lnternotionol Journal of Cos,netic Selence. 2004. 26. 47 59

Structure and rheology of semisolid 01w creamscontaining cetyl alcohollnon-ionic surfactant mixedemulsifier and difíerent polymers

H. M. Ribeiro J. A. Morais~ and G. M EcclestontLaboratório de Dermofarmácia e Cosmética. Faculdade dc Rarmácia da Universidade de Lisboa. Av.das Forças Armadas.

1600 Lisboa. Portugal. and fDepartmcnt of Pharmaceutical Sciences. Ilniversityof Strathclyde. Glasgow G1 1XW. UK

Received 10 September 2003. Accepted 10 September 2003

Keywords: calorimetry, creams. polymers, rheology. viscoelasticity

Synopsls

Oil-in-water (o/w) emulsions forcosmetic use. such aslotions and creams, are complex multiple-phase systems, which may contam a number of interactingsurfactants. fatty amphiphiles, polymers and otherexcipients. This study investigates the iniluence oftwo synthetic cationic polymers, Polyquaternium-7and Polyquaternium-11, and the natural anionicpolymer. gum of acacia. on the rheology and microstructure of creams prepared with a non-ionic mixedemulsifier (cetyl stearyl alcohol-12E0/cetyl alcobol)using rheology (continuous shear, and viscoelasticcreep and oscillation), microscopy and differentialscanning calorimetry (DSC). A control cream containing no polymer was also investigated. The semisolid control cream was structured by a swollenlameliar gel network phase formed from the interaction of cetyl alcohol and the POE surfactant, in cxccssofthat required to stabilize oil droplets. with continuous phase water. Endothermic transitions between25 and 100 C were identified as components of thisphase. Incorporation of cationic polymer into the formulation caused significant Ioss of structure to produce a mobile semisolid containing larger oudroplets. The microscopical and thermal data impliedthat the cationic polymer caused the swollen lamellar gel network phase to transform into non-swollencrystals of cetyl alcohol. In contrast, incorporation

Correspondence: Helena Margarida Ribeiro. Faculdade deFarmácia da Universidade de Lisboa. Av.das Forças Armadas.1649-003 Lisboa, Portugal. E-mali: helena.ribeiro ff.ul.pt

ofgum of acacia produced a thicker creatn than thecontrol, with smallcrdroplet sizes and little evidenceof the gel network. Microscopical and thermal dataimplied that although there were also interactionsbetween gum of acacia and both the surfactant andthe swollen gel network phase. the semisolid properties were probably because of the ability of the gumofacacia to stabulize and thicken the emulsion in theabsence of the swollen lamellar network.

Résumé

Les émulsions huile dans leau à usage cosmétique.telles que les lotions et les crêmes, sont des systàmesde multiples phases complexes. qui peuvent contenirplusieurs surfactants. amphiphiles gras, polymàreset d’autres excipients. Cette étude examine 1’ influencede deux polyméres cationiques synthétiques. Polyquaternium-7 et Polyquaternium-11. et d’un polymère anionique naturel, la gomme d’acacia, sur Iarhéologie et la microstructure de crèmes préparéesavec un émulsifiant non ionique (l’alcool de alcobol12E0/cetyl de stearyl de cetyl) utilisant la rhéologie(cisaillement continu. techniques de fluage et desoscillations), la microscopie et la calorimétrie (DSC).Une créme de contrôle sans aucun polymère a aussiété examinée. La crème de contróle semisolide a étéstructurée par une phase de réseau de gel lamellairegonflé formée deI’ interaction entre I’alcool cetyliqueet le surfactant POR, excédent qu’a exigé pour stabiliser les gouttelettes des émulsions, avec de l’eau dephase continue. Des transitions endothermiquesentre 25 et 100 C ont été identifiées comme les

© 2004 Blackwell Publishing Ltd

Structure and rhcology oí semisolid o/w creams H. ‘VI. Ribeiro cL ai.

composants de cette phase. Cincorporation des polymàres cationiques dans la formulation a causé Iaperte signilicative de structure pour produire dessemisolides contenant de plus grandes gouttelettesd’ huile. En raison des données microscopiques etthermales, les polyrnêres cationiques ont eu poureffet que Ia phase de réseau de gel lamellaire gonflé aété transformée en cristaux nongonflés d’alcool ceLylique. Au contraire. 1’ incorporation d’acacia a produitune crème plus épaisse que le contrâle, avec destalhes de gouttelette plus petites et peu d’apparencedu réseau de gel. Des données microscopiques et thermales ont suggéré que, malgré le FaiL qu’ ii yavait aussides interactions entre l’acacia et le surfactant eL laphase de réseau de gel gonflé. les propriétés semisolides ont été probablement dues à la capacité de lacada pour stabiliser et épaissir lémulsion en l’absencedu réseau lamellairegonhlé.

