Bài báo đánh giá độ ổn định trong nhũ tương

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/239554720 Physical stability assessment of emulsions and related disperse systems: A critical review Article  in  Journal of the Society of Cosmetic Chemists · January 1982 CITATIONS READS 16 1,489 author: George D Zografi University of Wisconsin–Madison 213 PUBLICATIONS   13,142 CITATIONS    SEE PROFILE Some of the authors of this publication are also working on these related projects: Stabilization of Biotherapeutics in the Solid State View project All content following this page was uploaded by George D Zografi on 19 July 2016 The user has requested enhancement of the downloaded file j Soc Cosmet Chem.,33, 345-358(November1982) Physical stabilityassessment of emulsions andrelateddisperse systems: a criticalreview GEORGE ZOGRAFI, Professor, School ofPharmacy,University of lVisconsin-Madison,Madison, IV153706 Received April 21, 1982.Presented at theAnnualScientific Meeting of theSCC,New YorkCity,December, 1981 Synopsis A criticalreviewof emulsionstabilityassesssment is presented Consideration is givento the needfor understanding whatbasicmechanisms areresponsible for the desiredstabilityof anemulsion Emphasis is placedon the role playedby flocculation in providingthe mechanical structurerequiredfor optimal stability.It is suggested that stabilityassessment is best carriedout only on the final product,as undisturbed as possible.It is furthersuggested that rheologicalmeasurement, particularlylow shear viscoelasticity, providesthe bestapproachbecause of its sensitivity to the flocculated structureof the emulsionandanysubtlechanges in structure whichmayoccur.The useof accelerated conditions, suchas elevated temperatures, isshownto bemisleading if thebasicmechanisms responsible for instability arenot understood or if suchmechanisms changeunderconditions accelerated relativeto thoseexpected during the shelf-lifeperiod INTRODUCTION Becauseof their thermodynamicinstability,emulsionshave alwaysrepresenteda significantchallengeto the formulator and manufacturerof cosmeticproducts Throughyearsof experience andwiththe availability of manyemulsifying agentsand othertypesof excipients, the formulator generally hashadlittledifficultyin comingup with a formulationwhichprovidesthephysicalandchemicalcharacteristics requiredof any cosmeticproduct.However,becauseof the often complexnatureof these cosmeticformulations and the relativelyuncertainconditionsto whichtheymay be exposedduringmanufacture and storage,the assessment of any changeswhichmay take placeoveran extendedperiodof storageand useis not asreadilycarriedout, particularly the predictionof long-termstability.In thispresentation I wouldlike to criticallyexaminethe processes whichare responsible for the physicalinstabilityof emulsionsand relateddispersesystemsand the variousoptionsopento us for both short-and long-termstabilityassessment In particular,I would like to addressthe followingquestions.Do our testsmeasurewhat we want to measureor shouldbe measuring? How predictiveof futureeventsarethe testswe use?Is it possibleto be 345 346 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS quantitatively predictive aboutemulsion instability? Beforewe lookat theproblems of instability, however, let usfirstdevelopa betterpictureof whatwewantin thewayof a stable emulsion PHYSICAL CHEMICAL STATE OF AN EMULSION Theclassical pictureof anemulsion, whetherit isoil-in-water or water-in-oil, isthatof a liquid systemwith discreetdropletsof one phasedispersed throughoutanother, showinglittleor no tendency to cream/sediment or coalesce Usingclassical colloidal scienceprincipleswe know that the stateof free energyfor two phasesdispersed togetherleadseventually to coalescence of the dropletsanda corresponding reduction in the interfacial areaof contactbetweenthephases Suchcoalescence arises through Distance I Figure Typicalpotentialenergydiagramfor emulsion drops:A) PrimaryMinimum;B) Repulsive Barrier;C) SecondaryMinimum STABILITY ASSESSMENT OF EMULSIONS 347 the attractiveforcesrepresented by the primaryminimumof the typicalpotential energydiagramin Figure1.It, therefore,isnecessary to providea barrierto coalescence