`© 1991 Elsevier Science Publishers B.V. 0378-5173/91/$03.50
`ADONIS 037851739100144G
`
`LIP 02366
`
`153
`
`The influence of surfactants on drug release
`from a hydrophobic matrix
`
`M. Efentakis 1, H. Al-Hmoud 1, G. Buckton 2 and Z. Rajan 2
`Department of Pharmaceutical Technology, The School of Pharmacy, University of Athens, 104 Solonos Street, Athens 106 80 (Greece)
`and 2 Department of Pharmaceutics, The School of Pharmacy, University of London, 29-39 Brunswick Square,
`London WCIN lAX (U.K.)
`
`(Received 15 October 1990)
`(Modified version received 28 November 1990)
`(Accepted 3 December 1990)
`
`Key words: Eudragit RL 100; Flurbiprofen; Controlled release; Surfactant; Wetting; Dissolution
`
`Summary
`
`Hydrophobic matrices were prepared using Eudragit RL 100. Flurbiprofen was used as a model drug, with sorbitol as a diluent.
`The effect of adding each of five surfactants (sodium lauryl sulphate, sodium taurocholate, cetylpyridinium chloride, cocamidopropyl
`betaine (CDB) and cetrimide) individually to the matrix was investigated. To investigate the mechanism by which the rate of drug
`release was increased following the incorporation of surfactants, experiments were undertaken to assess the wettability of the
`different formulations, and to measure drug release in the presence of submicellar and micellar concentrations of the surfactants.
`Three mechanisms were proposed by which drug release could be increased following the addition of surfactants: improved wetting,
`solubilisation, and the dissolution of the soluble surfactants to form pores in the matrix. When the surfactant was added to the
`dissolution fluid, only one surfactant (CDB) did not result in an increase in drug release; for the other surfactants a minor increase in
`drug release was observed. Therefore, in most cases, wetting plays a small role in aiding dissolution. There was no significant change
`in release rate when the experiment was performed in the presence of either sub-micellar or micellar concentrations of the surfactants,
`thus solubilisation of the drug does not seem to be implicated in the drug release mechanism. The most significant increase in drug
`release rate was caused by incorporating the most soluble surfactants (sodium taurcholate and cetrimide) within the matrix. As the
`increase was significantly greater than could be explained by wetting alone, it must be concluded that for these matrix systems the
`major mechanism by which surfactants increase the dissolution rate is by the formation of pores to aid the access of the dissolution
`fluid and egress of the dissolved drug. It is also possible that the presence of the relatively concentrated surfactant solution in the
`wetted tablet would reduce interparticle adhesion and thereby speed drug release rate as a result of an increased disintegration.
`
`Introduction
`
`In a recent publication (Efentakis et al., 1990),
`the influence of incorporating either sodium lauryl
`
`Correspondence.' G. Buckton, Dept of Pharmaceutics, The
`School of Pharmacy, University of London, 29-39 Brunswick
`Square, London WC1N lAX, U.K.
`
`sulphate or sodium taurocholate into hydrophobic
`matrices was investigated. Different matrix sys-
`tems were studied, consisting of either Eudragit
`RS 100 or RL 100, a diluent (lactose, dextrose,
`sorbitol or Avicel PH-101) and a model drug (flur-
`biprofen). The Eudragit polymers are biocompati-
`ble, non-degradable acrylic resins, consisting of
`copolymers of acrylic and methacrylic resins. In
`all combinations of polymer and diluent, the ad-
`
`MYLAN Ex 1046, Page 1
`
`(cid:9)
`
`
`154
`
`dition of either of the two surfactants resulted in a
`significant increase in the dissolution rate, the
`explanation for this effect was described as being
`either due to a change in wetting of the tablet
`formulation by the dissolution fluid, or due to the
`production of channels within the product, caus-
`ing a wicking effect and thus increasing the access
`of the dissolution fluid (Efentakis et al., 1990).
`The purpose of this study is to investigate the two
`proposed mechanisms and to identify which is of
`greatest significance.