Introduction

Cosmetic emulsions such as lotions and creams arerarely simple two-phase oih-and-water systems, andtheir study and development is one of the most difficuit and complex subjects in dermapharmacy andpharmaceutics. Such preparations often contam several interacting excipients and may be composed ol’additional phases to ou and water. In aqueous sysLems containing surfactant íatLy alcohol combinaLions, the addiLional phases generaily form when Lheemuisilier, in excess of that required Lo form a mono-molecular flIm at the oil droplet interface, interactswith continuous phase water Lo form a gel networkofvastly swolien bilayer structures. This swollen network controls the rheoiogical properLies of the cream[1 4]. In emulsions containing non-ionic surfactantand fatty aicohol. Lhe sweihing oí Lhe bilayers isbecause of Lhe hydration of the polyoxyeLhylenechains of the interpositioned surfactants that areorientated and extended into the interiamellar waterlayers [5]. tt has been established that the swelhingbehaviour and. hence. Lhe rheological properties ofsuch creams are proportional Lo Lhe surfactani polyoxyethylene chain length [6].

Water-solubie poiymers are used in dermatologicalpreparations as emuision stabilizers. thickeningagents and. sometimes. as Lhe active [7]. When suchwater-soluble polymers and suríactants are formulated Logether, they may interacL either Lo enhanceconsistency and/or sensory properLies. or provideunwanted elTects. However. it is difficult Lo predictwheLher such interactions might occur in a speciflc

formulation. Although Lhe association of polymersand surfactants in dilute aqueous solution has beenextensivelystudied over Lhe last 30 years. and severalreviews have been published [8.9]. most studies haveconcentrated on interactions between polyelectrolytes and ionic rather than non-ionic surfactants.Weak interactions between polyelectroiytes andnon-ionic surfactants have been reported only wherehydrogen bonding or hydrophobic interactionsbetween polymer and surfactant occur [10 12]. SuchinteracLions, although weak, sometimes cause dramatic changes in rheology [13]. The present workreports ao examination of the effect of cationic polymers, Polyquaternium-7 (Merquat 550’) and Polyguaternium-il (Gafquat 755”), and the naturalanionic polymer, gum of acacia, on the sirucLureand rheology (conLinuous shear and viscoelasLiccreep and oscillation) of a semisoiid emulsion formulation containing non-ionic surfactant ceLylstearyl alcohol 12E0 combined wiLh ceLyl alcohol.These caLionic polymers are often incorporatedas Lhe actives in shampoo and hair-conditioningformuiations. Gum of acacia is, in fact, a very peculiar poiymer with particular surface properLies,and is of a complex structure. In addition. it is notonly a polysaccharide, buL comprises of a proLeicstructure.

Materiais and ,nethods

Materiais

Deionized double distilled water was used throughout. Cetyl alcohol, propylenegiycol and gum ofacaciawere used according Lo Lhe F.RV (V PortuguesePharmacopea). Cetyl stearyl alcohol 12 moi EO(Eumuigin Bi” — non-ionic emulsilier) and thedecyl oieate (CeLiol V”) were obtained from Henkei,Dusseldorf, Germany. Diazolidinyi urea (as one preservative: Germali ti”) and Poliquaternium-li. aquarternized poly vinyl pyrrolidone dimethylaminoethyl meLhacrylate (Polyquaternium-11: Gafquat755 “)were obtained from ISP. Barcelona. Spain. Poliquaternium-Z a copoiymer of acrylamide and N,N

dimethyh-N-2-propenoi-1 ammonium chloride (Polyquaternium-7: Merquat 550”) was obtained fromCalgon Corporation. USA. The preservative solution.methyldibromogluLaronitrile and phenoxyeLhanol(Euxyi K 400”), was obtained from Schulke andMayr, Austria and Glycerine 97% was obtained fromMerck, Germany. AlI the excipients were used asreceived.

48 © 2004 Inter,,ational journal of Cosmeiir Sc’ience. 26. 47—59

Structure and rheoiogy of semisolid 01w creams 1-1. M. Ribeíio ei ai.

Methods

E,iiuLçíon processingPolymer geis (Poiyquaternium-Z Polyquaternium-1land gum of acacia) were prepared initiaily in waterin order Lo form geis with the same dry weight 10%(w/v). This was Lo ensure a previous hydration of Lhepoiymcr.

The poiymeric emulsions were prepared according Lo Lhe following formula w/w: surfactant (cetylstearyl alcohol 12 moi DO) 3%, (cetyi aicohol) 15%,polymer gei 10%, oiiy ingredient (decyl oieate) 6%,humectant (propyieneglycol) 3%. preservatives (diazolidinyl urea) 0.5% and (methyldibromogiutaroniLriIe and phenoxyethanol) 0.2% and deionized waterup to 100%. A control emuision. wherc poiymer geiwas replaced bydeionized water. was aiso prepared.Toprepare the emulsions, Lhe oiiy and aqueous phaseswere heated separateiy Lo 75 °C, Lhen the oiiy phasewas added Lo the water phase and Lhe system wasmixed(130 r.p.m)withconstantagitationuntil3ü C.