at the dropletsurfaceusingappropriate surfactants, polymers, or otheragentswhich accumulate at the interface.Our attention,therefore,is stronglydirectedtowardthe development of repulsive barriers, electrostatic and/or steric,as seenin Figure1, to keepthedropletsasfarapartaspossible In realitywe knowthatthispictureof discreet dropletsseparated fromoneanotherishighlyoversimplified Rather,we recognize that at just about any level of phasevolume ratio discreetdropletsare in a state of flocculation with an energyof interactionassociated with the secondary minimumin Figure It is this state of flocculationwhich is responsible for the rheological properties of the emulsion, givingit a higherapparentviscosity thanwouldoccurwith completelyseparated particles.It is this flocculation,and the level of rheological structureproduced,which also gives emulsionstheir desirablecharacteristics as cosmeticproductsand whichassists in resistinginstability Of considerable importance in determining theextentto whichsuchstructuring occurs withoutcoalescence arephasevolumeratio;thenatureof theinterfacialbarrier;andthe physicalstateof the systemmakingup the externalphase.The greaterthe coverage of internalphasedroplets,the betterthe chancethat highlyfloccultedemulsionswill not coalesce onstanding Importantconsiderations herearesurface charges (zetapotential); polymermolecularweight;andlevelof hydration.Alsoof considerable importance is thesurface rheologyof theinterfacial barrieranditsabilityto respond to stresses at the interfacewhichbringaboutcoalescence The stateof the bulk phaseexternalto the dropletsis also of considerable importancein stabilizingan emulsion.Higher bulk viscosityoccursbecauseof excesspolymerand surfactantdissolvedin the external phase.In the caseof concentrated surfactant andpolymersolutionsthe possibility of stabilizationbecauseof liquid-crystalline phasesmust also be considered (1) Such complexhighlyorderedphasesproducea highbulk viscosity,aswell asinfluencingthe state of flocculation To this we must also add situations where solid materials are presentto improvethe overallrheologicalcharacterof the system.Whetherthese represent highlyhydratednetworksof hydrophilicsolids,suchas clays,or crystalline waxesand otherfatty substances, the resultis an enhancedlevel of particle-particle interactionand an overallincreasein apparentviscosityand stability The picturepresented for a typicalliquidor semi-solid cosmeticemulsion, thus,is one in whichthe systemisin a highstateof flocculation Thisflocculationliterally"freezes" the emulsionparticlesinto a network which resistscreaming/sedimentation or coalescence, and producessignificantmechanical structureleadingto high apparent viscosities in the absenceof significantexternalstresses Basedupon this picture, therefore, I wouldarguethatanyapproach to theassessment of physical stabilitymust dealwiththe total emulsionasit setsup afterpreparation, andmusttakeinto account all ingredients, their physicalstateand their location,i.e., in the bulk or interfacial phases.Studyingisolatedpartsof the systemor dilutingthe emulsionto look at individualdropsis of limitedsignificance in stabilityassessment Subjecting emulsions to anysetof conditions duringevaluation whichaltersthephysicalor chemicalstateof ingredients,or their location,likewisewill lead to assessments which may not reflect the situationwhichthe evaluationis supposed to indicate.This ideawill be developed morefully laterin the discussion 348 WHAT JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS CAUSES PHYSICAL INSTABILITY OF EMULSIONS? Physicalinstabilityof disperse systems in general,andthat of emulsions in particular, is causedby a physicalphaseseparationof some type which leadsto changesin appearance; consistency; redispersability; and performance Of most criticalsignificancein cosmeticemulsionsystemsare the changesin appearance and consistency, which affectproductacceptance and performance Phaseseparationof emulsions occursbasicallyby oneof two mechanisms, coalescence and creaming/sedimentation It is importantto stressthe differencebetweenthese phenomena,becausethe methodologywe useto detectinstabilityoften doesnot differentiatebetweenthem and