]. Soc. Cosmet. Chem., 27, 509-531 (November 1976)

Characterizing Cosmetic Effects andSkinMorphology byScanning Electron Microscopy CHARLES A. GABBER, Ph.D. and C. T. NIGHTINGALE,

B.S.*

PresentedMay 30, 1975,SCC Seminar,St. Louis,MO Synopsis:The SCANNING ELECTRON MICROSCOPE has developed into an imr•ortant tool for characterizingthe effectsof COSMETIC PRODUCTS on HUMAN SKIN. Methods are describedfor "replicating" areas of stratum corncure as in vivo impressions,

or "negativereplicas."The "negatives" are replicatedagainto produce"positives," which are then studiedin the scanningelectronmicroscope.Through the use of selectedcontrol subjects,it has been possibleto definethe differencesbetween dry and normal skin, the former generally exhibiting larger uplifting desquamatinglayers (flakes) of stratum corneum.Experimentalfindingsnote that beneficialclinical effectsare generallyaccompaniedby a reductionin the amountand/or size of the desquam.ating material.Examples will be givenfor (1) moisturizingprodu,cts,(2) protein additives,(3) abradentcleansing products,and (4) moistnrizingsoaps.The effects,in someinstances,are so subtle that unlessthe microscopyis done as Before and After pairs on the same identical areas, the effectswill not be recognized.Once goodmicrographsare obtained,it is sometimes possibleto quantiratethe resuksthroughthe use of Quant/metImage AnalyzingCom-

puter,t a tool originallydevelopedby metallurgists, but which has great potentialfor quantifying cosmeticeffects. INTRODUCTION

The characterization anddefinition of the effectsof cosmetic products on humanhair andskinhasalwaysbeenoneof the mostimportantresponsibilitiesof the cosmetic chemist. Althoughthe actualformulation of new productsis generallybasedon fundamentalprinciplesof physicaland surface chemistry, the measurement of the efficacyof suchmaterialshasgenerally beenperformed by subjective clinicalevaluation. Onlyin recentyears,in part •StructureProbe,Inc., P.O. Box342, West Chester,PA 19380. timanco, Inc., Monsey, N.Y. 509

510

JOURNAL OF THE SOCIETY OF COSMETIC

CHEMISTS

due to the rising demandsfor the scientificdocumentationof advertising claims,hasinterestbeenfocusedon the development of reliablequantitative measurements for productefficacyon humanhair and skin. We have found electronmicroscopy,and in particular,scanningelectron microscopy(SEM) to be a powerful tool for scientificallydocumentingthe effectsof materialsapplied to human hair and skin. The transmission electron microscope(TEM) alsohas its place,but sinceSEM is basicallya surface (topographical)tool, and sincestudiesof productson hair and skin involve surfaces(protein depositson the outsideof a hair shaft;skin roughness is a surfacephenomonen),the SEM is generally the instrumentof choice. The SEM and TEM, rather than "competing"with each other, are actually complementary, andmostimportantstudiesultimatelyinvolvethe useof both. The presentwork is not the first applicationof SEM for cosmeticstudies. In fact, as early as the 1940s(1, 2) variousmethodsof replicatingthe outer cuticlestructureof hair were devisedthroughthe useof platinumand carbon techniques.The SEM, which really startedto developas a scientifictool in the late 19fi0s,was quickly utilized to examinehuman hair, and Wolfram and Lindemann(3) publisheda study of the structureand morphologyof hair. Further studieshave demonstratedyet additionalinformationon hair (4, 5) and, more recently,somevery novel studieshave been publishedon the elucidationof the actualstructureof the cuticle (fi). One of the mostin-

ventire approacheswas for the study of porositywithin humanhair (7), which describesa methodfor the polymerizationof methacrylatetoohomer in situ.within the holesor poresthat existwithin the hair structure.The hair is then degradedaway, leavingthe now "plasticpores"(or "ghosts")avail-

ablefor collection andcharacterization by SEM. A methodfor demonstrating the repair of damagedhair and a classification schemefor identifyingand quantitatingdifferenttypes of damagehas now been reportedby DiBianca (8). This latter method was used to substantiatethe claim that a hair care productcanrepairdamagedhair. Althoughthe amountthat is now knownand publishedabouthumanhair is substantial,by comparisonvery little SEM and/or TEM work has been publishedabouthuman skin.Virtually no SEM or TEM studieshave been publishedthat demonstratethe effectsor efficacyof specificproductson the stratum eorneum.This absenceof data is not for lack of interest, but because