Microscopy

Each cream was examined using a Poiyvar microscope (RheicharL-Jung. Austria) in bright fieid andbetween crossed poiars. Sampies were examined1 month after preparation.

Rheoiogical analysis

The rheologicai characteristics of Lhe emulsionswere examined aL high shear raLes using continuousshear techniques and in Lhe viscoeiastic regionusing creep and osciliation techniques.

Continuous shear experimenisPlow curves of shear rate against shear stress wereobtained at 25 C using a Perranti—Shiriey cone andpiate viscometer in automatic mode, with a 4-cmradius cone and a maximum shear rale of 1651 sThe eIïeci of draughts and sampie evaporation wasminimized with Lhe manufacturer’s antievaporationunit (solvent trough and vapour hood), and the resultant llow curves weredispiayed on a X—Yplotter. Preiiminary experiments were performed Lo investigateLhe influence of sampie size when flhiling the viscometer on flow curves by applying a icnown weightof cream onto Lhe plate using a 1-cm-diameler syringe. Flow curves were obtained after storage ai25 C for 1 month. Each test was performed ai leasiLhree Limes using freshiy ioaded sampies.

Viscoelastfc experimeiztsThe essence of viscoelastic anaiysis isto test variousstructures in Lhe sampie in the rheoiogicai groundstate where the method of testing does not destroyLhe siructure. Non-destructive creep and osciliiatoryexperiments were performed with a controiled stressCarri-Med CSL2 100 Rheometer (TA lnsLruments.[1K) using conc and piatc gcomctry (truncatcd coneangie 1 and radius 6 cm). Ali sampies were Lested1 month after preparation. and care was Laken Lo justliii Lhe cone and plate gap with Lhe correct amount ofcream. Tests were performed at ieast in duplicateusing fresh sampies for each repeal. Data were analysed using Lhe manufacturer’s software.

CreepCreep describes the siow deformation oí a materialwhen a stress is imposed on Lhe sampie at zero timeand then maintained constant. The time-dependentstrain or compliance response is plotted as Lhe creepcurve. The theory of viscoeiasticity can be appiied Loa creep Lest, provided thaL the strain response is proportionai to the applied stress, i.e. the BoiLzmannsuperposition principie is salislied [14, 15]. Preiiminary Lests for hnearity were performed aL short Limesby increasing the appiied stress and measuring Lhecompiiance afLer 5 mi In subsequent creep experiments, appiied siresses in the linear viscoeiasticregion of 0.5 (Lhinner creams) and 2 Pa (thickestcreams) were used. These stresses were near Lheminimum iimits of the insLrument under Lhese conditions.

AfLer Lhe viscometer was ioaded. each sampie wasieft for 5 mm for stresses induced by ioading Lo reiax.and the creep sLress was apphed for a minimum of30 mm. FuH 30-min recovery curves were aiso obtained and analysed to conlirm that each creep curvehad bcen ohLained in lhe linear viscoeiastic region.

Dynasnh (sinusoidal) testiiig

Osciiiialory Lesting provides informaLion at shortcreep times as a periodic experiment ai w rad s ‘isquahtativeiy equivaleni Lo a LransienL experiment aLLime t 11w s. it compiemenis the creep Lesi in thatii can anaiyse eiements with reiativeiy fasL responsetimes. The sampie is exposed Lo a forced osciliation(frequency). and the Lransmitted sLress is measured.Varying the frequency of lhe appiied waveform aiiow amplitudes will cause the diftereni eiements lorespond when their characteristic Limes maLch therale of change of stress associaLed with a particularfrequency.

© 2004 lnternotio,rnl jour;ial of Cosmetic Sciem’e, 26. 47 59 49

Structure and rheology oí semisolid o/w creams H. M. Ribeiro CL ai.

lnitially, Lhe linear viscoeiastic region was determmcd in manual mode by increasing the torque atconstani frequency (1 Hz) and measuring theresponse. In subsequent experiments, a stress of2 Pa was used for the frequency sweep test(5 x io—~ to 35 Hz) as iL is in the linear viscoeiasticregion. Data were analysed to give the storage modulus. C. and the dynamic viscosity. q. The storage modulus is related Lo the energy stored during a cycle, i.e.elastic energy. whiist the dynamic viscosity is relatedto the dissipation of energyduring a cycle, i.e. the viscous energy.

