]. Soc.Cosmet.Chem., 21,595-605 (August 19, 1970)

Toothpastes--Cleaning andAbrasion j. B. WILKINSON,

M.A., B.S.c.,F.R.I.C., and B. R. PUGH,

B.Sc., Ph.D.*

PresentedDecember1, 1969, New York City

Synopsis--It is generallyrecognized that the •najorityof peopleneedsomeform of abrasive in their TOOTHPASTES if theyare to maintainsatisfactory tooth cleanliness.SuchABRASIVESvary dramaticallyin their ability to a}bradedental mineral and also in their CLEANING PERFORMANCES in toothpastes, thus making it necessary that soundtest methodsare available for measuringsuchcharacteristics. Abrasion of dental mineral is measured by a RADIOTRACER

TECHNIQUE

in which

radioactiveteeth are brushedwith a toothpasteslurry and any abradedparticlesfound subsequentlyin such a slurry are monitored by normal radiotracer techniques. Measurement of cleaningis madeusinghigh-precision photography,in which picturesare takenof subjects' teeth after they have used each test paste for two weeks. A secondtechnique has been developedin which the soil differencesare examinedby an oral photometerconstructedfor this purpose.

For most abrasives,there is a direct relationship between the abrasivity towards dental enamel and the cleaningperformance. This relationship can, however,be broken with an abrasiveof correctlyselectedphysicalproperties. The high cleaningand low abrasivequalities of such an abrasive

are described.

INTRODUCTION

One of the primaryobjectives of a toothpaste is to cleanteeth [ectivelyand safely. This relativelysi•nplecriterionarisesbecause most individuals feela needto maintaintheir teethfreefrown unsightlystains andpotentiallyharmfulbacterialplaque. It mustbe recognized, however,that individualsvarygreatlyin their ability to build up a visible

stainon theirteeth(1). Heavystainers(so•ne 15-20%of thepopulation) can build up a visiblestain in as little ti•ne as one week, while at * Unilever Research Laboratory, 455London Road,Isleworth, Middlesex, England. 595

596

JOURNAL

OF THE SOCIETY OF COSMETIC

CHEMISTS

the other extreme, there are thosewho develop little or no visible stain. The heavy stainer thus needsa high-cleaningtoothpastebut the light staineris not socritical. With the vastmajority of productsnow on sale, the cleaningt•unctionis achievedby useof an abrasivepowderand greater cleaningability is invariablyassociated with greaterabrasivity. If it were an easymatter for an individual to recognizehis propensityto develop visiblestain,then he might be advisedwhich ot•the many brandsof toothpasteavailableto him would adequatelyclean his teeth. He could then be assuredthat the abrasivityof his chosenbrand would be no greater than necessaryto perform adequately,but it would seemnot yet practicable

to cater to the individual.

In this situation,a greaterunderstandingof the act of cleaningby abrasionwasclearlyvery desirable,particularlyto examinewhether or •lot higher cleaninginevitably meant higher abrasion. This, then, led to a tribologicalstudy to obtain knowledgeof how dental surfacesare worn by abrasiveparticlesand the relationshipof the nature of wear to the hardness, shape,and sizeof the particlesand to the propertiesot•the substrate. At the sametime, it wasnecessary to conductwork in the evaluation or assessmentarea so that in vitro results could be related to

performancein vivo. TOOTH

ABRASION

Abrasion ot•dentalmineralhasbeenexamined by manyworkers and thehistoryof itsdevelopment is tracedin a paperby Pughandhiscol-

laborators (2). Asdescribed bythese authors, thepreferred technique is directandreliable. Naturalteethareprepared with an evenenamel or dentinesurface exposed.Theyarethenirradiatedby neutronbom-

bardment andwearis measured by established radiotracer techniques. This, we believe,hasthe closest possible relationship with the in vivo situation in that the teeth are brushed with a normal toothbrush and a

slurryof toothpaste.The toothisnormalexceptinsofarthatit isradioactive,but it hasbeenestablished that,underthe right conditions of irradiation,thishasvirtuallyno effectuponthewearcharacteristics. Recentpublications by Stookeyand Muhler (3) and Wright

