SURFACTANT SYSTEMS
`
`Their chemistry, pharmacy and biology
`
`D. Attwood
`
`Department of Pharmacy
`University of Manchester
`
`A. T. Florence
`
`Department of Pharmacy
`University of Strathclyde
`
`LONDON
`
`NEW YORK
`
`CHAPMAN AND HALL
`
`|PR2015-01099
`
`|PR2015-01097
`
`|PR2015—01100
`|PR2015-01105
`
`Lupin EX1179
`Page 1
`
`Page 1
`
`

`
`First published 1983 by
`Chapman and Hall Ltd
`11 New Fetter Lane, London EC4P 4EE
`
`Published in the USA by
`Chapman and Hall
`
`733 Third Avenue, New York NY 10017
`
`© 1983 D. Attwood and A. T. Florence
`
`Softcover reprint of the hardcover lst edition 1983
`
`ISBN-13: 978-94-009-5777-0
`
`e-ISBN-13: 978-94-009-5775-6
`
`D01: 1 0. 1 007/978-94-009-5775-6
`
`All rights reserved. No part of this book may be reprinted, or
`reproduced or utilized in any form by any electronic, mechan-
`ical or other means, now known or hereafter invented, includ-
`
`ing photocopying and recording, or in any information storage
`and retrieval system, without permission in writing from the
`Publisher.
`
`British Library Cataloguing in Publication Data
`
`Attwood, D.
`Surfactant systems.
`1. Surface active agents
`1. Title
`II. Florence, A. T.
`668’.l
`TP994
`
`Page 2
`
`Page 2
`
`

`
`D. Attwood et al., Surfactant Systems
`© D. Attwood and A. T. Florence 1983
`
`Page 3
`
`

`
`Biological implications of surfactant presence
`
`-
`
`389
`
`Solubilizarion
`of active species
`
`Drug in
`
`formulorior\\
`
`l Drug in
`
`solution
`
`Membrane
`
`Drug in
`
`blood
`
`, Site of
`
`/ action
`
`_ aim.-nafion
`
`Effect on _
`d9°99feQ°r|9"
`and dI$SO|Ul'lO|'1
`
`Effect onrnembrane
`permeability or
`infegrify
`
`Effect on
`binding to
`recepror (7)
`
`Prevention of precipitation
`0" C°D"'°' 0? Dreclpilflfion
`
`Effecr on drug
`metabolizing enzymes (?)
`
`influence on drug absorption and activity.
`Figure 7.1 Possible sites of surfactant
`Utilization of a drug involves its release from the formulation, its solution in the body
`fluids, and its passage through barrier membranes into the systemic blood stream before
`transport into tissues and eventual arrival at the target organ. Release of poorly soluble
`drugs from tablets and capsules for oral use may be increased by the presence of
`surfactants, which may decrease the aggregation of the drug particles and therefore increase
`the area of particle available for dissolution. The lowering of surface tension may also be a
`factor in aiding the penetration of water into the drug mass; this wetting effect is operative
`at low concentrations. Above the critical micelle concentration (CMC) the increase in the
`saturation solubility of the drug substance by solubilization in the surfactant micelles can
`result in more rapid rates of drug solution. Where dissolution is the rate-limiting step in the
`absorption process, as it is with many poorly soluble drugs, an increase in rate of solution
`will increase the rate of drug entry into the blood and may affect peak blood levels. Very
`high concentrations of surfactant can decrease drug absorption by decreasing the chemical
`potential of the drug. This results when surfactant is present in excess of that required to
`solubilize the drug.
`
`since, but nonetheless the literature tends to be confused. The observed influences
`of surfactants depend on the concentration of the agent used (which is difficult to
`assess when the formulation has been administered to man or intact animal) and
`even in model systems this leads to complications in elucidating effects especially
`when the surface-active agent exerts several actions simultaneously. Much of the
`confusion in the literature on this subject arises from discussion of the influence
`of different concentrations of surfactant, and from attempts to generalize on the
`action of varied surfactants on many different types of biological membrane. As
`with the physical effects noted above, distinct changes in the activity of the
`surfactant can frequently be observed on increase of surfactant concentration.
`This can be demonstrated by experiments in model systems, for example, in
`goldfish immersed in solutions of drug and surfactant [3—5]. Low concentrations
`of polysorbate 80 increase the absorption of secobarbitone; concentrations above
`the CMC decrease absorption. Similarly, the influence of surfactant structure and
`properties on drug absorption can also be demonstrated with the goldfish; some
`of these experiments will be discussed later in this chapter.
`
`Page 4
`
`Page 4
`
`