Difl’erentiai scanning calorimetry (DSC)

Differentiai scanning calorimetry experiments wereperformed between ambient temperature and100 C on Lhe emulsions and raw materiais using aMettier DSC 30 system (Mettler. Switzeriand). Thesample and reference (air) were placed in hermetically seaied pans. A scan speed of 10 mia and 510 mg sample size gave the best compromise betweenresoiution, temperature accuracy and attenuation.

Results and discussion

Appearance of emulsions

The control and Lhe cream containing gum ofacaciawere white stiff semisolids. In contrasi, the creamscontaining the cationic polyrners Poiyquaternium-7and Polyquaternium-11 were white. giossy andmobile emulsions. Polyquaternium-7 cream appeared to be Lhe thinnest. lt was slightiy elastic.although still easiiy pourable.

Rheology

Conlinuous shear

Continuous sheardata show thatcorrect fluiingofthecone and plate gap is critical to obtain reproducibleflow curves. Figurei shows representative plotsobtained by 50% over- and under-flhling of Lhe coneand plate gap. The under-flhled curves obtained alilay considerably to Lhe Ieft of Lhe correctly flhled flowcurve, and a different ílow profile was shown. In contrast. over-fluling Lhe gap shifts the curve only slightlyto the right and its shape is identicai. Visual observation indicated that as soon as the cone began lo move,lhe excess cream was detached from the moving portion. Reproducibiiity was shown Lo be exceiient(within ±2%) when Lhe cone plate gap was correctly

Figurei Ferranti—Shírley cone and pinte viscometer.EiTecLs oíchanging Lhe quantity of cream in the rheometergap: (1) under-iiilling: (2) correct lillling: (3) over-fiuiing.

fihled. and ali Lhe piots are the mean of aL ieast tworuns.

Continuous shearexperiments measure the abiiityofeach system to resist structurai breakdown duringthe standardized shearing procedure. The llowcurves obtained for ali the samples (except gum ofacacia) were in the form of anticlockwise hysteresisloops, indicating shear thinning. Representativeplots are shown in Fig. 2 with apparent viscosityvalues calcuiated at Lhe apex of the ioop (1651 s 1),

which is also shown in the figure. Apparent viscosityvalues provide a comparison of the resistance tostructural breakdown between Lhe creams and theloop areas compare Lhe amount of structure thatfractures in the standardized cycie. The controlcream showed additional compiex flow with a spuron the up-curve. Such complex flow curves are typicai forcreams prepared with surfactant fattyalcoholmixed emuisifiers, where the spur. aithough, in part.instrumental, is a function of signiflcant eiasticity[16, 17]. The flow curves cieariy demonstrate Lhatincorporation of Lhe cationic poiymers ai least parLialiy destroys Lhe structural eiements of Lhe cream.The mobile poiymeric creams showed little hysteresisand had iower apparent viscosity values than Lhecontrol cream. with the Polyquaternium-7 creamshowing the iowest apparent viscosity and the narrowest hysteresis ioop.

Ia contrast, Lhe llow curve for the gum of acaciacream was complex and different from Lhose of theother creams. The up and down curves crossed

2

3

oLOCO

‘oa,co

‘oa,-cri)

oO Shear Stress (Pa) 3.3

50 © 2004 lnter,zotional journal of Cosmetie Scieuee, 26. 47 59

Structure and rheology of semisolid 01w creams II. M. Ribeiro et ai.

Figure 2 Ferranti—Shirley cone and piate viscometer. Representative flow curves oí Lhe four creams. The values inparenthesis represcnt apparent viscosities caiculated at loopapex. 1650 s ~: (1) Polyquaternium-7 cream (0.05 mPa s):(2) Poiyquaternium-11 cream (008 mPa s): (3) Controicream(1.59 mPa s):~4)gumofacaciacream(2.28 mPa s).

over, implying that Lhe shearing cycle itself mightcause structural build-up (rather than destructionas with the other creams). and the high apparentviscosity value indicated that this cream was thickerand more resistant to structure breakdown than thecontrol.

Creep

Ali Lhe creams were viscoelastic when they weretested in creep mode. and exhibited Lhe same generalshape. Figure 3 shows typical reproducibility dataand Fig. 4 shows Lhe typical creep curves for Lhe fourcreams. Analysis of Lhe data (Table 1) indicated thateach cream was a viscoeiastic liquid that could berepresented in modei form by a Maxwell unit in series

Figure 3 Creepcurvesofdifferentmeasurementsofgumofacacia cream to iliustrate reproducibility.

times(s) 4000

Figure 4 Creams creep curves: (1) l’olyquaternium-llcream: (2) control cream: (3) Polyquaternium-7 crearn: (4)gum of acacia cream.

with a number of Voigt units [1]. The rheologicalequation of state for this model is:

I(t)=Jo+ZIj(1 e (1)

whereJis Lhe ratío of shear strain Lo shear stress, J(t) isLhe total creep compliance aL Limei, lo is the residualshearcompliance. J~ is Lhe compliance of Lhe ithVoigt