Stevenson (4) havealsodescribed essentially identicaltechniques. Experi•nentally,the techniqueconsists of mountingthe radioactivetooth specimenin a trough and brushingit with a toothbrushimmersedin an

aqueous slurryof toothpaste.The brushis drivenmechanically in a reciprocating actionand is weightedto simulatethe pressure exerted

TOOTHPASTES--CLEANING

597

AND ABRASION

when the individual brusheshis teeth. Abraded debris from the tooth

surface passes intosuspension where it canbemonitored byanyoneof the standardradiochemicalcountingtechniques. In the normal course of events, two teeth seldom, if ever, have ex-

actlyidentical levels ofabrasion because theywillvaryin size,shape, and hardness.This makesit difficulttocompare toothpastes whichhavebeen used on different teeth. To overcome this, a standard toothpaste or

abrasive isusedandall products arecompared with thisstandard which is arbitrarilygivena valueof 100. Providedthe standardremainsunchanged overthe years,datacan be compared evenwhenthe experi•nentsare carriedout severalyearsapart. The experimental designis to abradefour teethusinga slurryof the standardpaste,then to follow thiswith slurriesof the testpaste.

Eachslurryis monitoredfor radioactivity(i.e., numberof countsper second). The ratio of the countsfor the test and standardpastesis called the abrasionvalue, which, when •nultiplied by 100, givesa scale on which the standardhas the value of 100. Some typical values for enameland dentine abrasionmay be seenin Table I. Table

I

Enamel and Dentine Abrasion of Various Toothpastes Enamel

Code A B C D E F G

Toothpaste ExperimentalanhydrousDCP paste Commercialchalk-basedpaste Referencepaste (40% coarsechalk) Commercialchalk-basedpaste Commercial Urlium • paste Commercialchalk/DCP2H20 b Commercialchalk/DCP2H20

Abrasion 208 4150 4100 73 445 431 q20 q-

35 29

20 17 15 14

Dentine

Abrasion 135 4118 q100 100 4105 q58 q41 q-

14.3 13.3

12.2 12.6 9.7 8.7

• "Urlium" is a trade mark of Unilever Ltd., used in this context to identify an aluminum

oxidetrihydrate. The registeredmark is not applicablein the USA at the time of publication. bDicalcium phosphatedihydrate.

The datawereanalyzedusinga standardstatisticalanalysisof variance to derive the confidencelimits. In the precedingtable, thesevaluesare seento vary with the sizeof the mean becauseof the conversionof the scaleused,toassignthe arbitraryvalue of 100 to the referencestandard paste. In suchan example, the limits are for the 95% confidencelevel. The table clearly demonstratesthe large differencesin abrasion capacitiesof varioustoothpastes with different abrasivecontents. Though

598

JOURNALOF THE SOCIETYOF COSMETICCHEMISTS

selected fromthe U K market,thesepastes are typicaland,asseenfrom Muhler'sdatafor USA pastes(3), the samerangeof differences occurs there. Normally,sucha rankingordercanbeusedto predictthelikely cleaning abilities. With theexception of theUrliumpastethisis true because, as;viii be seenin the cleaningsection,the rankingordergiven in cleaning assessment isnormallythesameasthatfor enamelabrasion or dentine

abrasion.

CLEANING

ASSESSMENT

The measurement of cleaningis much more difficult. One of the reasonsis becauseinterpersonalenvironmentalinfluencesare very large and alsothere are problemsof biologicalvariationswhich are met in suchfields,coupledwith a real difficultyin recording. Differentworkers have overcomethese problemsin different ways. For example, Loberie (5) wasable to differentiatebetweencleaningcapacitiesby carefully selectinga panel of individuals,eachof whom had visiblestain on histeeth,andgettinga dentalhygienistto cleantheir teethfor themwith a particular pasteand an electrictoothbrush. Plaque disclosingsolutions were used to show the different

levels of stain.