misinterpretation often follows.Creamingand sedimentation in liquid products dependprimarilyon such factorsas bulk phasedensitydifferences; bulk phase viscosity; particlesize;the stateof particleaggregation; and the phasevolumeratio Coalescence, on the otherhand,in additionto dependingon particlesize,viscosity and phasevolumeratio,uniquelydependson the natureof the emulsifierbarrierandthe physicalchemicalnatureof the externalphasesituatedbetweenthe dropletswhichare mostoften in a flocculatedstate.Consequently, in evaluatingcreamingor sedimentation we shouldbelookingat factorswhichrelatemostto particlemovementandto the floc structure.For coalescence theseare also important,but we must alsoconcern ourselves with thosefactorswhichinfluencethe emulsifierfilm andits abilityto resist the mechanicalstresseswhich lead to exposureof bare surfaceand subsequent coalescence (2) APPROACHES TO PHYSICAL STABILITY ASSESSMENT Beforedealingmorespecifically with emulsionstabilityassessment, it is importantto ask one more general question.What are we trying to accomplishwith such assessment? I seethree differentgoalswhich requiresomewhatdifferentstrategies First,we can usephysicalstabilityassessment to diagnoseproblemswhichmay occur duringpreliminaryformulationevaluation.In suchcasesthe changeswe are looking for haveto be verysignificantovera relativelyshorttime periodand/orthe techniques used must be extremelysensitiveto small changes.Second,we seek to monitor potentiallong-termproblemswhichmightariseduringthe desiredshelf-life,underthe environmental conditionsnormallyexpectedto be encountered Thus, for example, onemaysubjectthe productto a rangeof temperatures to whichthe productmightbe expectedto be exposedand test the productperiodicallyover a period of weeks, months,andevenyears.Thisis obviously the safestapproach towardestablishing long term stability,but it is also the costliestand most time-consuming approach.In the third strategywe can usestabilityassessment underacceleratedconditionsto shorten the time overwhichtestsare conducted.We would hope that the stresses appliedto accelerateinstabilitywould allow usto then extrapolateour resultsto normalstorage and useconditionsfor the purposeof predictinglong-termstability.In all of these casesthe importanceof knowingwhat stabilitymechanisms we aretestingis critical, and in the predictivetest we have the additionalconcernof being surethat the acceleratedconditionshave not introducednew and unanticipatedmechanisms of instability,bearinglittle relationship to the long-termsituation Much hasbeenwrittenon thevariousapproaches whichcanbe usedto assess physical instabilityof emulsions (3,4,5),soI will limit my comments to a morecriticalevaluation STABILITY ASSESSMENT OF EMULSIONS 349 of a fewspecificissues of significant importance Thesearetheproblems of accelerated emulsionstabilitytesting,and the rheologicalevaluationof emulsions, an approach whichI feelis the bestwayto dealwith emulsionstabilitytesting ACCELERATED TESTING To predictthelong-termstabilityof anemulsionsystemfromstudiesconducted overa relativelyshorttime periodrequiresthe introductionof a stresswhichwill accelerate instability, and methodology whichwill measure the processes leadingto instability Essentially, therearetwo majortypesof stresses whichhavebeenusedto accelerate instability, centrifugation andtemperature Centrifugation is obviously onlyapplicable to fairlyfluid emulsions whichcanbe forcedto separate underthe rangeof forces producedin commercially available centrifuges If the problemis clearlyoneof phase separation dueto creamingor sedimentation, it shouldbe possibleto run samples at variousratesof centrifugation, determinerateconstants for the process, andextrapolatetheseto forcesdueto gravity.An exampleof suchdata,takenfromthe work of Garrett(6), is shownin Figure2 It is important,however,to be surethat the phase 2o o} 18 ,,, 16 '"' 14 • I0 " • 4 I0 12 14 I6 FLOTATION RATES(SE• '1)vs.