`
`390
`
`-
`
`Surfactant systems
`
`7.2 Effect of surfactants on dissolution of drugs
`
`It is readily apparent that the rate of solution of poorly soluble drugs can be
`increased by the presence of surfactants in the dissolution medium. Most
`experiments have been carried out in vitro; the effect in vivo is more complex with
`the concomitant dilution of the surfactant by a complex medium, the absorption
`of the surfactant itself and the adsorption of other substances onto the dissolving
`particles.
`
`Surfactant adsorption on to hydrophobic drug particles below the critical
`micelle concentration can aid wetting of the particles and consequently increase
`the rate of solution of particulate agglomerates [6—10]. Surfactants may be
`incorporated into solid dosage forms [11] so that their solubilizing action comes
`into play as the disintegration process starts and water penetrates to form a
`concentrated surfactant through lowering of surface tension solution around the
`drug particles or granules. Both facilitation of wetting and solubility increase will
`aid dissolution of the drug. Finholt and Solvang’s results [12] on the dissolution
`in vitro of phenacetin and phenobarbitone in the presence of polysorbate 80
`show clearly the influence of surface tension (Fig. 7.2). The solubility of
`phenacetin is little affected by the concentrations of polysorbate 80 used and thus
`enhanced wetting is the primary cause of improved dissolution rates, a result in
`accord with the finding that sodium lauryl sulphate (NaLS) increased the rate of
`solution of salicylic acid from compressed tablets owing to better solvent
`penetration into the tablets and granules [13]. Finholt and Solvang [12]
`determined the pH and surface tension of gastric juice from 27 patients. Surface
`tension ranged between 35 and 50 mN In‘ 1 and pH between 1 and 7.5, and was
`independent of secretion rate. Such are the complications of the in vivo
`environment and the problems of determining the effect of synthetic surfactants
`on dissolution rates in vivo; the rate of solution of a drug such as phenobarbitone
`is significantly higher in diluted gastric juice than in 0.1N HCl because of the
`
`20
`
`_\. U1
`
`
`
`Time(min) 8
`
`
`
`40
`
`1 70
`so
`.
`so
`Surface tension (mN m‘ )
`
`Figure 7.2 Relationship between the surface tension of the dissolution medium and the
`time necessary for dissolution of 100mg phenacetin. Dissolution media: 0.1 N HCl
`containing different amounts of polysorbate 80. From Finholt and Solvang [12] with
`permission.
`
`Page 5
`
`Page 5
`
`

`
`Biological implications of surfactant presence
`
`-
`
`391
`
`difference in surface tension. In addition, the amount of a soluble salt such as
`phenobarbitone sodium dissolved in diluted gastric juice at 1 h has been shown to
`be considerably increased, presumably because the precipitation of the free acid is
`reduced by components in gastric fluid. Nevertheless increased absorption of
`paracetamol has been observed in vivo [14]. Enhanced absorption of digoxin and
`digitoxin [15] and sulphadiazine and sulphisoxazole [16] have been ascribed to
`increased dissolution rates of these drugs brought about by the incorporation of
`surfactants into the formulation. The effect of poloxamer 188 and dioctyl
`sulphosuccinate (DOSS) on absorption of sulphisoxazole from rat intestinal loops
`is shown in Table 7.1, and the influence of these surfactants on dissolution rate
`shown in Fig. 7.3. Poloxamer 188 and DOSS are both used below their critical
`micelle concentrations, at concentrations likely to be found in vivo where they are
`used as faecal softeners in laxative products. In some systems negligible effects are
`noted below the surfactant CMC. Such is the case with hydrocortisone [17];
`neither polysorbate 80 nor two Solulan surfactants (Solulan 25 and 16, American
`Cholesterol Products Inc., USA) increased the dissolution rate of this steroid
`until their respective CMCs were exceeded. However, the solubility of hydro-
`cortisone was increased much less than the increased solution rate would imply
`suggesting that the solubility increase was not of major importance in this case.
`Short et al. [8] have also considered the effect of surfactant on hydrocortisone
`dissolution. An increased dissolution rate constant below the CMC of poly-
`sorbate 80 is observed, this decreasing just above the CMC; Short et al. suggest
`that this might be related to a surface tension effect, the maximum in dissolution
`rate constant coinciding with the surface tension minimum of the polysorbate. A
`minimum surface tension around the CMC value implies the presence of surface-
`active impurities [18] which may adsorb preferentially on the drug particles
`decreasing dissolution rate.
`Concentrations of polysorbate 20 well in excess of the CMC have been used by
`Collett and Rees in their studies on salicylic acid dissolution [10, 19]. Dissolution
`rates were measured over a pH range from 1.0 to 4.0; the dissolution rate increases
`very slowly above 12 ‘Z, surfactant (Fig. 7.4) but there was no evidence of a
`decreased dissolution rate such as found by Parrott and Sharma [20], e.g. above
`
`Table 7.1 Effect of poloxamer 188 and dioctyl sodium
`sulphosuccinate on the absorption of sulphisoxazole from
`rat intestinal loops*
`
`Surfactant
`
`Concentration,
`"/3 w/v
`
`Dose absorbed,
`"/A +_- S.D.
`
`Control
`poloxamer 188
`
`Dioctyl sodium
`sulphosuccinate
`
`_—
`0.01
`0.10
`0.01
`0.10
`
`45.3 1- 6.5
`56.1 i 3.9
`57.3 i 10.1
`53.9 1 9.4
`55.0 i 8.4
`
`* Values represent mean of 6 animals.
`From [16].
`
`Page 6
`
`Page 6
`
`