We preferred to have each individual to use each and every toothpaste,thereby allowing intercomparisonsbetween pasteson each individual (2). This minimizesthe interpersonaldifferences. Sucha procedurehowevermade it necessaryto have permanent photographicrecordsof the individuals'teeth after the use of a paste. The difficulty in standardizing photographto photographto permit them to be judgedby assessors requiresprecisephotographicprocedures. The Lobene type of test measuresthe extent of removal of stained material from the teeth and contraststo the methodwe have adopted whichmeasures the extentof build-upovera 2-weekperiod. It hasbeen foundthat the rankingof productsis the samefor both techniques. An essentialfeature of the test describedhere is that the panelists shouldbe heavystainers. Suchindividualsare drawn from the general population,havingbeenselected in a preliminarytrial in whichtheyhad useda low-cleaning toothpaste for 2 weeks. Photographs of their upper and loweranteriorteethmustshowthe obviouspresence of stainedsoil.

Onceselected,their teethare scaledand polishedand a photographis takento recordthe initial "clean"state. Sufficientsuppliesof the toothpasteare handedout with instructionsthat panelistsshoulduseit in their normal manner, usuallyonce morning and night, for 2 weeks. At the

TOOTHPASTES--CLEANING

AND

ABRASION

599

end of this time, the subjects'teeth are again photographedto record the "after 2 weeks"use. Onceagain,eachsubject'steethare scaled,polished, and photographed. The next toothpasteis handedout to be usedfor 2 weeks. Theseprocedures are repeatedfor eachproductto be tested. Examinationof Fig. 1 will showthat for eachpastethere will be an

initial andan "after2 weeks"photograph for eachsubject. Comparisons are madebetweenthe "after 2 weeks"photographs, while the initial photographs are simultaneously examinedto ensurethat any visible stainsweredeveloped duringtheperiodof useandwerenot presentat the startof eachfortnight.

Figure1. Schematic flowchartof the variousoperations whichcotnprise the cleaning assessment tcchnique

This callsfor a highlevelof skillin assessment of photographs and carefulattentionmustbe givento selectingand trainingthe assessors. Figures2 and3 arefurtherillustrations of thespecialized instrumentation

whichgoes towards makingthistechnique reliableandalsoof thephotographswhich are used(6).

Assessments are normallycarriedout by two operators who work to-

gether,alternately takingtherolesof assessor andrecorder.In thisway, the assessor workswithoutknowledge of whichphotographs relate to which paste.

For eachpastetherewill be a pair of photographs, one the baseline

(prophylaxis) photograph andtheotherthe"after2 weeks use"photo-

600

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

"•..:•..•4 '"•

-..•

Figure 2. Intraoral camerasystemin useduring a cleaningtrial

t

"

Figure3. Typical photographprepared•or assessment

TOOTHPASTES--CLEANING

AND

kBRASION

601

Figure4. Photographic printsarrangedfor assessment

graph. The recorder setsup twopairsof suchphotographs in a simulated daylightcolor-matching cabinet (Fig. 4). The assessor compares

thetwo"after2 weeksuse"photographs for differences in soilbuild-up, takinginto accountanydifferences in print intensities and referringto the baseline photographs for any soil which maynot havebeencompletelyremoved at thestartof the2-weektest. Fourupperincisors, four lower incisors,and two lower canines,making ten teeth in all, are as-

sessed in eachpairedcomparison.This procedure is repeatedfor every intercomparison betweenpastesuntil all the panelmembershave been examined.

Eachtoothis comparedindividuallyagainstthe sametooth on a differentphotographanda decisionismadeon a yes/nobasis,i.e., the question is askedwhetheron onephotographa tooth is cleaneror dirtier than the same tooth on the compared photograph. Contrary to the proceduredescribedearlier (2), the assessor is not askedto assess the margin of differenceon an 0-3.0 scale,but only whether or not a differencedoes exist. This type of preferencetest makes the decisionsmore clear cut and therefore

easier for the assessor.