of the enormousexperimentaldifficultiesencounteredwith skinstudies.These difficultieshave, therefore,severelylimited the applicationof SEM and/or TEM techniquesto the skinarea. For TEM studies,a biopsysamplemustbe (1) "fixed"(hardenedto keep it from degrading);(2) embedded(so it can be sectioned);(3) sectioned; (4) "stained"to bring out contrast;and (5) photographedwithin the high vacuumof the electronmicroscope. Put quite simply,the changesbrought

aboutby the fixing,embedding,and stainingare enormous as comparedto

SCANNING

ELECTRON

MICROSCOPY

511

therelativelysubtlechanges broughtaboutby theapplication of a cosmetic product. In fact,evenanoilymaterialappliedto thesurface of a biopsysample will tendto diffuseawayinto the embeddingmaterialand not remainon the sample. Hence, for all of thesereasons,SEM tends to be the tool of choicefor most skin studies.

Our firstSEM studies onhumanskindid involvebiopsysamples, bothanimal aswell ashuman.We learnedquitequicklythat the severedehydration effectsof the ultrahighvacuumwithinthe SEM left the samplesoseverely shrunken, distorted, andchanged that anyrelationship betweenrealityand whatwe wereseeingwouldhavecertainlybeenfortuitous.Evensomeof the newerandmoresophisticated methodsof samplepreparation,suchascritical pointdryingor freezedrying,werenot particularlybetter.Consequently, we haveconcentrated virtuallyall of our effortstowardprocedures employing skinreplicas, whichhavethe advantage thatwe cannotonlycharacterize by SEM the in vivo stateof an areaof stratumcorncure,but we can alsofollow the sameidenticalareain the micrographs asa functionof time,in a sense,

likea highmagnification formoftimelapsephotography. In thisway,'wecan followthe effectsof a productas a functionof time, and evaluatethe efficacyof thisproductrelativeto a control. Surely,we •verenot the firstto makereplicasof humanskin.Facq (9, 10) and Berstein(11, 12) learned back in the late 1960sthat normal silicone

resin,eitherof the RTV-11type* or eventhe impression materialsusedin dentistry, wouldmake"acceptable" (relativelyfree of artifacts)replicasof the top surfaceof the skin (and, hence,alsothe top surfaceof the stratum comeurn).Oncehavingobtainedthe "impression" or "negative"aswe prefer to call it, the negativeis itselfreplicated,yieldinga "positive" replicaof the skinsurface. Therewasalsoa problemwith artifactsin the replicas,suchas air bubbles,whichtendedto limit, in certainrespects, the utility of these original efforts.

Unfortunately, artifactswerenot the onlydrawbacks tø the methodsoriginallyused.First,wasthe criticismthat too smallan areawasbeingstudied; how wasit knownthat this smallareawasindeedrepresentative of that area of stratumcorncureas a whole?In addition,the enormousheterogeneities presentfrom point to point on the stratumcorncurewere suchthat one could almostprove or disproveanythingby a carefulselectionof the area. Our approachwas to developa procedurethat would enableus to do the following.

1. produce high resolution(virtually) artifact-freenegative replicas; 2. convert the negative replicas to high resolutionpositive replicas; *General Electric Company,Waterœord,N.Y. 12188.

512

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

3. photograph the sameidenticalareas,beforeand after treatment,as a functionof time, and by followingthe sameidenticalarea, avoid some of the problemsof hetcrogeneity from pointto point; 4. makelargemontages, ratherthanrandomlyselected picturessoit would be obviousa largeenoughareawasbeingphotographed; 5. quantifythe resultingimageseithermanuallyor throughthe useof an imageanalyzingcomputer(suchas a Quantimet); 6. correlatethe results,quantitativelyaswell as qualitatively,with known clinical

conditions.

The purposeof thispaper,therefore, is to describe someof the methods we havedeveloped forthispurpose, to presentthe clinicalbasisfor justifyingthis approach, andto showhow the resultingdatais presentlybeingusedto demonstrateefficacyand documentadvertisingclaims. EXPERI1VIENTAL

A. Preparationo• NegativeReplicas

RTV-11 sfiiconresinis acceptablefor makingthe negativereplicas,althoughwe findthatbetterresultsareobtainedwith thehighmolecular weight fractionremovedand when an absoluteminimumof catalystis usedto setup the resin.Certainsilicone-compatible surfaceactiveagents(suchas "Aerosol OT,"*) seemto impartbetter releasecharacteristics and may have their maximumbenefitwhenreplicatingparticularlyroughskin (suchaspsoriatie subiects). B. Conversiono[ Negative Replicasto PositiveReplicas