`
`'9 0
`
`(mgml:)<501
`
`
`
`sulphisoxazoleconcentration 9 U1
`
`(G)
`
`'.° U‘
`
`5
`
`1O
`
`15
`
`2O
`(min)
`
`25
`
`30
`
`l
`
`'3’ o
`
`-1T‘) U1
`
`
`
`SuIphiso>Eazo|econcentration Qrtlgml
`
`U10
`
`('3)
`
`5
`
`10
`
`20
`15
`(min)
`
`25
`
`30
`
`Figure 7.3(a) Effect of poloxamer 188 on sulphisoxazole dissolution I, control; CI, 0.001 "A;
`A, 0.01 %; and O, 0.1 %. (b) Effect of dioctyl sodium sulphosuccinate on sulphisoxazole
`dissolution. 0, control; CI, 0.001 %; A, 0.01 %; and O, 0.1 %. From Reddy et al. [16] with
`permission.
`
`rateconstant(kg5'1x108)
`Dissolution
`
`20
`16
`12
`8
`4
`Polysorbare 20 ('I. w/v)
`
`Figure 7.4 Plot of dissolution rate constants (kgs" x 10") of salicylic acid against
`concentration of polysorbate 20 at several pH values vpH 10, <>pH 2.0,! pH 3.0, DpH
`4.0. From Rees and Collett [10] with permission.
`
`Page 7
`
`Page 7
`
`

`
`Biological implications of surfactant presence
`
`-
`
`393
`
`12% polysorbate 80 with benzoic acid. Collett and Rees [19] suggest that the
`decreased dissolution rates are not a function of the viscosity of the dissolution
`medium but rather an artefact due to lack of pH control in the system, the
`decreased pH resulting from the dissolution of benzoic acid leading to decreased
`solubility and thus solution rate. However, such an explanation cannot be put
`forward to discuss the decreased rate of solution of griseofulvin [21] at high
`concentrations of non-ionic surfactant.
`
`7.2.1 Theoretical approaches to dissolution rates
`in high concentrations of surfactant
`
`Higuchi [22] has analysed the dissolution process in the presence of micellar
`solutions. His equations predict that the effect of surfactant on dissolution rate
`will be less than predicted by the Noyes—Whitney equation on the assumption of
`increased bulk solubility. The Noyes—Whitney relation in the form
`
`3-: = kA(cs-c)
`
`(7.1)
`
`shows the rate of change of concentration of solute, c, related to its surface area,
`A, and its saturation solubility, cs. When c, > c there is a direct proportionality
`between the rate of solution, dc/dt and cs. The studies discussed above have
`shown that this is frequently not observed, as clearly demonstrated in Fig. 7.5.
`Higuchi [24] assumes that an equilibrium exists between the solute and the
`solution at the solid—liquid interface and that the rate of movement of solute into
`the bulk is governed by the diffusion of the free and solubilized solute across a
`stagnant diffusion layer. Drugs solubilized in micelles will have a lower diffusion
`coeflicient than free drug so that the effect of additive on dissolution rate will be
`related to the dependence of dissolution rate on the diffusion coefficients of the
`diffusing species, and not to their solubilities, as suggested by simple interpret-
`ation of Equation 7.1. The effective diffusion coefficient (Deg) is given by [24]:
`
`D
`
`eff
`
`= DfC'f+ Dm('m
`——————,
`Cs+Cm
`
`(
`
`7.2
`
`)
`
`subscripts f and m referring, respectively, to the free and micellar drug; c m is thus
`the increase in solubility due to the micellar phase. This leads to the following
`equation for dissolution of a solid at constant area A and under sink conditions,
`i.e.cs>c,
`
`DfCf DmCm
`dC'
`——= —-
`
`[11 + h
`
`dt
`
`]
`
`7.3
`
`(
`
`)
`
`where h is the diffusion layer thickness. Substituting Equation 7.2 into Equation
`7.3 gave, where C, is the total solute concentration,
`
`—~ = Dene./h.
`
`(7,4)
`
`Page 8
`
`Page 8
`
`