(R.P.M.) Figure RateConstantrs R.P.M for centrifugalevaluationof emulsioncreaming(6) separation occurringreflectsonlycreamingor sedimentation and not coalescence, as well.It is alsoimportantthat suchstudiesbe carriedout onlyon the finishedproduct Sincecoalescence canoccurundersufficientcentrifugalforce,suchtechniques might be usefulin accelerating the coalescence process, assuggested in a few studies(6,7).If usedfor this purpose,however,it is importantthat the centrifugation lead only to coalescence or that creamingandsedimentation be separated out asfactors.Also,it is 350 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS importantto be surethat the mechanisms involvedin causingcoalescence under ultracentrifugal forcesarethe sameasthoseoperatingundernormalconditions In my opinionit is reallydoubtfulthat thisis everthe casesinceultracentrifugal forcescan distortdropletsand place stresses on the emulsionbarrierin a way not normally encountered.It may be arguedthat survivalunder suchhigh stressassuressurvival underotherlessstrenuous conditions, andthusassures anacceptable shelf-life.Thereis no doubtthat the barrierat the dropletinterfacemustbe verygood if it prevents coalescence underultracentrifugal forces.However,it tellsusnothingaboutthe time periodoverwhichwe canbe assured of stabilityandit actuallymayeliminatenormally acceptable emulsions Thisis an exampleof a situationwhichexistsin anyaccelerated test,i.e.,a tendencyto "overkill"the emulsions because the testusedintroduces a new mechanism of instabilityor causesan unreasonably highstress By far the mostwidelyusedstressin emulsiontestingistemperature The rationalefor increasing temperatures to predictinstability comesfromthe well knownrelationship betweenthe rateconstant, k, for a chemicalreactionandtemperature, T, expressed as the Arrhenius equation Simplyplottingk vs 1IT allowsoneto determine k at any temperature, as longas themechanism of thereaction hasnotchanged Hereinliesthe criticalquestionconcerning the useof hightemperatures to accelerate instabilityin emulsions Is it properto assume thathightemperatures don'tchangesomeof thebasic mechanisms involvedin the instabilityprocess? Considerfirstsomeof the waysin whichtemperature may affectemulsions These includechanges in the viscosity of liquidphases; solubility;partitioningof molecules betweenbothphases; themeltingandfreezingof variousmaterials, particularly waxes; and the hydrationof polymersand colloidalsolids.Giventhat one or moreof these processes istheprimaryfactorin stabilizing theemulsion, it wouldnot besurprising if a highertemperature abruptlyeliminated thisasthe importantstabilizing factornormally observedat lowertemperatures An exampleof this is seenin Figure3, wherean 40,000 tO øG •'0,000 o I0,000 - tu 6,000 • 25 øG o• 4.,0oo • '"' 2,000 1,000 oC - 60O 5oC 4OO 2OO I00 • 0.04 0.1 0.4 I ELAPSED TIME I0 40 I00 (DAYS) Figure Logarithmof ApparentViscosityvs Logarithmof Time for an o/w lotion at various temperatures (22) STABILITY ASSESSMENT OF EMULSIONS 351 emulsionsubjectedto highertemperatures undergoes a dramaticdecrease in apparent viscosity, presumably whenthe meltingpointof someof the waxespresentis reached The partitioningof surfactants betweenphases, likewise,can changewith changing temperature in sucha wayasto createa situationunrelatedto that at roomtemperature Thisisnot to saythat stabilitytestingshouldnot be carriedout at highertemperatures Indeed,if thesetemperatures areor couldbe actuallyencountered suchtestsmustbe carriedout.What I am saying,however,is that predictionof long-termstabilityfrom suchhightemperature studiesis highlyunlikely,particularlyfor emulsions containing waxes,polymers, andcolloidalsolids.Again,asin the caseof centrifugation, survivors maybe takento be stablesystems, but we still mayrejectgood systems in the process andnot reallyhavea goodhandleon actualexpectedshelf-life Freezing of emulsions isalsoa widelyusedformof stress, particularly in the formof a freeze-thaw cyclingtest.In additionto theeffectsdueto temperature changediscussed above,in thistestwe alsohavethe freezingof waterto form ice crystals In general thereare3 factorswhichareimportantto recognize in interpreting themeaningof such tests(8,9).Theseareconcentration andsolidification of freeliquidwater;concentration andprecipitation of dissolved substances; andthinninganddisruption of theemulsifier film by icecrystals The criticalpartof the testcomesasonethawsthe samplesince