30%). These products had exceptional proper-

1-EO 70 10 4.8 1.46 1.7ties. In spite of the long polyethoxy chain, the1-EO 70 25 3.3 1.01 1.6material did not incorporate water, and the solubility1-EO 50 25 2.1 0.64 1.2

of salicylic acid was relatively low (Table 2). The 3-EO 80 10 35.1 10.6 11.3apparent distribution coefficient (5.5 at 208C) and the 3-EO 70 10 30.8 9.75 8.0

3-EO 70 25 26.3 8.16 14.9activation energy of the partition process (¯9 kJ /3-EO 50 10 20.1 6.10 4.5mol) were correspondingly low. In consequence, the3-EO 50 25 15.5 4.77 10.0permeability of this foil was lower than expected on5-EO 70 10 36.5 11.3 13.1

the basis of its structural parameters and with respect 5-EO 70 25 27.3 8.46 17.2to the properties of the other side-chain silastomers. 5-EO 50 10 25.9 8.03 -

5-EO 50 25 16.0 4.96 6.3As the causes of these observations have not been7-EO 50 25 22.8 7.02 11.0investigated, the 9-EO silastomer is not further7-EO 40 10 28.0 8.67 6.6treated in the following.7-EO 30 10 15.4 4.74 3.37-EO 30 25 10.6 3.26 4.2

3.2. Thermal studies 9-EO 30 25 5.5 1.69 5.0Silgel 600 0.7 0.22 2.9

Some fundamental properties of the silastomers The aqueous donor phase for permeability and partition measure-with relevance to their application as diffusion-con- ments had a pH 2.9.


trolling matrices or membranes have been investi-gated. The identification of liquid-crystalline phasesin linear polymers and much more in cross-linkedmaterials is difficult [15,18]. The synthesized silas-tomers appear as isotropic substances under thepolarization microscope. The 5-EO and 7-EO typestake up water; an equilibrium is reached within about2 h. Double refraction and a few areas with texturescan be detected only in materials with high contentsof side-chains after storage in the swollen state forsome days. The restricted mobility of the side-chainswithin the polymer network may hinder their ar-rangement to lc-domains. For the same reason, phasetransition heats can be hardly detected by DSCmeasurements. The diagram of the 5-EO 70 25 typeonly has an endothermic peak at about 50–608Cwhich is broadened and flattened in comparison to

Fig. 2. Increase in weight of several side-chain silastomers bythe diagram of the corresponding biphenyl com-water uptake in dependence on the weight fractions of polyethoxypound substituted at the main chain [4-propenyloxy-units at 208C. The points marked by an arrow are not included in

(49-methoxy-pentaethoxy)-biphenyl]. Furthermore, the correlation calculation.the height of the peak decreases with increasingdegree of cross-linking, i.e. with decreasing substitu-tion by the side-chain. sent the silastomers 7-EO 50 25 and 5-EO 70 25

which have relatively high contents of voluminous3.3. Swelling by uptake of water side-chains so that their composition is close to the

stoichiometrical limits of the silastomer synthesisSwelling by uptake of water depends on the length (see above). In these cases, the incorporated amounts

of the methoxy-polyethoxy group and is influenced of water are relatively high in relation to the contentsby the portion of side-chains bound to the main- of ethoxy units. That deviating points are lacking, ifchain as well as by the temperature. Uptake of water thickness changes are measured, is not necessarily aby silastomer foils without side-chains (Silgel 600) contradiction, because volume expansion is limitedand with methoxy-monoethoxy and methoxy-tri- by the siloxane network. On the other hand, theethoxy groups could not be detected by measurement incorporated amount of water may be additionallyof weight or thickness. On the other hand, polymers dependent on the arrangement of the side-chainwith n55 and 7 show an increase in weight up to domains which is influenced by the degree ofabout 29% accompanied by an increase in thickness substitution with side-chains and by the cross-linkerof films up to about 18%. The uptake of water chain length.decreases with increasing temperature. A linearcorrelation of the measured values versus the con- 3.4. Drug solubility and distributiontents of side-chains, methoxy-polyethoxy groups orethoxy units, respectively, was tested. The best Using salicylic acid as model substance, its parti-results were obtained with the weight fractions of the tion between the polymer foils as organic phases andethoxy units. A coefficient of r50.9348 was found a buffered solution which was saturated with thisfor the increase in thickness. The increase in weight drug and had the same pH (2.9) as the donor phasein relation to the weight fraction of ethoxy units used for permeation experiments was measured. Asshows 2 deviating points marked by an arrow in Fig. undissociated molecules only penetrate the polymer2. The regression line drawn is calculated ignoring (see below), the partition is pH dependent, and thethese values (r50.9457). The deviating points repre- given apparent distribution coefficients are valid for