Almostanythermoplastic resincanbe usedfor thispurpose,but we prefer fine spray-dried or asopolymerized polyolefin"bead"polymer.After the positive replicais completed,it is metallizedto makethe replicasurfaceconductive to preventsamplechargingin the electronbeam. C. Preparationo[ ReplicaCross-Sections

For replica cross-section work, the positivereplicasare diamondknifetrimmed, using a duPont/SorvallMT2-B Diamond Knife Ultramicrotome.t The knife speedand anglemustbe adjustedfor minimumchatterand thinnestsections(even thoughthe sectionsare not important) as theseare the conditionsfor obtainingthe best undistortedand unsmearedcross-sections.

Unfortunately, the sorer resinsyield the bestdetail in the positives; hence, we sometimes mustdo the sectioningin a cryostageat liquid nitrogentem-

*American Cyanamid Co., Stamford,CT. rE. I. dupontde Nemoursand Co., Inc., Wilmington,DE 19898.

SCANNING

ELECTRON

MICROSCOPY

513

peraturesasa meansof minimizingsmearingandpreserving the cross-section-

al edge.We havenotbeensuccessful usingglass knivesforthispurpose. D.SEM

The workdescribed in thispaperwasdoneon a JEOL Model#JSM-U3 SEM.* We havefoundthisequipment to be particularly well-suited for this typeof workfor severalreasons: (1) it canbe modifiedwithouttoomuchdifficultyto providea virtuallydistortion-free imagein the15to 20 X range,an absolute necessity for puttingtogether distortion freemontages; (2) it hasan extremely flexiblegoniometer stage,with all of the degreesof freedomnecessaryto relocatethe sameidenticalareasof a replicawithoutexceptional difficulty;and (3) the electronic stabilityof the instrument is suchthat micrographscanbe takenwithoutdiscernable drift in brightness or contrast, an extremelyimportantfeaturexvhenmontages arebeingmade. Themostimportant aspect of ourprocedure is thateverysuccessive replica in a seriesis doneon the very sameidenticalarea,sametilt, samemagnification,in short,everyaspectof the imagesare identical.We have eliminated

any chancethat a particularfeatureof the resultsmightbe due to the way the microscopy was doneratherthan due to real changeson the skin surface.Thisalsomeans, therefore, that all microscopy, whichis donegenerally in the form of largemontages, is clearlyfrom the sameidenticalareasand this fact is so obvious,that evena nonmicroscopy-oriented observerwould realize

it.

E. QuantitativeImage Analysis

All imageanalysis wasdoneon a Quantimet720BImageAnalysingComputer equippedwith the automaticdetector,optical and epidiascope input modes,andclassifier-collector modules with directteletypeoutput. RESULTS

An exampleof what onecantypicallyobtainis shownon Fig. 1. This is an areafromthe backof thehandof a youngfemalesubjectin her mid-twenties, and the microscopy was originallydone as a montageat 100 X from the scanningelectronmicroscope(but which was reducedfor publication). The originalindividualfields that make up the montagecan still be seen. The "hill andvalley"typeof topography is "normal," particularlyfor younger subjects. The onemajorexceptionis what is sometimes calledthe "cobblestone"textureand an exampleof it is shownin Fig. 2. The term "cobblestone"

*Jeol (USA), Inc., Medford, MA.

514

JOURNALOF THE SOCIETYO1•"COSMETICCHEMISTS

Figure I. SEM montagetakenfrom positiveskin replicafrom back of hand of female subiectin mid-20s.Exampleof "good" skin. Scalebar represents•00 txm.

texturewasoriginally coinedby D. R. Highley(13)whohasfour•dthatthis textureoccursin approximately 5 percentof the population. For studieson skincareproducts, we generally try to avoidusing"cobblestone" subiects, becausethe effectsof cosmetic products aremoredifficultto discernin the final montages andthequantitative analysis is moredifficult. Onemustbe extremely carefulwhenmakingboththe negativeas well as thepositive replicas, soasnottoproduce artifacts. Figure2 wasactually made in mid-1972, andwasoneof ourearliestattempts. It alsocontains the 3 most commonly foundartifacts:(1) air bubblesin the negativereplica;(2) air bubblesin the positivereplica;and (3) a glazedareaof improperly mixed replicating material(in the negative).Evenour mostexperienced replica makersstill occasionally producetheseartifacts,but the goalis alwaysto minimizetheirfrequency of occurrence. In thisway, eventhoughsomeareas of a replicamightbe bad,we cangenerally findat leastsomeareaswhich

havereplicated wellin allthereplicas of a sequence. TypeA artifacts canbe minimized byusingonlythefreshest resinandcatalyst, whicharethenmixed smoothly withoutwhipping anyunnecessarily largeamount of air intothe

SCANNING

ELECTRON

515

MICROSCOPY

..