`A number of factors could influence the drug
`release profile from a hydrophobic matrix. Firstly,
`the wetting of the dosage form may be improved if
`a surfactant is present, to study this, products with
`each of five surfactants incorporated individually
`were compared with results in which the surfac-
`tant was omitted from the tablet. Secondly, when
`surfactant is present in the tablet (1% w/w) it may
`dissolve to form a solution of indeterminate con-
`centration in the micro-environment around the
`preparation; it is, therefore, necessary to investi-
`gate the effect of surfactant concentrations of less
`than, and greater than the CMC. The use of
`dissolution media containing sub-micellar and
`micellar surfactant solutions and tablets with no
`added surfactant will achieve this objective.
`Thirdly, it is possible that it is the solubility of the
`surfactant that is important, and that as it is
`dissolved it forms pores (or other disruptions) in
`the matrix, thus facilitating drug release. Finally,
`it is possible that a number of these mechanisms
`may occur simultaneously for different products.
`
`Materials and Methods
`
`Materials
`
`One diluent (sorbitol) and one polymer
`(Eudragit RL 100) were selected for study, with
`the same model drug as used before (flurbiprofen)
`(Efentakis et al., 1990). Five different surfactants
`were used, two were anionic (sodium salts), two
`were cationic and one ampholytic (cocamidopro-
`pyl betaine).
`Matrices were prepared using flurbiprofen
`(a gift of The Boots Co. Ltd), Eudragit (a gift
`
`of Rohm-Pharma), sorbitol (Merck), magnesium
`stearate (BDH), sodium taurocholate (ST) (Fluka),
`sodium lauryl sulphate (SLS) (BDH), cetylpyri-
`dinium chloride (CP) (Fluka), cocamidopropyl be-
`taine (CDB)(a gift of Goldsmith) and cetrimide
`(CET)(Serva). All chemicals used were reagent
`grade. Water was double distilled (for surface
`tension experiments) or reverse osmosis (for wet-
`ting experiments). Ethanediol (BDH) was used as
`a probe liquid for the wetting experiments.
`
`Methods
`
`Preparation of the tablets
`The Eudragit RL 100 was powdered in a ball
`mill and sieved through a 300 pm sieve. The
`tablets were formed from a mix of the powdered
`polymer (25%), drug (49%), sorbitol (25%) and
`magnesium stearate (1%). Six different batches of
`tablets were produced, one without surfactant,
`and the others with each of the five surfactants
`incorporated individually.
`The powders were compressed to prepare 500
`mg tablets on a single punch tablet machine
`(Korch-Erweka). The ratio between the diameter
`and thickness of the cylindrical flat faced tablets
`was between 0.7 and 0.9. The hardness of the
`tablets was controlled such that a breaking force
`of between 9 and 10 kg was required (Schleuniger-2
`hardness tester).
`
`Wettability
`Powder mixtures of the same composition as
`those used to produce the tablets were prepared
`and compacted into rectangular beams of nominal
`dimensions 4 cm x 1 cm X 1 mm. These beams
`were attached to a microbalance system, and used
`as a Wilhelmy plate to measure the contact angle
`of water and ethanediol on the formulation. The
`method was exactly as previously described (Zajic
`and Buckton, 1990).
`By measuring the contact angle formed by two
`liquids of known surface tension and polarity, it is
`possible to calculate the surface energy and polar-
`ity of the test solid (see for example, Zajic and
`Buckton, 1990). Having obtained the surface en-
`ergy (y) and the polar (p) and dispersion (d)
`components for a solid, it is possible to calculate
`
`MYLAN Ex 1046, Page 2
`
`
`
`spreading coefficients (Al2 of phase 1 over phase
`2) to indicate the extent of interaction between
`any two phases of interest. In this work, the
`surface energies of the formulations were de-
`termined, and then the spreading coefficients of
`water (subscript 1) over these surfaces (subscript
`2) were calculated, using Eqn 1:
`
`Yip • Y2
`
`d
`(cid:9) Y1
` • Y2
`
`yr ± yf
`
`yid ± Ili
`
`Yt
`2
`
`(1)
`
`The values for the spreading coefficients will re-
`flect the extent to which water will interact with
`the different formulations.