Table 1 Analysísofcreepcurves

Voigt unit 1 Voigt unit 2

Creams 110(Pas) Ja(Nm2) ‘ii (Pas) J1 (Nm 2) ~. (s) q2(Pas) J2(Nm 2) ‘2(S) .4(Nm2)

Standard 2.83 x io~ 3.00x lo” 3.30 x io~ 7.04 x 10 227 120 x 10 1.27 x 10 12,8 1.38 x l0_2Gum ol acacia 1.29 x io~ 4.12 x io-~ 1.81 x io~ 1.40 x 10 244 7.70 x lo3 1.97 x 10 14.9 1.60 x lo_2Polyquaternium-7 2.83 x i0~ 3.20 x io-~ 4.89 x io~ 2.00 x 10 6 120 — — — 6.36 x 10_2Polyquaternium-I1 4.42 x lo3 2.60 x io-~ 1.86 x io~ 5.00 x lO 6 96 — — — 4.07 x l0_l

Creep time: 30 mm.

1651

o)0

Oa,CO

o1.6 2.5

Shear Stress (Pa)

times (s) 4000

o‘o0?o

EO)oa

oEoo.

oo

//EZ

© 2004 Internalional journal of Cosmebe Science. 26. 47—59 51

Structure and rheology of semisolid 01w creams H. M. Ribeiro CL ai.

unit. Lis Lhe retardation time of the ith Voigt unit and~ is the residual shear viscosity.

The initial elastic component of the creep curve isassociated with Lhe residual (uncoupied) Hookeanspring. which represents bonds being stretched elastically, and simulates the elasticity of the gel nelworkstructure in Lhe emulsion. Tabie 1 shows that theinstantaneous compIiancesJ~, (which are the reciprocal of Lhe modulii of eiasticity) of the Polyquaternium-il and Polyquaternium-7 creams are muchlarger than either gum of acacia or control creams,implying that Polyquaternium-11 and Polyquaternium-7 creams possessed a more delicaLe networkthan Lhe others. The curved portion of Lhe creepcurve is the region where Lhe flow is viscoelastic,and Lhe model representation consisLed of Voigt uniLsconnected in series. They represent that part of Lhestructure in Lhe gei network in which secondarybonds were breaking and reforming during the tesL.In Lheory, Lhe number of Voigt units may be inlinite,but in the present work, one (for Poiyquaternium-7and Polyquaternium-11 creams) and two (for the conLrol and gum of acacia creams) VoigL units wererequired to reproduce Lhe creep curves. The retardaLion time t represents Lhe time during which bondsbreak and reform, and as ali honds did not do this atthe same raLe. the weaker bonds break aL smailervalues of t Lhan Lhe sLronger ones, a specLrum ofretardation time exists.

The viscoelastic data imply that the gel networkstructure of Lhe conLrol cream is essenLialiydestroyed when caLionic polymer is incorporatedinto Lhe cream. whiist Lhe addition of gum ofacaciadoes not prevent structure formaLion. The retardation Limes for PolyquaLernium-1I and Polyquaternium-7 creams were much lower than Lhe control,whilst those of Lhe gum of acacia cream were similarLo the control. In addition. the viscosiLies and compliances of the Polyquaternium-11 and FoiyquaLernium-7 creams were much lower and higher,respectively, than those of Lhe conLrol. whereas thecreep data for Lhe gum of acacia cream were similarto Lhe control. This implies thaL the different Lypes ofbonds were breaking and reforming in emuisionsincorporating caLionic poiymer and suggest that Lhegel network structure was essentially destroyed bycationic polymer.

When Lhe stress had been applied for suflicienttime toensure thaL ali theVoigt units were essenLiallyfully extended. further extension was in the natureof a viscous fiow. This strain was represented by Lheresidual dashpoL. In this region. Lhe creams were

llowing as dispersions and the viscosities were high.Again, although the residual Newtonian viscositiesofcontrol and gum of acacia were similar, Lhe valuesfor Polyquaternium-11 and Poiyquaternium-7 weremuch iower, because Lhe weak network exisLingin Lhese creams caused less interference with Lheprocess of llow [18J.

Dynanzic (sinusoidal) LestiugThe resulLs shown in Fig. 5 show Lhe variation ofstorage modulus. G’, and Lhe dynamic viscosity 1’ withfrequency for Lhe creams. For each cream.increases and 1’ decreases wiLh increase in frequency. This behaviour is typical of a viscoelasticliquid and can be described by a mechanical modelmade up from a combination of springs (eiastic elements) and dashpots (viscous elements). At high freguency. Lhe springs can elongate and contract underimposed shear, but Lhe dashpots have very liLLie Limein which to move. Thus, Lhe system behaves essentially as an elastic solid SO that values of O’ are highand of1’ are Iow. AL low frequency. Lhe springs alsocan extend, but in this case. the dashpoLs have ampleLime Lo move and their extension greatiy exceedsLhaL for the springs. Thus, values for ‘q’ are high aLiow frequency.