`
`394
`
`-
`
`Surfactant systems
`
`‘D O
`
`9‘ o
`
`*9 o
`
`9’ o
`
`_/
`
`)
`
`D
`
`(
`
`1
`
`“P 0 Ratioofdissolution
`rateandsolubility
`
`
`Ratioofdissolutionrateandsolubility
`
`2
`
`3
`
`4
`
`5
`
`6
`
`OO
`Concentration of sodium Iciuryl sulphate
`(°/o,w/V)
`
`/
`
`‘F o
`
`9’ o
`
`_.NOO
`
`O
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`Concentration of sodium lciuryl sulphate
`
`(°/0| VV/V)
`
`Figure 7.5(a) Ratio of dissolution rates and solubilities of sulphamethizole in surfactant
`solution to those in distilled water. (b) Ratio of dissolution rates and solubilities of
`sulphadiazine in surfactant solution to those in distilled water.
`A: ratio of dissolution rate constant.
`0: ratio of solubility.
`From Watari and Kaneniwa [23] with permission.
`
`However, both Collett and Rees [19] and Gibaldi et al. [25] find that dissolution
`rate is proportional to the effective diffusion coefficient raised to the power
`0.5 to 1.0,
`thus placing in some doubt
`the diffusion coeflicients of
`salicylic acid calculated assuming Equation 7.4 to hold [20]. The lack of
`agreement between the dissolution data and the predictions of Equation
`
`Page 9
`
`Page 9
`
`

`
`Biological implications of surfactant presence
`
`°
`
`395
`
`7.4 leads to the conclusion that alternative models are required. A ‘film-
`penetration’ model incorporating the surface renewal concepts of Danckwerts
`[26] has been proposed [27]. In this, mass transfer from the surface is believed to
`occur by two simultaneous processes—one involving a stagnant film in which
`steady state molecular transfer occurs, and the other encompassing non-steady
`state mass transfer by eddy formation in the surface layer. The film-penetration
`model predicts a dependence of dissolution rate on diffusion coefficient with an
`exponent between 0.5 and 1.0 [25, 27].
`Predictions of dissolution rate may be made using diffusion coeflicients of the
`solutes in their solubilized state by applying the Stokes—Einstein equation.
`
`D_ RT
`~—6m1NA 3
`
`41tNA
`3M5 ’
`
`75)
`(-
`
`where D is diffusion coefiicient, R is the molar gas constant, T is the absolute
`temperature, 71 is the viscosity of the solvent in poise, E is the partial specific
`volume of the micelles, M is the micellar molecular weight, and N is Avogadro’s
`number. More direct measurements of Dm are now possible by photon
`correlation spectroscopy and this should lead to a better analysis of dissolution
`models for solubilizing systems.
`Elworthy and Lipscomb [28] considered dissolution to consist of two
`processes occurring simultaneously:
`
`(1) a zero order reaction for the transfer of griseofulvin molecules from the solid
`surface into the solution, with rate constant k,;
`(2) a first order reaction for the deposition of solute from solution to solid
`surface, with rate constant k2.
`
`The rate of increase of concentration in solution:
`
`dc
`— = k — k
`dt
`1
`
`.
`
`2C
`
`The solution to this equation with the condition that at I = 0, c = 0 is
`
`k
`c = —1—(1-e”‘2').
`k2
`
`Expanding the exponential term and rearranging gives
`
`c
`k
`_: l_
`I
`
`k1k§t3
`k1k2t+k1k§t2
`_
`2
`6 24
`
`7.6
`
`l
`
`(
`
`(7.7)
`
`At fairly early times in the dissolution process, terms in t2 and t3 etc. can be
`neglected giving:
`k k t
`1
`2
`2
`
`(7.8)
`
`—:—= k, —
`
`.
`
`A plot ofc/t versus t will have an intercept k, , and a slope kl k2/2, enabling both
`constants to be evaluated. Trial calculations show that Equation 7.8 gives 1 ‘Z,
`
`Page 10
`
`Page 10
`
`