`
`Surface tension measurement
`Aqueous solutions of the surfactants were pre-
`pared by serial dilution, and their surface tensions
`were measured using a Du Nouy tensiometer
`(Kruss), in thoroughly cleaned glass apparatus.
`
`Dissolution testing
`The in vitro drug release from the formulations
`was assessed using a USP dissolution test appara-
`tus (Hanson model 72R), paddle method, with
`1000 ml of a pH 7.4 phosphate buffer (USP) at
`37 °C. The rotation speed was set at 100 rpm.
`
`54
`E 52
`50
`48-
`• 46
`0 -(7) 44
`a) • 42-
`co 40-
`tie 38-
`36-
`340
`
`0!5
`1.5
`Concentration (% W/V)
`Fig. 1. Surface tension as a function of concentration for the
`surfactants used. (*) ST, (o) CET, (+) SLS, (a) CP, (x )
`CDB.
`
`155
`
`0
`
`0
`
`70-
`
`60-
`
`50
`
`ce
`a)
`-F3 40-
`
`b-030
`
`20-
`
`10-
`
`0
`
`X
`A
`
`0
`
`X
`A
`
`0
`•
`X
`•
`
`X
`
`9
`2
`
`0
`0
`
`6
`4 (cid:9)
`2 (cid:9)
`Time (hours)
`Fig. 2. Drug release as a function of time for the matrix
`formulations in buffer. (A) No added surfactant (blank); (*)
`ST, (o) CET, ( + ) SLS, (A) CP, ( x ) CDB.
`
`8
`
`Samples were taken every hour, filtered and as-
`sayed at 248 nm using a Perkin Elmer Lambda
`series ultraviolet spectrophotometer.
`Dissolution was undertaken on tablets without
`surfactant present, and on those with 1% of the
`different surfactants incorporated in the product
`(i.e. 1% SLS or 1% CET, etc.). Also the tablets
`without added surfactant were studied using dis-
`solution fluids with added surfactant at concentra-
`tions of 0.25 and 1.25% w/v. For all the surfac-
`tants the critical micelle concentration appeared
`to fall in the range 0.5-1.0% w/v (Fig. 1), thus
`0.25% will be below the CMC and 1.25% will be
`above the CMC.
`All experiments were performed in triplicate
`and the average value was recorded.
`
`Results
`
`The dissolution profiles for the tablets with
`incorporated surfactant are presented in Fig. 2.
`The release data for the tablets with incorporated
`surfactant (in buffer), and without incorporated
`surfactant (in buffer and surfactant solutions) are
`presented in Table 1.
`
`MYLAN Ex 1046, Page 3
`
`
`
`156
`
`TABLE 1
`
`Percentage release of drug from the tablets (cid:9)
`
`Time (h)
`
`1
`
`10
`
`14
`11
`12
`
`19
`12
`13
`
`16
`10
`10
`
`9
`9
`9
`
`9
`11
`12
`
`2
`
`15
`
`21
`17
`18
`
`29
`18
`19
`
`22
`15
`16
`
`17
`14
`15
`
`16
`17
`18
`
`No surf.
`
`SLS (a)
`(b)
`(c)
`
`ST (a)
`(b)
`(c)
`
`CET (a)
`(b)
`(c)
`
`CP (cid:9)
`
`(a)
`(b)
`(c)
`
`CDB (a)
`(b)
`(c)
`
`3
`
`20
`
`28
`24
`25
`
`36
`24
`25
`
`30
`22
`24
`
`28
`19
`20
`
`22
`23
`24
`
`4
`
`26
`
`35
`31
`32
`
`46
`32
`33
`
`38
`29
`30
`
`34
`24
`25
`
`30
`29
`30
`
`5
`
`30
`
`40
`36
`37
`
`56
`37
`38
`
`45
`34
`35
`
`38
`29
`29
`
`35
`35
`35
`
`6
`
`35
`
`44
`40
`41
`
`60
`40
`42
`
`51
`38
`39
`
`42
`33
`33
`
`39
`39
`39
`
`7
`
`39
`
`48
`43
`45
`
`63
`43
`46
`
`56
`41
`41
`
`45
`37
`38
`
`42
`42
`42
`
`8
`
`43
`
`52
`46
`48
`
`67
`47
`50
`
`61
`44
`44
`
`48
`41
`43
`
`45
`44
`45
`
`(a) Surfactant incorporated in tablet (1% w/w of tablet).