pH 2.9 (Table 2). The drug solubilities in the silastomers than for Silgel 600, a cross-linked silas-silastomers were calculated from the apparent dis- tomer without side-chains, and in part also highertribution coefficients (c 5k ?c ; c 5concentration than for the thermotropic lc-silastomers. The solu-o D w o

in the polymer at saturation; c 5concentration in bilities in cross-linked polysiloxanes at 208C are asw

the aqueous phase at saturation). Solubilities, c , and follows: Silgel 0.22%, thermotropic side-chain silas-o

partition coefficients, k , decrease with increasing tomers 2.2–4.8% [10], EO-side-chain silastomersD

temperature; this indicates that polar interactions are 0.64–11.3%. Some solubility values and apparentprimarily involved in the incorporation of salicylic distribution coefficients for distinct silastomer typesacid molecules into the polymer matrix just as in the are given in Table 2.uptake of water. Further support for this fact is given For Silgel 600 and the silastomer types with n51by correlation calculations of the apparent partition or 3 ethoxy units, the Arrhenius equation for thecoefficients (separately for t520, 30 and 608C) partition process log k 5log A2E /2.303 RT isD A

versus the weight fractions of side-chains, ethoxy valid in the temperature range from 20–808C (Fig.units or methoxy-polyethoxy groups. The highest 4). The activation energies calculated from thesecoefficients were obtained for the last mentioned plots are mostly greater than 14 kJ /mol and mainlygroups (r.0.93). In consequence of the propor- between 19 and 25 kJ /mol. The activation energy fortionality of both the parameters, c and k , the Silgel is about 7 kJ /mol for comparison. This valueo D

correlations with the saturation concentrations to the indicates that polar interactions between salicyliccontents of these polar groups gave the same results acid molecules and siloxane units of the network are(Fig. 3). This shows that interactions with aromatic involved to a minor degree only.ring systems and nonpolar groups are of minor The other polymer types do not show linearinfluence. This observation corresponds with the Arrhenius plots (Fig. 5). Regarding the silastomershydrogen-bonding parameter suggested by Chen and with n55 ethoxy units, the slopes of the curvesMatheson [19]. decrease with decreasing temperature, i.e. with in-

Generally, the distribution coefficients of salicylic creasing 1/T-values. The diagrams of the 7-EOacid in the polymer /buffer systems and its solubility polymers show a leap of the curves in the tempera-in the polymers are higher for the EO-side-chain

Fig. 3. Saturation concentration of salicylic acid in severalsilastomer types correlated to the weight fractions of the methoxy- Fig. 4. Arrhenius plots of the apparent distribution coefficients ofpolyethoxy group at 208C. some 1-EO and 3-EO silastomer types.


favours the perfect sink property of the receptorphase at pH 7.4. The substance transport in thequasi-steady state, therefore, was evaluated by thefirst-order kinetic equation:

Log c 5log c 2(P?A /2.303?V ?h )?t (c 5D D,0 D M D

diffusant concentration in the donor phase at time t;c 5concentration in the donor phase at time t50;D,0

A5cross section area; V 5volume of the donorD

phase; h 5thickness of the membrane). The diffu-M

sion coefficients were calculated from the relation-ship P5D?k .D

The temperature dependence of the diffusioncoefficients can be described with the Arrheniusequation, i.e. the diagrams log D versus 1 /T of allpolymer types are linear within the measured rangefrom 20–808C. In the tested silastomer matrices,whether able or unable to incorporate water, theD-values of salicylic acid have the same magnitudeFig. 5. Arrhenius plots of the apparent distribution coefficients of

28 28 2 21some 5-EO and 7-EO silastomer types. (2?10 to 9?10 cm s at 208C). In comparison,lower diffusion coefficients varying in a wider range

28 28 2 21(0.006?10 to 1?10 cm s ) have been found inture range between 30 and 608C. The reasons for this thermotropic lc-silastomers [11], whereas the highest

28 2 21exceptional behaviour may be that the side-chain value (42?10 cm s ) has been observed indomains exist in different arrangements or modi- Silgel. There is no linear correlation between thefications depending on the temperature, but weak diffusion coefficients and the weight or molar frac-evidence could be detected only, as discussed above. tions of the side-chains, methoxy-polyethoxy groups