`
`396
`
`-
`
`Surfactant systems
`
`error in c compared to the exact Equation 7.7 provided that the kzt term does not
`exceed 0.25.
`
`Equation 7.8 reduces to the Noyes—Whitney equation. When equilibrium is
`reached, i.e. a steady state between dissolution and redeposition,
`
`d'
`
`&§=O= kl _k2Csa
`
`where cs is the saturation solubility,
`
`cs = kl/k2,
`
`(7.9)
`
`and from Equation 7.7
`
`or,
`
`c = cs(1—e"‘*‘)
`
`k2 =1ln( C‘
`
`t
`
`cs-c
`
`(7.10)
`
`which is the more usual form of the Noyes~Whitney equation. The rate constant
`of Equation 7.6 thus appears to be the first order constant arising in the
`consideration of the dissolution—redeposition process. Equation 7.8 is useful if
`the saturation solubility is not known; when it is, Equation 7.9 can be used to
`evaluate one constant when the other has been determined from Equation 7.8 or
`7.10.
`
`A result of this analysis is shown in Fig. 7.6 for the cetomacrogol—griseofu1vin
`system [29]. The considerable effect of stirring rate on the dissolution rate of
`the powdered drug is seen, leading to the conclusion that it is necessary to choose
`
`18
`
`(0)
`
`11.
`
`10
`
`6
`
`2
`
`go
`
`°
`
`(b)
`
`12 1-5
`
`3
`
`1-2
`
`1-
`
`0-9
`
`0
`
`O-L
`
`o
`
`2
`
`I.
`
`5
`
`8
`
`1o
`
`12
`
`o
`
`2
`
`I.
`
`5
`
`8
`
`Cetomacrogol concentration (% w/w)
`
`2-0
`
`1-5
`
`1-0
`
`0-5
`
`0
`
`1o
`
`Figure 7.6 Effect of oetomacrogol concentration on k 1 (O) and k2 ( x )at a stirring rate of
`(a) 200 rev min“ (b) 60 rev min " ‘. Left hand ordinates 107k,. Right hand ordinates
`103 k1. The solute griseofulvin, is in powdered form. From Elworthy and Lipscomb [29]
`with permission.
`,
`
`Page 11
`
`Page 11
`
`

`
`Biological implications of surfactant presence
`
`-
`
`397
`
`carefully the rate of stirring in attempts to obtain in vitro——in vivo correlations. It
`has been found [30] that in vitro rates of methyl prednisolone, for example,
`correlated with in viva absorption rates only when the rate of stirring employed in
`the dissolution test was low.
`
`It seems likely [28] that the presence of surfactants facilitates the transfer of
`drug molecules from the crystal surface into solution as the activation energy for
`this process was found to be lower in surfactant than in water. In the case of k2,
`the activation energy increases in the surfactant solution which probably reflects
`the viscosity increase and also the possibility that a layer of adsorbed surfactant
`molecules interferes with the redeposition process.
`Chan et al. [31] have presented a theory of solubilization kinetics and its
`relation to the flow of dissolution medium, based on an analysis of five steps
`depicted in Fig. 7.7. Surfactant molecules diffuse to the surface as micellar species
`(step 1). These molecules are adsorbed on the surface of the solid (step 2) and on
`the surface the surfactant and solubilizate form a mixed micelle (step 3). In step 4
`the mixed micelle is dissolved and it diffuses away into the bulk solution in the last
`step (step 5). The solubilization rate is assumed to be controlled by steps 4 and 5 in
`Fig. 7.7. If these steps are rate controlling
`
`dl:A4]
`dt
`
`= k,-A[Mi]
`
`(7.11)
`
`where [M] is the concentration of mixed micelles in the bulk solution, and [Mg]
`is the concentration of micelles at the interface. A is the surface area per volume, k,-
`is the forward reaction rate constant for step i.
`
`<1[A4i]
`dz
`
`=k4[Ms]‘k—4[Mi][S]‘k5A[Mi] =0
`
`(7-12)
`
`
`
`Figure 7.7 Schematic mechanism for initial solubilization. Mixed micelle desorption and
`diffusion (steps 4 to 5) are assumed to control stearic acid solubilization. From Chan et al.
`[31].
`
`Page 12
`
`Page 12
`
`