`(b) Surfactant 0.25% w/v in dissolution fluid, no surfactant in
`tablet.
`(c) Surfactant 1.25% w/v in dissolution fluid, no surfactant in
`tablet.
`
`The contact angles, the surface energies and
`dispersion and polar components of surface en-
`ergy, of the tablets with added surfactants are
`presented in Table 2, as are the spreading coeffi-
`
`TABLE 2
`
`Contact angles measured on the formulations containing the
`surfactants, the surface energies calculated for the formulations,
`and the spreading coefficients of water over the different formula-
`tions
`
`8Nv
`
`48
`54
`59
`68
`74
`
`0E.
`(°)
`40
`39
`35
`35
`38
`
`7 p
`1d
`(mN/m)
`
`52.3
`48.2
`45.7
`41.8
`39.8
`
`36.6
`30.7
`25.3
`18.0
`13.7
`
`15.7
`17.5
`20.4
`23.8
`26.1
`
`X12
`
`17.0
`19.4
`19.4
`16.9
`13.2
`
`SLS (a)
`ST (a)
`CET (a)
`CDB (a)
`CP (a)
`
`Reproducibility of the contact angle data was at worst ±3'.
`(a) Results for formulation with 1.0% w/w surfactant added.
`W, water; E, ethanediol.
`
`cients for water over the surface of these formula-
`tiOnS.
`
`Discussion
`
`The drug release profiles obtained with all of
`these formulations follow a pseudo zero order
`release profile for about 5 h, and then deviate to
`form another near linear release profile which is of
`a slower rate than the initial release. For the
`formulations that have been investigated here, the
`release profiles remain parallel (Fig. 2) throughout
`the 8 h experiment. The experimental error associ-
`ated with the dissolution results was extremely
`small, such that the replicate determinations where
`almost superimposable.
`Two possible mechanisms have been postulated
`as to why surfactants increase the rate of drug
`release from matrix formulations (e.g. Desai et al.,
`1965; Dakkuri et al., 1978). Firstly, it is possible
`that the surfactant lowers the interfacial tension
`between the product and the dissolution fluid,
`secondly, it is possible that the surfactant acts as a
`wicking agent, causing the fluid to enter the dosage
`form, the surfactant may then dissolve and form
`pores (or other disruptions) from which the drug
`release may be effected (Dakkuri et al., 1978). In a
`previous study, dissolution profiles of flurbiprofen
`release from matrix systems containing Eudragit
`polymers demonstrated that surfactants increase
`drug release, but the mechanism(s) for this
`was/were not investigated (Efentakis et al., 1990).
`In the introduction, four possibilities have been
`postulated as mechanisms by which drug release
`from matrix systems can be increased due to the
`presence of a surfactant.
`In this study we have investigated the effect of
`five surfactants, as expected the anionic/cationic
`nature of these excipients does not seem to have a
`major effect on the release data (e.g. ST (anionic)
`and CET (cationic) have the fastest release rates,
`and CP (cationic) and SLS (anionic) have slower
`release rates). In circumstances where anionic/
`cationic interactions are expected it is possible to
`achieve a reduction in the dissolution rate by the
`addition of a charged surfactant (e.g. anionic
`surfactant/cationic drug (Feely and Davis, 1988)).