The data from theseassessments are then analyzedusing a standard statisticaltechnique which is often referred to as a multiple regression techniqueleadingto an analysisof variance. From this analysisit is possible to rank dentifricesin order of their cleaningabilities and to examine

602

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

differencesbetweenproductsfor statisticalsignificance. To aid in this, the data output is expressedas an hierarchy. For example, a typical result may be communicated as shown in Table II. Table

II

Ranking of Various Dentifricesin Order of Cleaning Ability Paste G

Paste D

Paste E

Paste B

Paste A

0

129

321

399

414

Significant differences Level

Difference

1 : 20

146.5

1 : 100

193.8

1 : 1000

249.3

The numbersassignedto each pastecan be usedto showwhether a statistical differenceexistsbetweenproducts.Thus,in thecitedexample, PasteG is significantlyinferior to PasteE, but PasteE is not significantly different from Paste13. It must be appreciatedthat such numerical valuesserveonly to establisha testfor significance.They do not representan absolutescaleof cleaning,nor can they be usedto sayone product cleansso many times better than another. This is because,when constructing suchan hierarchy,the lowestpasteis arbitrarily given the

value of zero. This would be obviousif all the pasteswere high cleaners,for, whicheverpastehad the lowestscore,it would be madezero despitethe obviousfact that, in anothertest,that pastecouldbe shown to be superiorto otherlowercleaningtoothpastes. A measure of the absolute differences between dentifrices can be

madewhenthereis a considerable wealthof datafor eachproduct. In thiscase,and in contrastto the comparison methodalreadydescribed,

each"after2 weeks"photograph is compared with each"afterprophylaxis"photograph for everysubject whohaseverbeenaskedto undergo thetest. Assuming thata dentalprophylaxis represents the stateof best practicalcleanliness,it is possibleto showthat one dentifrice will build

up x% lessstainin 2 weeksthananotherproduct. In thisexperiment, each product is comparedwith a standardclean state,which contrasts

withtheearlierdescribed testin whicheachproduct wascompared with anotherproduct. Statistically, thelatterprocedure ismoresensitive and capableof greater discrimination.

To referonceagainto Table I in whichproducts wererankedin order

of theirenamel abrasivities, it isclearthattherankingof cleaning abili-

TOOTHPASTES--CLEANING

AND

4_BRASION

603

tiesseenin Table II is in exactagreementwith the exceptionof PasteE which,contraryto its relativelylow enamelabrasivity,is rankedas a

highcleaner. On the otherhand,thereis a muchstrongercorrelation betweencleaningand dentine abrasion.

Alongsidethismethod,a secondtechniquewasdeveloped. This was becausewe recognizedthat the "photographic"techniquedescribedis largelysubjectivein nature and that greaterreliability may be achieved by instrumentation,particularly of the assessment of tooth cleanliness. For thisreason,a "photometric"techniquewasdeveloped. In this paper only its salientfeatureswill be illustrated(Fig. 5).

Figure 5. Photometricdevicewhich quantitativelymeasuresthe presenceof tooth soil

Esssentially, the instrumentlooksat an areaof the lower anterior teeth and detectschangesin optical brightnesswith the presenceof tooth soils. Suchchangesare monitored by an optical systemcoupled to a sensitive photomultiplier.

The photographic stages of the testpreviouslyoutlinedare replaced by photometric readings. Thus, insteadof theinitial photograph, an initial readingof the cleanstateis taken. This is followedup a fortnight later by a secondphotometricreading. As before,eachpanel member useseachtoothpastefor a fortnightly period. A typical result is illustrated in Table

III.

604

JOURNAL

OF THE SOCIETY OF COSMETIC Table

CHEMISTS

III

BrightnessChangesfor Various Toothpastes

% Change in Brightncss Paste

after 2 Weeks'

A

3.1

B

3.4

E

3.5

D

10.9

G

13.2

H (based on PVC) •

18.6

Use

This is a polyvinylchloridepowdermanufacturedby ICI as"Corvic" P65/50.