thawingpermitsthe release of wateranditsrapidmovementthroughouttheemulsion If theemulsifier filmcan"heal"itselfuponrelease of thestresses inducedbyicecrystals beforecoalescence occurs,the systemwill survivethe test However,if the rate of redissolution of ingredients istoo slow,for example,instability mayoccurin a manner not relatedto normal usageat highertemperatures, so use of this techniquefor predictivepurposes stillsuffersfrom the uncertainties alreadydiscussed In summary, withregardto predictive testingof emulsions, I wouldconcludethatthere reallyisno clearcutbasisfor expecting thataccelerated studies allowextrapolation to normalstorageconditionsandquantitative expirationdating.Carefulanalysis of how increasing or decreasing temperatures mightmechanistically effectvarious processes of importance in stabilizing emulsions couldhelponeto chooseprocedures whichbest reflectanticipated effects.If, for example,we knowwaxesarepresent, andthat their degreeof crystallinity is important, we shouldbuildthatintothetypeof stress testwe develop.This obviouslycallsfor muchmore researchon the thermalpropertiesof emulsions andtheirisolatedparts RHEOLOGICAL EVALUATION OF EMULSIONS Sincethe propertiesof cosmeticemulsions,be they pourableliquidsor semi-solid creams, dependprimarily ontheextentof structure developed throughparticle-particle interactions, anychangein thisstructurecanbe considered a form of instabilityor an indication of sometypeof instability Changes in thisstructure maybebroughtabout, for example, by suchprocesses ascreaming; sedimentation, coalescence; fiocculation; crystallization or meltingof waxes;hydradonchangesin colloidalsolidslike clays;or changesin the solutionstate (aggregation) and concentrationof surfactantsand polymersin the two bulk phases In viewoftheimportance of structure inemulsion system, it followsthatanytechnique usedto monitorinstabilityshouldbecarriedout on thefinishedproductratherthanon isolated partsof thesystem or afterthesystem hasbeendilutedsoasto changethereal 352 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS structure.This is not to say that studyingisolatedpartsof the system,e.g., zeta potentials or surfacefilm rheology,wouldnot be helpfulin understanding how the systemis actuallyfunctioningand hencehelpfulin morerationallyformulatingthe product.However,in evaluatingphysicalstabilityof an emulsionproductnothing shortof workingwith the finalproductwill be veryuseful If structureis indeedas importantas suggested here,then the bestapproachto evaluatingemulsionstabilityand the underlyingcausesof instabilityis the use of rheological measurement Rheological measurement basicallyinvolvesthe imposition of sometypeof mechanical stress onthesystem andtheobservation of theresponse to this stressin the form of a dimensionalchange or deformation.What makes rheologicalmeasurement so attractivefor our purposesis that instrumentation is availablewhich allowsthis stressto be variedcontinuouslyover a verywide range Dependingon theextentof structure in oursystemwe maywantverysubtlestresses or extremelyhigh stresses This abilityto alter stresses and measureresponses is also importantbecause emulsions normallyencounter differentlevelsof mechanical stress duringmanufacture, storage,and use,and amongotherfactors,one wouldwant to knowin advancehowtheemulsionwill respondundersuchconditions The science of rheologyand its applications to dispersed systems is verywell documented in the literature at boththetheoretical andexperimental levels(10-13).Consequently, in this concluding sectionI wouldsimplyliketo reviewsomecurrentthinkingaboutthebasis and applicationof rheologicalmeasurement to emulsionassessment and to suggest carefulreviewof approaches reported in theliterature for differenttypesof situations It is safeto saythatmost,if not all, emulsionsystems we dealwith appearto exhibit non-Newtonian behaviorwhenapparentviscosities are measured at variouslevelsof stress andrateof shear.This behaviorarisesprimarilybecause the flocculated gel-like structure exhibitedbymostcosmetic emulsions breaksdownunderincreasing levelsof shear.Thustheverycommonpracticeof measuring "one-point"apparent viscosities, at best,canonlyprovidesomeindication