~ntrol~acacia.—*—poly7 —..—poiy II

0.00i 0.01 0.1 i lO 00(b) Prequency (Hz)

Figure 5 Osciliatory data: Piots of(a) storage modulus and(b) dynamic viscosity wíth frequency.

10000

1000

100

lO

otoEULI

o

0,001 0.01 0.1 i lO 00(a) Frequency (Hz)

10000

1000

100

lo

LIo

oo

oELIo

~0

controlacaciapoiy 7

— poiy II

52 © 2004 Internatiomil Journal ai’ Cosmelie Science. 26, 47—59

Structure and rheology of semisolid o/w creams H. M. Ribeiro eL ai.

The osciliation data show Lhe sarne general trendsas Lhe conttnuous shear and creep data in Lhat aL eachfrequency. Lhe oscillatory pararneters for the cationicpolyrner creams were much Iower Lhan those ofthe control. indicating less structure, whereas Lhe

— e

• e.

60.

parameters for the gum of acacia cream were similarto the control. The data also irnply that the Polyquaternium-li cream is more mobile than Lhe Polyquaternium-7 emulsion and LhaL the gum of acaciacream is slightly Lhinner than the control. In creep.

80.

factanL (cetyl stearyl alcohol 12 moi EO): (b) cetyl alcohol.

© 2004 Inlernutionul journul of Cos,netic Scie,we. 26. 47—59

Structure and rheology of semisolid 01w creams H. M. Ribeiro ei ai,

the opposite effects were found. These discrepanciesare minor and are possibly a result of both viscoelastic tests being dose to their limits of linearity at stresses dose to the minimum available under theseconditions.

galquat cream10.050 mg Rate: 10.0 °Cfmin

49 mJ4.9 J/g

43.3°C-1.4 mW

E

(b) 30.

standard cream5.714 mg

a Onset 34.7 °C~ Siope —0.78 mW/K

40.

Rato: 10.0 °Cfmin

Onset 48.7°CSiope —0.71 mW/K

Onset 54.4 °CSiope —1.25 mW)K

Figure 7 Thermograms of the íourcreams: (a) Poiyquaternium-11 cream: (b) control cream: (e) Polyquaternium-7 cream: (d)gum of acacia cream.

DSC

Ijifíerential scanning caiorimetry thermogramsbetween ambieni and 100 C for raw materiais(Fig. 6) and the creams (Fig. 7) were obtained. The

Ao

integrationDeita H

Peak

Ed

Onset 37.4 °CSiope —0.27 mW/K

integiationDeita H 68 mJ

6.8 J/gPeak 56.5°C

-3.1 mW

ønset 53.6 °CSiope —0.91 mW/K

(a) 40. 50. 60. 70.,..I,,,,i,,,,

80. 90.

36 ni/J6.2 J/9

Peak 36.7 °C-1.3mW

IntegrationDeita H 108 ni’J

18.9 J/gPeak 55.5 °C

-3.2mW

50. 60. 70. 80. °C

© 2004 lnternatíonal jow’nai of Cosmetie Scíence. 26, 47—59

Structure and rheoiogy oí semisolid 01w creams H. M. Ribeiro eL ai.

merquat cream7.273 mg Rate: 10.0 °Cfmin

~ Oriset 37.2 ~~ Siope —0.16 rnW/K

Onset 54.3 °CSiope —0.92 mW/K

acacia cream7.325 mg

Á

e

ELCi

Rate: 10.0 °C/min

polymeric excipients (gum oí acacia. Polyquaternium-7 and Polyquaternium-11 polymers) did notshow endotherms in this range. With cetyi alcobol.there was a broad endotherm peaking at 51 C witha broad shoulder from 40 to 50 C. The broadshoulder probably represented the meiting of 3 and

y crystais and the endotherm peaking at 51 thefusion of the a crystais L3. 19. 20]. The meltingof the suríactant crystals (cetyi stearyl alcohol12 moi E0) occurred at iower temperatures witha broad double endotherm peaking aL 34 and40 C.

IntegrationDeita H

Peak

30 m)J4.2 J/g

43.3 °c—0.9 mW

IntegrationDeita H 58 r&J

7,9 J/gPeak 56.8

—2.2mW

90. °c40. 50. 60. 70. 80.(o)

‘i(ci) 40.

Figure 7 continued

Onsel 55.2 °cSiope —4.63 rnW?K

integrationDeita H 276 rnJJ

37.7 J/gPeak 57.0 °c

-8.4mW

50. 60. 70. 80. 90. °c

© 2004 Inlernntionol journol of Cosmetfc Scienre. 26. 47—39

Structure and rheology of semisolíd o/~ creams H. M. Ribeíro et ai.