`
`MYLAN Ex 1046, Page 4
`
`(cid:9)
`(cid:9)
`
`
`The incorporation of the individual surfactants
`into the formulation results in a range of drug
`release profiles (Fig. 2 and Table 1). As mentioned
`above, there is a change in rate of release at — 5 h
`for each formulation. The ranking of the release
`rates can be made from Fig. 1 and going from the
`fastest to slowest release is: ST, CET, SLS, CP,
`CDB/no surfactant.
`It is generally regarded that the surfactants will
`result in improved wettability of the surface of the
`preparation, and that this is the major factor
`causing the increased release rate. If wetting of the
`surface was of greatest significance, then a tablet
`without added surfactant, which was allowed to
`release drug in the presence of a surfactant solu-
`tion, should have a very similar release rate to a
`tablet in which that surfactant was incorporated
`into the matrix. The results in Table 1 allow a
`comparison of the dissolution data for tablets in
`which the different surfactants have been incorpo-
`rated, with those tablets which have been disso-
`luted in the presence of surfactant solutions. For
`CDB there was no difference between the data for
`tablets with incorporated surfactant, and those in
`which the surfactant was added to the dissolution
`fluid. The data for CDB are all slightly higher
`than the product without surfactant added
`(hereafter termed the 'blank') up to 5 h, after
`which the dissolution rate fell significantly and at
`8 h was indistinguishable from the blank. These
`data indicate that the increased dissolution rate
`caused by CDB (either incorporated in the prod-
`uct, or in the dissolution fluid) is due to increased
`wetting of the tablet surface. The increase in rate
`over the initial period is probably entirely due to
`improved wettability, after about 5 h, the slower
`release rate is probably due to the dissolution
`front receding from the surface of the tablet, into
`the body of the matrix. Drug release in the later
`stages will be linked to the diffusion of the dis-
`solved drug away from the dissolution front. The
`addition of CDB does not aid this diffusion pro-
`cess. Dissolution experiments that were performed
`in the presence of surfactants other than CDB,
`showed similar responses, that is the amount dis-
`solved was higher than the blank at 5 h, and
`rather more similar to the blank at 8 h.
`The addition of CP to the dissolution fluid did
`
`157
`
`not result in any increase in the drug release from
`the blank tablets, however, when CP was incorpo-
`rated there was a slight increase in release. For
`this surfactant it can be concluded that wetting
`does not play a significant role in the dissolution
`process, and the minor acceleration with the incor-
`porated CP must be due to the surfactant dissolv-
`ing and forming pores/channels, thus increasing
`the effective surface area by a method other than
`wetting. The results in Table 2 confirm these find-
`ings, the formulation with incorporated CP has
`the highest contact angle with water, and the
`lowest spreading coefficient (i.e. is the most hy-
`drophobic of the surfaces studied).
`When the other surfactants (SLS, ST, CET)
`were incorporated in the matrices, they produced
`different dissolution profiles to those obtained
`when the blank was studied in the corresponding
`surfactant solutions. Furthermore, the results ob-
`tained in the surfactant solutions were different to
`the results for the blank in buffer alone. There was
`no significant difference between the drug release
`following dissolution in sub-micellar or micellar
`concentrations of either SLS, ST, CET or CDB.
`This leads to the conclusion that improved wetting
`of the tablet surface is achieved with these four
`surfactants (but not CP), and that the potential
`resultant increase in drug release due to improved
`wetting is finite. After adequate wetting of the
`surface has been achieved, there must be further
`mechanisms which facilitate more rapid dissolu-
`tion i.e. there must be a soluble component in the
`matrix which will dissolve easily to form pores, or
`to disrupt the matrix in some other fashion. The
`similarity between results obtained in sub-micellar
`and micellar surfactant solutions provides further
`evidence to support this hypothesis: the drug re-
`lease is controlled by the permeation through the
`matrix, and is not, after a certain point, affected
`by the wetting of the surface, and is not affected
`by the solubility of the drug, i.e. the potential for
`solubilisation does not increase the dissolution
`rate to any serious extent. The data presented in
`Table 2 demonstrate that the spreading coeffi-
`cients of water over the tablets with incorporated
`surfactant fall in a rank order that correlates with
`the rank order of drug release from these formula-
`tions. As has been explained above, this is of
`
`MYLAN Ex 1046, Page 5
`
`
`
`158
`
`interest, but does not form the major reason for
`the changes in dissolution.