of structure changefor the conditions usedin the measurement.If this is understood,however,and if the objectivesof such measurements aredefinedcarefully,apparent viscosity measurement maybeveryuseful To ãOO, OOẵ oã 60,0OO]oã 40ã000 r 20,OOO r - '-' ,ooo 1_ ø2 6ool- '-* t t • ,,oool_ 6oop o.m 0.04 , o• 0.4 • ELAPSED TIME •o 40 •oo 400 (DAYS) Figure Logarithmof ApparentViscosityvs.Logarithmof Time for o/w lotionsexhibitinghardening (22) STABILITY ASSESSMENT OF EMULSIONS 353 in rapidlyspottingproblems.For example,asshownin Figure4, an apparentviscosity changeof somesignificance occursastheproductcoolssometimeafteremulsification of a wax-containing lotion,apparently dueto a changein therateof waxrecrystallization.We sawearlierin Figure3 thatanabruptdropin apparent viscosity occurredwhen the lotion wassubjectedto a temperature at or abovethe meltingpoint of some ingredientor ingredients The Brookfieldviscometerwith a helipathattachmenthas beenusedto detectthe tendencyfor creamingor sedimentation (14).By measuring apparentviscosity at variousdepthswithinthe disperse system,usingverylow spindle speeds,it waspossibleto detectdifferences in concentration of dispersed particles arisingfrom particlemovementupwardor downwardat earlystagesbeforethiswas apparentby visualmeans Becauseemulsionsexhibit non-Newtonianrheology,continuousshearrheometry, which allowsone to relateshearstressto rate of shearwith a known and continually variablerateof shear,is widelyusedin emulsionevaluation (15).Emulsions ranging frompourableliquidsto semi-solids will typicallyexhibitthecurvesshownin Figure5 SHEAR STRESS Figure Rate of Shearvs.ShearStressfor semi-solido/w emulsions varyingin emulsifierconcentration Hysteresisbetweenvaluesobtained upon increasingand decreasingsheararises becauseof structurebreakdownand a time dependencyfor restructuring If this processis reversible and isothermal we speakof thixotropicbehavior.More often,as seenin Figure6, however,when we deal with emulsionsthe processinvolvesan 354 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS t4 t3 t2 SHEAR STRESS Figure Rateof Shearvs.ShearStress for a semi-solid emulsion sheared afterdifferenttimeperiodsof rest: t• < t < t irreversible breakdowndue to majorstructuralchanges In suchcases,afterthe first rheologicalrun we are no longer really dealingwith the original emulsion,and subsequent testingmay be meaningless Despitethislimitation,however,continuous shearrheometry doesallowsomeinterpretation, particularly if theemulsion exhibits a well-defined yieldvalueor "spur"in the rheogram, asshownin Figures5 and6 The yieldvalueor spurclearlyreflectthebasicgel-likestructure of anemulsion Therefore, anychangein itsvalueor appearance mustbea reflection of a changein structure and instability Unfortunately, manyinvestigators concentrate theireffortson studying the regionsof highshearanddo not seethe subtlechanges whichoftenshowup at the lowershearvalues.I wouldargue,therefore,thatwe wouldprobablylearnmuchmore about emulsioninstability,and even have a chanceat predictingfailure,if we concentrate on rheologicalmeasurement at as low a shearingstressas possible Hiestand (14),forexample, hassuggested a relatively simpletestof a dispersion's ability to withstand"curdling"or otherchangesin floc structurewhichmight be brought about by vibrationor other forms of gentle movementduringthe shipmentof products.The dispersion is placedin a bottlewhichrotatesslowly,e.g.,4 revolutions per hour.The slowagitationtendsto causebreakingof the flocat somepoints,but allowstimefor reestablishing flocculation at otherpoints.Thisapparently resultsin a totallydifferentflocstructure anda changein the characteristics of the dispersion The type of bond breakingand makingjust described is characteristic of systems exhibitingthe phenomenon of viscoelasticity for whichthereis a soundtheoretical STABILITY ASSESSMENT OF EMULSIONS 355 basis(11).Recentpublications (16-18)haveclearlyindicatedthe valueof evaluating emulsionsby measuringtheir viscoelastic propertiesand have providedsuggested experimental procedures whichcouldbe adaptedto stabilityassessment Despitethis information,however,the application of suchtechniques doesnot appearwidespread