Order or structure of the side-chain domains not or ethoxy units of the polymers. A tendency, how-investigated in this study does not only depend on ever, is observed that somewhat higher coefficientsthe length of the methoxy-polyethoxy groups, but are found in the silastomers with the longer cross-also on the length of the cross-linker chains and the linker (z525) than in those with z510, because thedegree of cross-linking. These parameters controlled latter form a denser network. Further, a low degreeby the chemical composition have important in- of cross-linking seems to favour somewhat higherfluences on the properties of the polymer materials, diffusion coefficients. These observations are inas these results show, which may be better under- accordance with the results of Gander et al. [20]. Itstood, when more information about the molecular is, however, not only the density of the networkarrangement of the polymer network and side-chain which mainly controls the permeability, because thedomains is available. side-chain domains seem to be used as the pre-

dominant pathway through the matrix by migrating3.5. Drug permeation measurements molecules.

In comparison to Silgel, the permeability coeffi-Diffusion measurements were run using a 2-com- cients of salicylic acid in 1-EO polymers are lower,

partment cell with a silastomer foil as membrane whereas the permeability values in all the other typeswhich is only passed by undissociated salicylic acid are higher (Table 2). This means that the highmolecules, but not by salicylate ions. This is con- permeability of the methoxy-polyethoxy side-chaincluded from diffusion experiments with donor phase silastomers is caused by the relatively high solubilitypH 7.4 (salicylic acid dissociation .99.9%) which of the salicylic acid. The Arrhenius diagrams logdid not show membrane permeation independent of P(permeability) versus 1 /T of the 1-EO and 3-EOthe ability of the polymer to take up water. This silastomer types are linear (Fig. 6); this is in


Fig. 6. Arrhenius plots of the permeability coefficients of some Fig. 7. Arrhenius plots of the permeability coefficients of some1-EO and 3-EO silastomer types. 5-EO silastomer types.

accordance with the linear slopes of corresponding deviation. The 5-EO 70 25 silastomer showed anplots of the diffusion and partition coefficients. endothermic process at 50 to 608C by DSC measure-Silastomers with the longer cross-linker chain and ments (see above). This means that the side-chainwith the longer methoxy-polyethoxy group tend to arrangement may change in this temperature range.give higher P-values. If materials with equal degrees If the degree of cross-linking is relatively low and,of substitution by side-chains (s-values) and length consequently, the side-chain content relatively high,of the methoxy-polyethoxy groups (n) are compared, the mobility and the closeness of the side-chains maythe activation energies are little higher in polymers be high enough to allow discernibly different molec-with the cross-linker chain z510 (steeper slopes of ular arrangements in dependence on the temperaturethe regression lines) than in those with z525 (Fig. influencing the incorporation of dissolved molecules.6). Permeation of a heterogeneous membrane depends

Silastomers of the 7-EO type which can be on the density of the penetrated phases and on thesynthesized only up to about 50% substitution with solubility of the migrating substance in these do-side-chains show approximately linear plots for log P mains [20,21]. Using side-chain silastomers, theversus 1 /T despite the deviation of the partition density can be adjusted by the degree of cross-coefficients described above. If the main chain of the linking and by the length of the cross-linker chain.5-EO polymer types is substituted by side-chains to a The solubility of the diffusant is dependent on thedegree of #50%, the Arrhenius diagrams of the degree of substitution with sidechains (their portionpermeability coefficients likewise show linearity. of the membrane mass) and on their polarity which isThis seems to be a consequence of the relatively low controlled by their structure. Polymorphic modifica-content of side-chains. On the other hand, 5-EO tions may occur in some side-chain silastomer typespolysiloxane types having a substitution degree of which are structurally characterized by a low degree70% show 2 regions with differing slopes (Fig. 7). of cross-linking, high content of side-chains, andOn account of P5D?k , this is caused by the sufficient length of polyethoxy chains (type 5-EO 70D

nonlinear plots of the partition coefficients reported 25 as an example from this work). Polymorphismabove. It is, therefore, the solubility of salicylic acid further influences the membrane permeability andin the polymer matrix which gives rise to the causes a special temperature dependence which may


¨[9] A. Euschen, Diffusion in Flussigkristallinen Silastomeren,be useful for varying drug release. The effects of all¨Doctoral Thesis, Saarbrucken, Germany, 1988.these parameters arise in an analogous manner, if the

[10] H. Loth, A. Euschen, Diffusion flux control by liquid-foils were used as matrices or reservoirs for drug crystalline side-chain polysiloxane elastomer foils, Makro-delivery. mol. Chem. Rapid Commun. 9 (1988) 35–38.