`The aqueous solubilities of these surfactants
`are: ST 2 in 1, CET 1 in 2, SLS and CDB 1 in 10
`and CP 1 in 20 (data from Martindale, except for
`CDB which was measured). Thus the release rates
`correlate to a reasonable extent with the solubili-
`ties of the surfactants; the only exception is that
`of CDB which neither improves wettability signifi-
`cantly, nor dissolution performance. The most sig-
`nificant increases in drug release were for the
`tablets with added ST (solubility 2 in 1) and CET
`(solubility 1 in 2), it follows that the major in-
`fluence on the drug release profile is the dissolu-
`tion of the soluble surfactants to produce pores
`(or disruptions) in the matrix. If pores are formed,
`this will aid the access of dissolution fluid, and the
`egress of dissolved drug, however, in a product
`with a considerable quantity of soluble matter, it
`is likely that the dissolving surfactant will not alter
`the number of pores significantly; the effect of the
`surfactant dissolving may be to produce a high
`local concentration of surfactant solution in the
`matrix, which in turn may result in disruption to
`the matrix (perhaps in the form of reduced inter-
`particle adhesion).
`
`contact angle for water intermediate between that
`of CDB (68° ) and CP (74° )) any further im-
`provement in wetting does not result in faster
`dissolution rates. This effect is probably due to the
`dissolution being controlled by diffusion within
`the matrix.
`The inclusion of surfactants within the formu-
`lation results in increases in the extent of drug
`release which cannot be explained purely on the
`basis of wetting. The major influence is the solu-
`bility of the surfactant, the more soluble the
`surfactant, the more rapid the drug release. This
`effect is due to the formation of pores (or disrup-
`tions) in the matrix which allow access for the
`dissolution fluid, and aid removal of dissolved
`drug, perhaps by a mechanism which involves
`reduced interparticle adhesion.
`
`Acknowledgement
`
`The British Council are gratefully acknowl-
`edged for the provision of a twinning grant.
`
`References
`
`Conclusions
`
`There are three possible mechanisms by which
`drug release can be accelerated following inclusion
`of a surfactant in the formulation, these are im-
`proved wettability, solubilisation and the forma-
`tion of pores (or disruptions) in the matrix due to
`the surfactant dissolving. For the matrix system
`investigated here (Eudragit RL 100/ sorbitol) the
`results indicate that solubilisation is not a signifi-
`cant factor.
`The effect of improved wetting of the tablet
`surface by the dissolution media is of some minor
`significance. However, once a certain degree of
`wetting has been achieved (corresponding to a
`
`Dakkuri, A., Schroeder, H.G. and DeLuca, P.P., Sustained
`release from inert wax matrixes. II: Effect of surfactants on
`tripelennamine hydrochloride release. J. Pharm. Sci., 67
`(1978) 354-357.
`Desai, S.J., Simonelli, A.P. and Higuchi, W.I., Investigation of
`factors influencing the release of solid drug dispersed in
`inert matrices. J. Pharm. Sci., 54 (1965) 1459-1464.
`Efentakis, M., Al-Hmoud, H. and Choulis, N.H., Effects of
`additives on flurbiprofen controlled release preparation.
`Acta Pharm. Technol., 36 (1990) 237-239.
`Feely, L.C. and Davis, S.S., Influence of surfactants on drug
`release from hydroxypropylmethylcellulose matrices. Int. J.
`Pharm., 41 (1988) 83-90.
`Martindale, The Extra Pharmacopoeia, 28th Edn, The Phar-
`maceutical Press, London.
`Zajic, L. and Buckton, G., The use of surface energy values to
`predict optimum binder selection for granulations. Int. J.
`Pharm., 59 (1990) 155-164.
`
`MYLAN Ex 1046, Page 6
`
`