Consequently, I would like to concludethis discussion of stabilityassessment by reviewingwhat hasbeenandwhat couldbe done Basically, manystructured systems suchaspolymers, gels,suspensions, andemulsions under very low levelsof stresswill exhibit linear elasticdeformationbecauseof relatively strongintermolecular or interparticle bonding•In suchcases thestress applied isdirectlyproportional to thestrain(extentof deformation) andindependent of therate of strain.Aswithperfectlyelasticsolids,thislinearrelationship isdescribed byHooke's law In the caseof a true elasticbody the strainwill remainconstantunderconstant stress, anduponrelease of thestress, thesolidwill fullyrecoveritsoriginalstructure In the caseof a viscoelasticsubstance,under such conditions a slow relaxation or increase in strainoccursoveran extendedtime period.This arisesbecause the structuredbody undergoes a process of bond breakingand remakingin sucha wayasto modifythe structure.Eventually,with continuedapplicationof the stressviscousflow will occur Whenthe stressis smallenoughto causeHookeanelasticityand NewtonJan viscosity, we speakof linearviscoelasticity, for whichtheoretical concepts werewellworkedout (11,18).Uponrelease of the stress in suchsystems somerecovery of structure occurs because of theelasticity, butthisrecovery cannotbecompletebecause of therelaxation whichhastakenplace.A typicalplotof compliance (strainperunitstress) versus time, oftencalleda creepcurve,isshownin Figure7 What makesthisapproach soattractive for emulsionstabilityassessment is that we are able to obtainfundamentalmaterial constantssuchas elasticityand viscositywithout significantlystressing the system mechanically Indeed,Barry(18) hasreferredto this as an evaluationof the "ground state"of anemulsionor suspension Therefore,wearenot onlyin a positionto observe verysubtlestructuralchanges whichtakeplaceuponstorageand use,but we alsocan determinethe valuesof fundamentalconstantswhich correspondto acceptable stability.This allows us, for example,to developquality control specifications translatable from laboratoryto laboratory.Because of the highsensitivity of viscoelastic measurement, it shouldalsobe possibleto attemptto predictlong-termstability utilizing a much less extreme acceleratedcondition and, hence, minimizing the possibility of a changein the mechanism of instability Althoughtheapplication of viscoelastic measurements to anunderstanding of polymer solutionbehaviorat a fundamental levelis well documented (11),reportson the useof suchtechniquesfor dispersesystemsof interestto cosmeticand pharmaceutical scientists havebeenextremelylimited.Barry(19) hascarriedout extensivestudieson somesemi-solid o/w emulsions andhasbeenableto monitorverywellthe significant changesin viscoelasticity broughtabout by the effectof emulsifierconcentration on emulsionstructure Sherman (20)hasusedviscoelastic creepteststo studyliquidw/o emulsionsstabilizedby sorbitanmonooleate.He has correlatedthe tendencyfor coalescence duringshelflife for emulsions rangingfrom 30%to 65%waterwith their viscoelastic properties Hiestand(14),usingthe risingsphereviscometer developed by McVeanandMattocks(21),hasapplieda verysmallstress on a 5%solidsuspension stabilizedwith methylcellulose and hasmeasuredviscoelastic relaxation.He wasableto showthatthe relaxation process arosefromthe flocculated structure of the suspension and not the vehiclealone.Thus we seethe potentialutility of thesetechniques in 356 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 280 x 240 m,m (.• 160 120 80 4O I I I I I I I I0 12 14 TIME (min.) Figure Compliance vs Time for a viscoelastic emulsionsystem: A) Elasticity; B) Viscoelasticity; C) VisousFlow;D) Recovery studying disperse systems rangingfromhighto low in thedegreeof overallconsistency and structure We maynowask,whatisthe nextstep?It seems to methatthemostpressing problem facingcosmetic formulatots is theirinabilityto predictwhethera disperse systemwill or will not hold up overits expectedshelflife Observing viscoelastic properties of a given formulationunderstorageconditionsover a periodof time, and/or under minimalstress conditions, shouldbe carriedout.The leastthatwill happenisthatone will