`
`398
`
`-
`
`Surfactant systems
`
`where [Ms] is the concentration of mixed micelle on the surface and [S] is the
`number of free sites for micelle adsorption
`
`[Ms] = Ks[Bs]
`
`[Bs] = K2[B] [S]
`
`[S0] = [S]+[Bs]+[Ms]-
`
`(7-13)
`
`(7.14)
`
`(7-15)
`
`[B] is the concentration of surfactant micelles in bulk, [BS] at the surface and
`[Bi] in the interface. [So] are the total number of sites in the surfaces. K ,~ is the
`equilibrium rate constant for step i.
`Combining Equations 7.11 to 7.15 we obtain
`
`{k4K3[So:i/(1+K3)}[B]
`d[M] _
`dr
`" {k~4[So]+ ksA/k5AI<2(1+ 1(3)} + [B] ‘
`
`(7.16)
`
`d [M] /dt is difficult to measure. It is assumed that the solubilizate concentration
`[F] is proportional to [M] and that d[F]/dt oc d[M]/dt.
`Obtaining [Fsat] and [B] by experiment, Equation 7.16 can be rewritten in the
`form,
`
`("SW = l—~m.ki2.’E§.]l
`+{("K[21;<s:t[]B])|:k4[lSo] + ks /iK4:|} [Fiat].
`
`(717)
`
`This equation predicts that, providing steps 4 and 5 are rate controlling, a plot of
`(d[F]/dt)" 1 versus [Fm] " will be linear;
`the intercept of the plot
`is
`independent of k5 and hence independent of flow; the slope of the plot is flow
`dependent, being dependent on ks.
`In experimental studies of fatty acid
`dissolution into NaLS solutions the validity of the first two predictions was
`established (see Fig. 7.8).
`The model on which the above derivations are based is by no means
`unequivocal. There is no proof that micelles diffuse to the surface and adsorb, or,
`indeed, that hemi-micelles as depicted in Fig. 7.7 form, although Somasundaran
`er al.
`[32] have previously postulated their existence. The transfer of
`solute molecules to the micelle at
`the surface probably involves complex
`interactions between surfactant, fatty acid and water perhaps with liquid crystal
`formation as an intermediate stage following penetration of surfactant molecules.
`As the earlier steps in the process are not rate limiting their formulation is perhaps
`less important. Diffusion of the solubilizate-laden micelle is a process which must
`occur.
`
`Higuchi’s analysis [24] predicts that substantial effects on dissolution rate will
`only be evident when the drug concentration in solution approaches or exceeds
`saturation solubility. The dissolution model used by Higuchi assumes that an
`equilibrium exists between the solid and the solution at the interface and that the
`rate is controlled by the diffusion of free and solubilized solute across the
`diffusion layer which has a thickness 5.
`
`Page 13
`
`Page 13
`
`

`
`Biological implications of surfactant presence
`
`-
`
`399
`
`
`
`1/flux(1O3cm2h")
`
`Increasing
`
`O
`
`2
`
`4
`
`1/solubility (1o3cm3g“)
`
`Figure 7.8 Solubilization kinetics of stearic acid. These data support the hypothesis that
`mixed micelle desorption and diffusion are rate controlling. From Chan et al. [31] with
`permission. Re is the Reynolds number.
`
`Provided sink conditions obtain (i.e. c < 0.1 cs);
`
`dc/dt = A[(DcS/5) + (Dmcm/6)]
`
`(7.18)
`
`where cm is the increase in solubility due to the surfactant and Dm is the diffusion
`coefficient of the drug in the micelle, it being assumed that 6 is the same for both.
`
`7.2.2 Dissolution from drug—surfactant mixtures
`
`The work on dissolution rate, rather than solubility, tends to be of rather
`academic interest as a drug is rarely to be found dissolving into concentrated
`surfactant solutions. It is of more practical interest to consider dissolution from
`intimate mixtures of drugs and surfactants into water [34]. Application of the
`technique of formation of solid dispersions by fusing poorly soluble‘ drugs with
`water-soluble carrier has been shown to increase the solution rate of drugs;
`carriers used include polyoxyethylene glycol, polyvinylpyrrolidone [34] but also
`surfactants [33, 35] in their solid or waxy state. The enhanced rate of dissolution
`of testosterone [35] from Myrj 51 (but also from polyoxyethylene glycol 1000
`and PVP 11500 dispersions) was attributed to the small particle size of the drug in
`the solidified melt and to a lesser degree to the increased solubility in the carrier
`solution which formed. Ford and Rubinstein [33] made a more detailed study of
`a glutethimide—non-ionic surfactant system using Renex 650, a nonylphenyl-
`
`Page 14
`
`Page 14
`
`

`
`400
`
`'
`
`Surfactant systems
`
`polyoxyethylene condensate. Phase diagrams showed the presence of a eutectic at
`21 "/0 of the drug, 79% surfactant with a eutectic temperature of 35° C. Solid
`solutions of the drug in the surfactant and of Renex in the drug also existed. When
`placed in water, drug and carrier do not dissolve at rates directly proportional to
`their concentration in the dispersion and the dissolution rate of the drug is
`maximal when the drug concentration reaches about 25 ‘X, in the disc (Fig. 7.9).
`Dissolution of digitoxin from co-precipitates of the drug with poloxamer 188 or
`deoxycholic acid has been shown to be enhanced over dissolution from physical
`mixtures and administration of the co-precipitates to mice significantly increased
`the oral toxicity [15] (Table 7.2).
`Other techniques involving attempts to utilize the properties of surfactants
`have included crystallization of poorly soluble drugs such as sulphathiazole,
`prednisone and chloramphenicol in the presence of small amounts of surfactants
`[36]. Increases in the rate of solution were observed in each case when
`polysorbate was used as a 2.5 ‘X, solution as the crystallization medium. While the
`result might be partly ascribed to adsorption of surfactant molecules on to the
`hydrophobic crystal surface, differential thermal analysis also suggests that some
`surfactant is incorporated into the crystal structure. Interference of a surfactant in
`the crystallization process could lead to defect formation. Model studies with
`
`25
`
`N C
`
`.. U‘
`
`_. O
`
`U‘
`
`
`
`Intrinsicdissolutionrate(mgmin'1cm'2)
`
`
`
`
`
`0 ¢
`0
`
`,
`.
`.
`60
`40
`'
`20
`‘I. Gluterhimide in disc
`
`80
`
`100
`
`Figure 7.9 Dissolution rate—composition profile. Effect of glutethimide-Renex com-
`position on the intrinsic dissolution rates of 1 h old resolidified melts into distilled water at
`30°C. I Renex 650. O Glutethimide. From Ford and Rubinstein [33].
`
`Page 15
`
`Page 15
`
`