Such broad endotherins are often diíficult to interpret. but when considered in conjunction with therheoiogical data, a possibie explanation emerges[3, 21]. The control cream shows two endotherms.The broad endotherm at higher temperature is characteristic of creams containing swoiien lameilarphase. with the Iow temperature shouider representing crystaiiine phasc and the high temperature

(b)

‘.3.

endotherm because of swollen cz-crystalline gelphase transitions [19]. The Iow temperatureendotherm peaking at 36 C is consistent with thatofíree surfactant.

The creams with cationic polymers (Polyquaternium-7 and Polyquaterníum-11) show the surfactantendotherm peaking (±37 C). suggesting that Lhecationic polymers do not interact with the surfactant

4

~I2

Figure 8 Photomicrographs of creams: (a) control cream: (b) gumofacaciacream: (c~ I’olyquaterniuin-Ii cream: (d) Polyquaternium-7 creani.

© 2004 l,iternaLional Jounuil of Cosmetic Scieuce. 26. 47—59

Structurc and rheology of sernisolid o/w creams H. M, Ribeiro et ai.

alone. However. the broad shoulder on the majorendotherm of cetyl alcohol (approximately 55 C) tsnow almost gone (small onset at 43.3 C).VisuaI andrheological data indicate Iack oí structure in thesecreams.This suggests that thesc polymers mightprevent the swelling of the a-crystals of cetyl alcohol,and hence the formation of swollen lameliar gelphase.

(d)

In contrast, DSC data imply that there is interaction between the gum of acacia and non-ionic surfactant for the endotherm represenhing the surfactanthas disappeared in the gum of acacia cream. As withthe cationics. the broad shoulder on the majorendotherm ofcetyl alcohol has also disappeared. Thissuggests that the lameliar structure in the gum ofacacia cream has algo been destroyed. 50 that the

Figure 8 continued

© 2004 Internnlional journal of Cosmetie Scie,we, 26. 47—59

Structure and rheology of semisolid 01w creams H. M. Ribeiro ei ai.

structurai properlies in this cream may be simply aresultofgeiation of the gum [22].

Microscopy

Representativephotomicrographs of lhe fourcreamsare shown lo Fig. 8. lo ordtnary Iight, ali creams cootained distorted oU droplets.The gum of acacia creamcontained smallerdroplets than the control, whereasdroplet sizes were largest in the creams containingcationic poiymer. Crystalline structures were alsomore prominent in these creams. Multiple w o w droplets appeared lo be present in the Polyquaternium11 cream.

Conclusions

The iníluence of three diiTerent polymeric geis. PoIyquaternium-Z Polyquaternium-11 and gumofacaciaon the microstructure of creams prepared with lhenon-ionic surfactant cetyi stearyl alcohol-12 moi £0in combination with cetyi aicohol has been investigated using rheological (continuous shear. and viscoelastic oscilialion and creep Lechniques) andmicroscopicai and thermai DSC techniques. The datawere compared with that of a semisolid controlcream containing no polymer.

The data indicate that the cationic poiymers interact with otherexcipients in lhe cream to cause significant ioss of structure with an increase in dropietsizes. lo continuous shear. apparent viscosities forthe calionic creams were signiflcanlly Iower thanthose of lhe control. and in creep and oscilialion. theiower elastic and viscous moduli confirmed thatlhese creams were less structured than lhe control.DSC data implied that the cationic polymers did notassociate with the non-ionic surfactant alone, butpossibiy reduced lhe swelling of lhe iamellar crystalune gei phase. thereby increasing the conlinuousphase-free water content.

ln contrasl. the gum of acacia cream remainedsemisoiid. contained smaller droplels and appearedto be somewhat thicker than the controi. In continuous shear. the gum of acacia cream was Lhe mostresistant Lo structure breakdown, giving a very complex clockwise llow curve, although viscoelastic datadid not differ markediy from lhe control. DSC spectra.however. showed that gum of acacia also interactedwith excipienls in lhe crearn. As with lhe calionicpolymer creams, the gum of acacia reduced lhe sweluing of lhe lameilar gel phase. However, unlike lhecationic polymer creams, lhere was aiso an inlerac

tion between gum of acacia and the non-ionic surfactant, as lhe surfactant endotherm had disappearedaltogether. The semisolid nature of lhe gum ofacaciacream was possibiy simply because of ils ability toboth slabiiize and thícken emulsions in Lhe absenceof Lhe Iameuiarnetwork.

Acknowledgements

The authors wish to lhank Mr Massoud Bakkhashaefor his Lechnical assistance.

References

1. Barry. B.W. and Eccicslon, G.M. Influencc of gel networksin controliing consistency ofo/w emulsions stabilised bymixed emulsiliers. 1~ Texture Slud. 4.53—81(1973).