.•.

"'"

....

'?'¾'"?•;;L•'•:•'4•:•(. • %;,-:;,•z ..... =.•'....•:' '"'.'?"./' ..... -- j •'•-'•:?ß-•.

'.:.?'.''

: -.•, ,,• "•--":;:..:'. -, .-:.....•

' . .... .... -. . ,....'?,:' "" .:"'•••..' "u•::4•-•.¾%... •.:'•;.?:--:,• '.z ...... •'--•--.•..•?.:' •.'- '"-•-•'••',:'.•.½ •:.';"• •'•'::.½:•';?.•:•'..•,.;.•(¾•-?•,-. :' , ,'7 ß .....-.•'-:::•'•':,•&:? p•½•.'

"-'½'-•-'•

•':•

'.•

.:::-'t.½:' ',. ' .s::---:,'..'..., ....... ...-...-'.'"..•.:'.': •-'•'t•

ß

•

'•••'•:..•..•.... A:.: .. •½:: ...?:.•..,.:;47.,.• :.: •.,..,.•,;•..•:•... ••.:•.-.•-.•.•.•.•-:½•.

¾;.,,•-%---•* --•-••••-••••..:.. :?:t,•.-..½...•b•.•;-.,?•.,;.--•x•Z;... •f -:.?..•..."•:7'-.'.•'.::t'";;;•;-'• '•-...•:••(':..:: •: ....•'•,•:• ..,•::?':--.,•;;;•:'t/;X•*½...:-,,¾':-;•;,:,.•.•. -' ..... :•........' .?. •;;;;'•-r-y.?...:•?•, :-:-:.....•:•., .•:..•?j;::;•.•j•...-':-.•-.•.:.. :•½:•:•z: .... •t;:•:Z?....:..;.•..: ,.,•.•;-;.;...-,.•:•,½.•._.... • ..,•.._•:. .";:%•.:-*:•..•-::;:•'.';..: ..... .:.,• . ,?..• ß.-':::':.•,,½•.:•-'• .,'.½• .....

•. ½•.••..•..

:.... ß .

•...

.- ..•.,•.,:•.:'" ...:•.. • ..........'•••••••••;,...; ...... ......... -.•-•--:.. ..' ,: .' .. ß,.•.-•... •-.,• ..... , . •.:..f•.. ß ,. ß,.. ß '-U•., • -•-•..:'.'.' . ß .... Figure 2. SEM montage o[ "cobblestone"ski•, •ositive replica, male subjectabout 30 years old. T•ie• artifacts:(A) ab bles in negative;(B) air bubblesi• positivea•d (C) glazed o{ improperly mixed replicating material. Scale bar represents1000 •m.

resin.Type B artifactsare causedby eitherunstablepositivereplicatingmaterial,the useof toohigha temperature whenthe positives aremade,or not sufficient contactpressure of the positivereplicamaterialwith the negative. TypeB artifactsarenot asdevastating astypeA, asthe optionalwaysexists

fortryingto makea betterpositive fromthesamenegative. TypeC artifacts arethe mostfrustrating, sincetheycanbe soeasilyconfused with real cosroeticeffects. An emollient, forexample, whichformsa coatingcanlooksur-

prisingly similar totypeC artifacts. Thistypeofartifact isalmost always the resultofeitherimproper mixing oftheresinwiththecatalyst, lessthanactive catalyst, or a resintemperature bel-ow roomtemperature (resinshouldbe storedunderrefrigeration to minimizeautopolymerization of resin).Since manyof themicrographs usedin thispaperwereprepared aslongas4 years ago,notallwfilbeofthesame highqualityasFig.1,andspecial mention will not, in general,be madeof theseartifacts.

Theultimateobjective of thiswork,fromtheverybeginning, wasto use thisreplicaapproach to demonstrate improvement in the condition of the stratumcorncure asa resultof theapplication of a cosmetic product.Hence,

516

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

..

. . •-..,: .•:•.:::½. ...."'•.,, [. " ...:...:•½? ....•1•:.::•.•:.

.•....:.s•;;?;7:d.•.......