`
`Biological implications of surfactant presence
`
`-
`
`401
`
`Table 7.2 Oral toxicity of various digitoxin preparations in mice*
`
`Test system
`
`Number of
`animals deadl
`
`Mortality
`( 2;)
`
`20
`97
`
`100
`
`37
`
`30
`
`0
`0
`
`6
`29
`
`30
`
`ll
`
`9
`
`0 O
`
`Digitoxin
`Digitoxin—poloxamer 188*
`co-precipitate
`Digitoxin—deoxycholic acid?‘
`co-precipitate
`Digitoxin—poloxamer 188*
`physical mixture
`Digitoxin—deoxycholic acid*
`physical mixture
`Poloxamer 1885
`
`Deoxycholic acidll
`
`* A dose of 70 mg of digitoxin/kg was administered as a suspension in 0.5 "/5
`methylcellulose. Thirty animals were used for each test system.
`* Animals
`were observed for 7 days post-administration.
`* A 700mg/kg dose was
`. administered containing 10% (w/w) digitoxin.
`‘5 A 2.7 g/kg dose was
`used.
`‘H A 630 mg/kg dose was used.
`From [15].
`
`adipic acid have shown that surfactant adsorption on to growing crystal faces can
`change crystal habit [37, 38] (see Chapter 9).
`
`7.3 Effect of surfactants on membrane permeability
`
`Before we discuss some of the work which has been carried out on surfactant
`
`effects on drug absorption in whole animals, we review in this section some of the
`work which has been done using model systems. Foremost amongst these has
`been the goldfish Carassius auratus. In choosing this system Levy et al. [39]
`explain: ‘Most of the studies of surfactant effects on drug absorption have been
`carried out on microbial systems. The results thus obtained may have limited
`applicability to multicellular organisms, since the latter are able to maintain
`homeostasis much more effectively. Moreover, the presence of enzymes and other
`vital cell constituents in the cell membrane makes unicellular organisms
`particularly sensitive to direct effects of surfactants.’
`Use of small animals or humans presents great difficulties, not the least being
`the difliculty of maintaining a constant, known concentration of surface—active
`agent and drug. The major advantage of the fish system is that large quantities of
`test solution can be used, permitting the maintenance of constant concentration
`gradients across the membranes, which behave, as far as passive diffusion
`characteristics are concerned, in a similar way to human membranes. Fig. 7.10
`shows the effect of polysorbate 80 on the time of death of goldfish immersed in
`sodium secobarbitone solution. The results show an enhancement of activity of
`the barbiturate at low concentrations and a decrease at higher concentrations, in
`common with other studies using alternative systems.
`The end point in the experiment is the turnover time or death time of the fish.
`
`Page 16
`
`Page 16
`
`

`
`402
`
`-
`
`Surfactant systems
`
`60
`
`50
`
`40
`
`30
`
`
`
`Timeofdeath(min) 20
`
`10
`
`2-0
`10
`0-02
`0-01
`0
`Polysorbate 80 concentration (°/o w/v)
`
`Figure 7.10 The effect of polysorbate 80(1) on the time of death of goldfish immersed in
`0.02% sodium secobarbitone solution at pH 5.9 and 20° C. Mean values of 10 fish are
`shown. Vertical bars indicate i 1 standard deviation. Arrows connect values which differ
`significantly (p < 0.05) from one another. From Levy et al. [39].
`
`The reciprocal death time (T‘ ‘) is proportional to the rate of absorption of the
`drug,k1
`
`%= k1cB/cF—k2/2,
`
`(7.19)
`
`where c3 and cF are the concentrations in the bathing solution and the threshold
`concentration in the fish, respectively, and k 2 is the rate of elimination of the drug.
`A range of non-ionic surfactants has been studied for their effect on absorption
`of drugs in goldfish. Not all surfactants do increase absorption [40—42] some
`exhibiting only an inhibiting effect as seen in Fig. 7.11. Three main types of
`activity have been noted [43] when surfactant concentration is increased
`(Fig. 7.12), namely (a) the increase and decrease depicted in Fig. 7.12 when a drug
`is solubilized in the surfactant micelles (e.g. thioridazine—Renex 650 mixtures); (b)
`an overall decrease in activity when solubilization occurs, the surfactant having
`no influence on membrane permeability (e.g. thioridazine—Cremophor EL 120
`(Fig. 7.12b), and (c) (Fig. 7.12c) an overall increase in activity when the surfactant
`increases the flux through the membrane and the drug is not associated with the
`micelles (e.g. paraquat—non-ionic surfactant systems) [44].
`In some systems where the drug concerned interacts to a small degree with a
`surfactant which has a significant effect on permeability, only the increase in
`absorption is detectable. This is the case with thiopentone and a series of non-
`
`Page 17
`
`Page 17
`
`