2. Eccieston, CM. The microstructure of semísolidcreams. Pharm. Internar. 7.63—70(1986).

3. Eccieston. CM. MullipIe-phase oil-in-water emuisions.

J. Soe. Cosmet. Cheia. 41.1—22(1990).4. Eccieston. CM. Functions of mixed emulsifiers and

emulsifying waxes in dermalologicai lolions andcreams. Cofloids and Surfaces: Physicochemieal andEngineering Aspecis 123—124, 169—182 (1997).

5. Vringer.T.. JoosLen, J.G.H. andJunginger. H.E. Characterisation oí Lhe gel structure in a non-ionic ointmenLby smail angle x-ray diffracLion. 1. Colloid liii. Sei. 262,55—60(1984).

6. EcciesLon, CM. and Beatlie. L. Microstructural changesduring Lhe sorage of systems conlaining cetostearylaicohoi poiyoxyethylene aikyl elher surfactants. DrugDcv. Ind. Phar,n. 14.15—17(1988).

Z Reibeira H.M.. Rodrigues, L and Momi~ J.A. Haircare Pbiymers Appiied to o/w Creams Stabilily StudieaProceedings of lhe VIII Coti. Nae. Derinofarinacia, Caceres(1996).

8. Kwak. J.C.T. Polynier—Surfactant Systems. Marcel Dekker. NewYork (1998).

9. Goddard, E.D. and Ananthapadmanabhan. K. InleracLions o.)’ Surfactants with Polyrners and l’roteins. CRCPress Boca Rabo, London (1993).

10. Sarrazin-CarLalas. A.. Iliopoulos. 1.. Audebert. R. andOisson. U. Association and thermai gelation in mixluresoí hydrophobicaily modifled polyeleclrolyLes and nonionicsuríactanls. Lang,nnirlO, 1421—1426(1994).

11. Kamenka, N., Kapiun. A.,Talmon.Y. and Zana. R. Interactions belween polysoaps and surfacLants in aqueoussoluLions. Langnis,ir 10,2960—2964 (1994).

12. Ihopouios. 1. and Olsson. U. Polyeieclrolyte associationLo miceiles and bilayers. 1, Phys. Chein, 98. 1500 1505(1994).

iS. Jocic. Ii. Julia. M.R. and Erra. RThe time dependence ofchitosanfnonionic surfactant solution viscosily. 1, Ain.OiI Cheia. Soe. 274.375—383 (1996).

58 © 2004 International journal of Cosinetie Selence. 26. 47 59

Structure and rheology of semisolíd ojw creams H. M. Ribeiro ct ai.

14. Barry. B.W. Rheology of pharmaceutical semisoiids? In:Advances ia Pharmaceutical Sciences (Bean. H.S.. Beckett. A.W and Carless. J,E.. eds). Academic Press. I~ondon(1974).

15. Van ~A~zer. j.R.. Lyons. J.W.. Kim. K.Y. and CoiweII, R.E.Viscosity and Flo~~’ Measurnnent. A Lul~ori1ory HandbookofRheology mx 362. Interscience. London (1963).

1(x Barry. B.W. and Saunders. CM. The self-bodying actionof Lhe míxed emulsiliercetrimide ceLostearyl alcohol. J.Colloid liii. Sei. 34.300—315 (1970).

17. Davis. 8.5.. Shotton, E. and Warburton. B. Some limitations ofcontinuousshear methods for Lhesludyofpharmaceutical semi-soiids. J. Pharm. l’hannacoi. 1575—167S (1968).

18. Barry. B.W. and Saunders, CM. Rheology of systemscon(aining cetomacrogol 1000 cetostearyl alcohol.

Ii. Variation with temperature. 1~ Colloid 1111. Sei. 38.626—632 (1972).

19. Eccieston, CM. and Beattie. L.The iniluence ofemulsilier composition on Lhe microstructure oí non-ionicsemisolids. Br. Phar,naccul, Coiif 12. 771’ (1990).

20. Schambil. E. josL. E and Schwuger. M.J. Inlerfacial andcoiloidal properties of cosmetic emulsions conlainingfaLLy alcohol and fa(ty aicohol polyglycol ethers. Pro~XIV I.R&C.C. Cong. 2.583 584(1986).

21. Eccieston. C.M. Phase transiLion in ternary systems andoil-in-water emulsions containg cetrimide and fattyalcohols. mi, J. l’har,n. 27.311 323 (1985).

22. Shotton. E. and Davis, 8.8. The effect of emulsilierconcentration on the rheological properties ofacacia emulsions. 1~ Pharin. Pharmacol. 20, 780—789(1968).

© 2004 In(ernalio,snl Journal of Cosmetic Scicnee, 26, 47—59