behaviour of perfumery ingredients in products

8/11/2019 Behaviour of Perfumery Ingredients in Products

1/13

J. Soc.Cosmet. hem. 5,325-3371974) 1974Society f Cosmetichemistsf GreatBritain

The behaviour f perfumery

ingredientsn products

J. W. K. BURRELL*

Presentedat the 2nd Joint Perfumery Symposiumorganizedby

the British Society of Perfumersand the Society of Cosmetic

Chemistsof Great Britain at Eastbourneon 7-9th May 1973

Synopsis-A new technique nvolving the GLC ANALYSIS of PERFUMERY INGREDIENTS

directly from product baseshas been developed.This technique has been used to study how

materials behave in SOAPS and laundry powders after storage under various conditions.

INTRODUCTION

Perfumers and cosmetic chemists have for some time been interested

in predicting ow perfumeswill behave n the variousbasesnto which hey

are to be incorporated.Until now, the perfumerhas had to resort to many

hours of tedious rial and error to ensurea final pleasingeffectand even

then he could not be certain that the effect will hold constant over months of

storageunder possiblyadverseconditions.

Apart from simple frustration and tedium, this situation presents

problemsof economics.t is of little value incorporating elatively ex-

pensivengredientsn a product f theycontributenothing o the odour over

time because f incompatibilitywith the baseor because hey are lost by

chemical eaction or evaporation. t therefore became mperative that

reliableobjectivemethodsshouldbe developed o study heseproblemsof

behaviour n greater detail.

*Proprietary PerfumesLtd, Ashford, Kent.

325


2/13

326

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

HISTORY

Several workers have tried to develop nstrumentalanalytical tech-

niqueso study he ossof perfumeryngredients,articularlywith respecto

soap.Stirasand Demeilliers 1), for instance, eveloped technique f

analysing y glc the vapour surrounding nwrapped erfumed oapafter

storageunder various conditions.Although some conclusions ould be

drawn rom the data obtained, he authorswereunableactually o observe

what was akingplace nside he soap tself.

One question hey were trying to answerwas which materialshad the

highestperfumery alue'.Derivichian 2) had statedearlier hat esters ave

higher perfumeryvalues han alcohols.Stirasand Demeilliers ound from

theirexperimentshat hevapour rom hesoap ontainedigher roportions

of esters e.g. benzylacetate) o alcohols han that whichwas found in the

vapour from the perfuming iquid itself.

Whether this proves a higher 'perfumery value' for the esters s

debatable. t could be argued hat the higherproportionof estersn the

vapour emittedby soap shows hat thesematerialsare lost to a greater

extentand that thereforeess s available or the physical ransferof per-

fume rom the soapbase o the skin.Therehasbeenconjecturehat there s

a virtually nexhaustibleeservoir f perfume n the centralcoreof a soap

tablet and therefore he perfume ngredients re alwaysavailable n their

originalproportion,but no concrete videncewas found to support his

theory. t therefore ecame erydesirableo discover xactlywhat happens

inside he soap,by analysing amples irectly.

TECHNIQUESOF ANALYSIS

Some early work that was carried out in Unilever involved he use of

uv spectroscopy s a method of assessinghe behaviour of perfumery

ingredients. he materialsbenzophenonend anisaldehyde ere incor-

porated separatelyn soap and it was found when the ingredientwas

extractedwith aqueoussopropanolhat a reasonable v spectrum f the

ingredient ouldbe obtained.Using this techniquet wasproved hat for

thosematerialsstudied, he rate of loss of the ingredientsrom an un-

wrappedbar of soap s determined y the rate of diffusionof thosematerials

through he soapmatrix.Thiswasachieved y measuringhe concentration

of the ingredients t variouspointsalong the axes rom the centre o the

surface.t was ound hat the concentrationf the ingredients ashighest


3/13

BEHAVIOUR OF PERFUMERY NGREDIENTS IN PRODUCTS 327

at the centreand lowestnear the surface.Had the rate of evaporation rom

the surfacebeen he rate determining tep, hen of course he concentrations

at pointsalong the axis would have-beendentical.

For some ngredients hat were studied,e.g. benzophenone, o measur-

able change n the concentration t the centreof the soapbar was observed

after 24 weeksstorage.