[11] H. Loth, A. Euschen, Liquid-crystalline side-chain silastom-ers, largely-modifiable diffusion media for the control ofdrug delivery, Drug Devel. Ind. Pharm. 16 (1990) 2077–Acknowledgements2095.

[12] Y.H. Bae, T. Okano, R. Hsu, S.W. Kim, ThermosensitiveThis work has been supported by Deutsche For- polymers as on–off switches for drug release, Makromol.

schungsgemeinschaft, Fonds der Chemischen Indus- Chem. Rapid Commun. 8 (1987) 481–485.[13] H. Finkelmann, H.J. Kock, G. Rehage, Liquid crystallinetrie, Wacker-Chemie GmbH, and Bayer AG to which

elastomers—a new type of liquid crystalline material, Ma-we express our thanks.kromol. Chem. Rapid Commun. 2 (1981) 317–322.

[14] J. Chatt, L.M. Vallarino, L.M. Venanzi, Olefin coordinationcompounds, part IV, diene complexes of platinum(II), the

References structure of Hofmann and von Narbutt’s [dicyclopentadie-ne(RO) Pt Cl], J. Chem. Soc. (1957) 2496–2505.

¨[15] B. Luhmann, H. Finkelmann, G. Rehage, Phase behaviour[1] R.J. Kostelnik (Ed.), Polymeric Delivery Systems, Gordonand structure of polymer surfactants in aqueous solution. Theand Breach Science Publishers, New York, London, Paris,occurrence of lyotropic nematic phases, Makromol. Chem.1976.186 (1985) 1059–1073.[2] Y.W. Chien, in: D.R. Paul, F.W. Harris (Eds.), Controlled-

[16] M.M. Feldstein, V.N. Tohmakhchi, L.B. Malkhazov, A.E.release Polymeric Formulations, American Chemical Socie-´Vasiliev, N.A. Plate, Hydrophilic polymeric matrices forty, Washington, D.C., 1976, p. 53.

enhanced transdermal drug delivery, Int. J. Pharm. 131[3] D.S.T. Hsieh, C.C. Chiang, D.S. Desai, Controlled release of(1996) 229–242.macromolecules from silicone elastomer, Pharm. Technol.

[17] K. Dusek, W. Prins, Structure and elasticity of noncrystallineJune (1985) 39–49.polymer networks, Adv. Polymer Sci. 6 (1969) 1–102.[4] J.W. McGinity, L.A. Hunke, A.B. Combs, Effect of water-

[18] C. Noel, Macroscopic structural characterization of sidechainsoluble carriers on morphine sulfate release from a siliconeliquid-crystal polymers, in: C.B. McArdle (Ed.), Sidechainpolymer, J. Pharm. Sci. 68 (1979) 662–664.Liquid Crystals, Glasgow 1989, pp. 159–195.[5] V. Carelli, G. Di Colo, Effect of different water-soluble

[19] Y. Chen, L.E. Matheson, Quantitative structure–transpor-additives on water sorption into silicone rubber, J. Pharm.tability relationship for the release of a series of substitutedSci. 72 (1983) 316–317.pyridines from a planar polydimethylsiloxane matrix, Int. J.¨[6] B.C. Lippold, P. Kurka, Fullstoffgesteuerte Freisetzung ausPharm. 94 (1993) 153–160.Suspensionsmatrix–Laminaten auf Basis von Silicon mit

[20] B. Gander, R. Gurny, E. Doelker, N.A. Peppas, Effect ofkonstanter Abgaberate, Act. Pharm. Technol. 35 (1989) 82–polymeric network structure on drug release from cross-84.linked poly(vinyl alcohol) micromatrices, Pharm. Res. 6[7] M. Rehula, Effect of polyethylene glycol on the pharma-(1989) 578–584.ceutical availability of papaverine hydrochloride from

[21] B.H. Lippold, J.F. Lichey, Arzneistofftransport durchsilicone matrices, Pharmazie 48 (1993) 126–128.lipophile Polymermembranen, Act. Pharm. Technol. 30[8] R. Toddywala, Y.W. Chien, Evaluation of silicone-based(1984) 140–154.pressure-sensitive adhesives for transdermal drug delivery. I:

Effect of penetrant hydrophilicity, J. Control. Release 14(1990) 29–41.

Documents

Methoxy-polyethoxy side-chain silastomers as materials controlling drug delivery by diffusion flux