havefundamental theological parameters whichdescribe a stableemulsion system Evenbetteris the possibility that this techniquewill pick up changein theological behaviorwhichcanthen be extrapolated to shelf-lifeconditions It remainsfor such studiesto be carriedout beforethe potentialvaluedescribedcan be realistically STABILITY ASSESSMENT OF EMULSIONS 357 assessed, and it is my hopethat this reviewwill stimulatefurtherinquiryinto this topic SUMMARY AND CONCLUSIONS Most references to emulsionstabilitydeal with relativelysimpleliquid emulsions whichbearlittle relationship to typicalcosmeticlotionsandsemisolids Sincemost cosmeticemulsionsare aggregatedgel structures,any test must only minimally disturb this system relative to the environmentalsituation to be encountered duringshelf-lifeanduse Most stressesused to accelerateinstability cause it by unrelated mechanisms Thermalstressis usefulonlyif relatedto the temperature rangeto be encountered Thereisthe dangerof over-killingacceptable emulsions Low sheartheologicalstudieswhichmeasureviscoelasticity, coupledwith continuous shearrheometry,are the bestmethodsfor monitoringphysicalstabilityof emulsions REFERENCES (1) S.Friberg,FoodEmulsions (MarcelDekker,Inc.,New York, 1976) (2) J A KitchenerandP R Mussellwhite, "The Theoryof Stabilityof Emulsions," Ch in Emulsion Science, ed P Sherman(AcademicPress,London, 1%8) (3) K.J Lissant, Emulsions andEmulsion Technology PartsI andII, (MarcelDekker,Inc.,1974) (4) G Zografi,"Emulsions," Ch 16in TheTheory andPractice ofIndustrialPharmacy, 1stedition,ed L Lachman,H A Lieberman, andJ L Kanig,(LeaandFebiger,Philadelphia, 1970) (5) P Sherman, "Accelerated Testingof Emulsion Stability,"Soap, Perfumery and Cosmetics, 44, 693-699 (1971) (6) E R Garrett,"Prediction and Evaluation of EmulsionStabilitywith Ultracentrifugal Stress,"J Soc Cosm.Chem.,21,393-415 (1970) (7) R C Vold andR C Grott, "Parameters of EmulsionStability,"J.Soc.Cosm Chem.,14, 233(1%3) (8) T G RochowandC W Mason,"Breaking Emulsions By Freezing," Ind Eng.Chem.,28, 12%-1300 (1936) (9) J.P Cordiez,G Grange,andB Mutaftschiev, "DropletFreezingExperiments in StearicAcid-Water Emulsions, Roleof Droplet-Medium Interface,"J Coll.Interf Sci.,85,451-441 (1982) (10) H R Kruyt, ColloidScience, Vols.I andII (ElsevierPubl.Co.,New York, 1949) (11)J D Ferry,Viscoelastic Propetites ofPolymers, 3rdedition(JohnWileyandSons,New York, 1980) (12) H Schott,"Rheology," Ch 22 in Remingtons Pharmaceutical Sciences, ed A Osol(MackPuN Co., Easton,PA, 1980),323-339 (13) P Sherman, "Rheologyof Emulsions," Ch.4 in Emulsion Science, ed P Sherman (AcademicPress, London,1968) (14) E N Hiestand, "Physical Properties of CoarseSuspensions,"J Pharm.Sci.,61,268-272(1972) (15) B W Barryand E Shotton,"Structure and Rheologyand SodiumDodecylSulfate-Cetyl AlcoholWater:Continuous ShearExperiments,"J Pharm.Pharmac.,19, 1105-1205 (1%7) (16) B W BarryandE Shotton,"Rheologyof SodiumDodecylSulfate-Cetyl Alcohol-Water: SmallStrain Experiments,"J Pharm.Pharmac., 19, 1215-1295 (1%7) (17)B W BarryandB Warburton, "SomeRheological Aspects of Cosmetics,"J Soc Cosm Chem., I9, 725-744 (1968) (18) B W Barry,"Evaluationof RheologicalStatesof Semisolids," J Soc.Costa.Chem.,22, 487-503 (1971) (19) B W Barry,"The Self BodyingActionof the Mixed Emulsifier SodiumDodecylSulfate/Cetyl Alcohol,"J.Coll.and lnterf Sci.,28, 82-91(1%8) 358 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS (20) P Sherman,"Rheological Changesin Emulsions on Aging,"J Coll.and Interf Sci.,24, 107-114 (1967) (21) D E McVeanandA.M Mattocks,"SimpleRheometer for Definitionof Viscoelastic Properties of Structured Liquids,"J Pharm.Sci.,50, 785-788(1961) (22)J H Wood andG Catacalos, "Prediction of The Rheological Agingof Cosmetic Lotions,"J.Soc Cosm.Chem.,14, 147-156(1963) View publication stats ... surfactants, polymers, or otheragentswhich accumulate at the interface.Our attention,therefore,is stronglydirectedtowardthe development of repulsive barriers, electrostatic and/or steric,as seenin... andemulsions under very low levelsof stresswill exhibit linear elasticdeformationbecauseof relatively strongintermolecular or interparticle bonding•In suchcases thestress applied isdirectlyproportional