This technique ad manyrestrictions nd could not be appliedgenerally

to the studyof all perfumeryngredients. irstly, thereare only a limited

number f materials sedn perfumery hichhaveuv absorbingroperties.

Secondly,he preparationof each sample or analysiss time consuming;

and, thirdly, the method does not identify possiblebreakdown products

which may be either falsely dentifiedas the original material or may be

missed ltogether. ny complex reakdown ouldof course e impossible

to diagnose.

A more generallyapplicable echniquewas thereforerequired and glc

appeared o be the most promising.Previousworkers have used this tech-

nique, but they have all employedsomecumbersomeorm of extraction

procedure o isolate the perfume material prior to analysis.Although

preliminary extraction techniques an be used for studiesof a limited

numberof materials, t becamedesirable o developa simpler,more

straightforward ystemor studying largenumberof perfumerymaterials

under various storageconditions n different products. Such a straight-

forward systemhas been usedby Proprietary PerfumesLimited for about

4000 analyseswith goodresults,'-

The approach hat wasusedwas very simple. t involvedplacingsmall

samples f the soap or detergent) n the top of the glc column n the flash-

heatedzone, the idea being hat the perfumery ngredients vaporatento

the carriergas, eaving he soapbehind.Many who have ried this echnique

have ound that the chromatograms ruinedby the presence f largepeaks

from the soap base,as shown n Fig. 1. This problem was overcomeby

utilizinga modified njectionsystem. he samples placed n a sample ube

which s then inserted n the injectionhead; the latter is then flushedwith

carrier gas. Once the correctcolumn nlet pressures reached he sample

tube is injected nto the flash-heated one of the glc column.After a short

period the sample ube is removed.

In thisway t was oundpossibleo eluteall of the perfumeryngredients

from the samplewithout getting he undesired omponentsrom the base,

someof which are probablyproducedby thermalbreakdown.Complete

elutionof the ingredients as provedby the re-injection f the sample


4/13

328


o

O O O O O O

13 (D b-- 0D 03 --

I I I I

o o o o o o o o o


5/13

BEHAVIOUR OF PERFUMERY INGREDIENTS IN PRODUCTS 329

when no residual racesof the ingredientscould be found. Lossesof the

majority of perfumerymaterialsduring the weighingand sample oading

procedurewereminimaland a coefficient f variationof d-7 relativewas

obtained for the method when analysinga mixture of cineole, inalool,

benzyl acetate,benzylbenzoateand diethyl phthalate n superratted oap.

To carry out controlledstorage ests t was necessaryo use a stable

referencematerial which would be incorporated nto the product with the

ingredient under investigation. t was also desirablethat this reference

material shouldbe odourless o that parallel odour assessmentsould be

made.

Diethyl phthaIate DEP) was selected s it is both odourless nd stable

in soap bars under the conditionsof storage hat were to be used n the

test. The suitabilityof DEP was confirmedby incorporatinga mixture of

benzophenone which had previouslybeen shown to be stable by uv

measurements)nd diethyl phthalate n soapbars and analysing amples

from the centreof eachbar after variousperiodsof storage.

The behaviourof each ngredientwas studied n the followingmanner.

A mixture of the ingredientand diethyl phthalate (ratio of about 1: 1)

was incorporatedn superratted oap at 0.2; the soapwas plodded nto

bars 3 x 1 and wrapped n standardglassine nd glazedpaper wrapping.

The barswere then stored or both 12 and 24 weeksboth at room tempera-

ture (c. 20) and at 37.Samples ere aken rom the centreof eachbar and

analysed by glc. Chromatogramswere compared with those obtained

from freshlypreparedbars of perfumedsoap 1 week old to allow equili-

bration of distribution).The incorporationand storagewere arrangedso

that analyses ould be carriedout within a 24 h period.

RESULTS

Table gives he results btained or somecommonperfumerymaterials.