`
`Biological implications of surfactant presence
`
`-
`
`403
`
`(
`
`Cl )
`
`( b)
`
`(c)
`
`0
`001
`01
`0-2
`0-4
`
`0
`0-001
`om
`o-025
`0-05
`o-:
`0-5
`
`-1
`
`‘I
`
`(f)
`
`I
`
`I
`
`I
`
`I
`
`I
`
`O
`0-01
`0'02!)
`005
`-1
`.5
`
`00
`
`|
`
`I
`
`I
`
`l
`
`I
`
`O 1
`
`0 - 01
`0
`25
`A
`
`0
`0
`
`0
`0-01
`0025
`0'05
`0-09
`
`0
`0-01
`0-1
`0'3
`04.
`
`(d)
`
`(e )
`
`Figure 7.11 Absorption of thioridazine in goldfish in the presence of increasing
`concentrations of various non-ionic detergents, the rate of absorption being proportional
`to the reciprocal of the death time of the fish, reciprocal death time is plotted on the
`ordinate concentrations of surfactants ("/0 w/v) are marked. Lack of enhancement of
`absorption by some surfactants is probably due to poor ability to penetrate lipid
`membranes because of shape factors. Decrease in absorption is due to non-ionic micelle
`formation. From Florence and Gillan [41] with permission. The surfactants are all Atlas
`products (Honeywill-Atlas, UK).
`(a) Atlas G2162 (II); (b) Renex 650 (III); (c) Atlas G1790; (d) G1295 (IV); (e) G1300 (IV); (f)
`Cremophor EL.
`
`Polysorbate so
`
`A
`
`62:62
`
`Renex 650
`
`re‘--fi
`H('ZO (CH2CH2O).H
`H(OCH2CH2), ocu
`”$
`H(‘3O(CH2CH2O),H
`H2CO(CH2CH2Ol.,OCR
`
`o
`'
`
`R
`
`[CH3
`RCOOCH2—CHO(CH2CH‘2O)25H
`
`O(CH2CH2O)3oH
`
`x+y+z+w=2O
`( I )
`
`(II)
`
`(111)
`
`(‘A295 ond (31300
`
`(EH2 O(CH2CH2O)~R
`(‘tel O(CH2 CH2Ol,R
`CH2 O(CH2CH:_O):R
`(IV)
`
`for G1295 x+y+ 2 =50 I50
`
`for Gl3OO x+ y+z =ca 200
`
`ionic surfactants studied in goldfish [45] using mean reciprocal overturn time as
`an index of the rate of absorption. Some results are shown in Table 7.3.
`There are several competing mechanisms for surfactant-induced effects when
`solid oral dosage forms are administered. When solutions are administered,
`
`Page 18
`
`Page 18
`
`

`
`404
`
`-
`
`Surfactant systems
`
`(0)
`
`(b)
`
`CMC
`
`(c)
`
`I
`
`Absorption
`
`1
`
`1
`
`1
`
`Surtacrcmr concn. ji-
`
`Figure 7.12 Representation of three forms of absorption—surfactant concentration
`profile (see text for discussion).
`
`Table 7.3 Reciprocal turnover times (min“‘) (is.D.) of thiopentone in
`presence of surfactants
`
`Surfactant
`
`HLB
`
`0.0005 "/3
`
`0.1 ‘X,
`
`None
`POE (4) lauryl ether
`POE (10) lauryl ether
`POE (23) lauryl ether
`POE (2) stearyl ether
`POE (10) stearyl ether
`POE (20) stearyl ether
`POE (2) oleyl ether
`POE (10) oleyl ether
`POE (20) oleyl ether
`
`——
`9.7
`12.0
`16.9
`4.9
`12.4
`15.3
`4.9
`12.4
`15.3
`
`0.11 i 0.02
`0.54