It will be noted that the lowestchangein brightnessoccurswith the highestcleaner. This is becausethe subjects'teeth start in the dean state,i.e., with a dental prophylaxis,and therefore the bestcleanerswill show least tendencyfor the teeth to become stained. Toothpastes such as H, containing PVC as the main abrasive,and normally consideredto be ineffectivecleaners,will showthe greatestchangebecause the teeth build up considerablestain and hence move further from the initial

clean

state.

For any one testusinga selectedpanelof people,the resultsare consistentand reproducible. Similarly,the ranking order is the samewith differentpanels. However,experience hasshownit to be dangerous to placetoomuchemphasis on themagnitude of the brightness changes becausetheseare dependentto a degreeon the composition of the panel. For example,the brightness changes for a panelof very heavystainers wouldbe greaterthan for a panelof medium-heavy stainersthough,of course,the overallrankingof productswouldbe unchanged.

Agreement betweenthetechniques is excellent;in fact,therankings areidentical. Also,thesensitivity of thephotometric technique isof the sameorderasfor the photographic, i.e., productscan be categorized as high, medium, or low cleaners. CONCLUSIONS

Heavystainers canreadilydifferentiate between thecleaning propertiesofdentifrices.Whilethose dentifrices containing chalk,anhydrous DGP,oraluminum oxidetrihydrate arelikelytohavesatisfactory clean~ ingproperties, it is doubtfulwhetherproducts based on hydrated DCP or PVCwould. On theotherhand,individuals whohavevirtuallyno

TOOTHPASTES--CLEANiNG

AND

ABRASION

605

stainingproblem do not need high cleaningdentifricesand, in fact, wouldnot be expectedto noticeanydifferencesin cleaningperformances. Cleaningis, for the mostpart, related to the abrasiveness of the agent used. Lobene found that the stain removing property of a dentifrice wasnot solelyrelatedto its abrasiveness towardsdentine (5), thoughother researchers(2-4) have shown that the largest contributory factor for cleaningperformanceis the abrasiveand, further, that there is a relation. ship between cleaning and dentine abrasion. Enamel abrasion and

cleaningare usuallysimilarlyrelated,but the relationshiphasbeenshown to be brokenfor a particularabrasive. Presumably,sucha deviationoccursbecause of the propertiesof the abrasive(hardness, shape,size,etc.). SUMMARY

Dentifrices

of various formulations

have been examined

for their

abrasivehess towards bothdentineandenamelusinga standardized radiotracer technique. Cleaningdata have been obtained for the samedenti-

fricesbya precision photographic technique andalsoby usingan instrument designedto detectchanges in the opticalbrightness of teethwith varying degreesof associatedstain.

Cleaningand abrasionare, for the mostpart, directlyrelated,and higher cleaningcorresponds to greater abrasivehess.In the caseof

enamelabrasion, thisrelationship canbe broken,presumably asa result of theproperties of theabrasive (i.e.,hardness, shape, andsize). (Received December 9, 1969) REFERENCES

(1) Kitchin, P. C.,andRobinson, H. B. G.,Howabrasive needa toothpaste be? ]. Dent. Res.,27, 501-6 (1948).

(2)Bull,W.H.,Callender, R. M.,Pugh, B.R.,andl¾ood, G.D.,Theabrasion andcleaning properties of dentifrices, Brit.Dent.J., 125,331-7(1968).

(3) Stookey, G. K.,andMuhler,J. C.,Laboratory studies concerning theenamelanddentine abrasionproperties of commondentifricepolishingagents,J. Dent. Res.,47, 524-32 (1•68).

(4) Wright,K. H. R.,andStevenson, J. I., The measurement andinterpretation of dentifrice abrasiveness, J. Soc.Cosmet. Chem.,18,397-411(1967).

(5) Lobene, R. R., Effectof dentifrices on toothstains withcontrolled brushing, J. Amer. Dent. Ass.,77, 849-55(1968).

(6)Callender, R. M.,Techniques ofintra-oral photography, Photog. J.,106,327-g2 (1966).