Although some esultshave beenomitted, no generalconclusions ould be

drawn as to relationships etweenbehaviourand chemicalclassexcept n

the caseof formates,which all showeda tendency o hydrolyseand pro-

duce the parent alcohol. Low boiling materials generallywere lost more

readily than the higherboiling materials.Samples aken from the surface

layersof the wrappedbars had the samecomposition s those aken from

the centre,whichproves hat any lossother than by chemicalbreakdown

is governed,n the caseof the wrappedbar, by the rate at which the ingre-

dientmigrateshrough he wrapper nd hencento the atmosphere.part


6/13

330


Table I

material incorporated present after:

Material in soap 12 weeks 24 weeks

at RT at 37C at RT at 37C

Hydrocarbons

Diphenylmethane 95 67 82 61

Camphene 62 62 50 38

Limonene 85 41 81 17

Myrcene 100 20 50 10

a-Pinene 100 40 50 4

Alcohols

n-Nonanol 100 100 100 100

cis-p-tert-Butylcyclohexanol 100 100 100 100

trans-p-tert-Butylcyclohexanol 100 100 100 93

tt-Terpineol 100 100 100 90

Bomeol 100 96 100 89

Benzyl alcohol 100 87 100 81

Caryophyllenealcohol 100 81 88 75

Linalool 100 78 100 53

cis-Hex-3-enol 86 53 46 35

trans-Hex-3-enol 88 56 49 37

Esters

Citronellyl acetate 100 100 100 88

Bomyl acetate 100 88 83 83

cis-p-tert-Butylcyclohexyl cetate 100 100 100 89

trans-p-tert-Butylcyclohexylcetate 100 100 100 78

Terpinyl acetate 97 97 85 82

Inonyl acetate 84 56 64 49

Methyl heptine carbonate 67 31 59 13

Benzyl acetate 64 41 49 0

Benzyl formate 0 0 0 0

Citronellyl formate 0 0 0 0

Trimethylhexyl formate 0 0 0 0

Linalyl formate 0 0 0 0

Phenylethyl ormate 0 0 0 0

Aldehydes

Amylcinnamaldehyde 100 100 100 100

Hydroxycitronellal 100 100 100 98

Undecenal 100 96 100 86

Cinnamaldehyde 70 76 63 60

cis Citral, neral 95 58 68 53

trans Citral, granial 97 58 74 52

Hydratropic aldehyde 62 36 33 16

Benzaldehyde 0 0 0 0

Phenylacetaldehyde 0 0 0 0


7/13


Table I--continued

material ncorporated resent fter:



Ketones

t-Ionone 100 100 100 100

Benzophenone 100 100 100 100

Jasmalone 100 93 100 89

Methyl nonyl ketone 83 75 79 64

Methyl hexyl ketone 40 20 14 0

Ethyl amyl ketone 0 0 0 0

Ethers

Amyl benzyl ether 100 100 100 79

[3-Naphthylmethyl ether 94 79 96 78

Anther 100 77 73 58

Rose oxide 83 52 65 35

p-Cresyl methyl ether 100 63 80 9

Phenylethyl methyl ether 53 44 25 9

from the formatesonly a few materialsshowedsignsof chemicalbreak-

down.

Phenylacetaldehydes sounstablehat about60o was ost 1 weekafter

incorporation.Benzylacetate,whichhas often beenstudied n this context

in the past by others 3), is among hosematerialswhich showchemical

instability,but the amount of benzyl alcohol produceddoesnot account

for all the loss and thereforeevaporationmust also be a contributory

factor.

After the initial screening f individual materials t became nteresting

to examinesome ime-honoured eliefs hat perfumershave held in con-

nection with the behaviour of various ingredientsand mixtures. For

example, t has beenwidely believed hat the behaviourof aldehydes an

be improvedby the additionof an alcohol,and in particular he corre-

sponding lcohol, because f the supposedormation of hemiacetals.n

order to test the validity of the argument,mixturesof the aidehyde, he

correspondinglcohol and diethyl phthalate in the ratio 1: 1: 1) were

incorporated nto soap bars and stored under the same conditionsas

described reviously.Table H gives he resultsof the analyses nd shows,

within experimentalerror, that there is no difference n the behaviour of

aldehydes ith or without he alcohol.


8/13

332


Table II

material incorporated present after



n-Octanal 80 35 30 4

n-Octanal (+ n-Octanol) 90 40 28 6

n-Nonanal 100 75 100 55

n-Nonanal (+ n-Nonanol) 100 72 100 50

Citronellal 91 93 83 59

Citronellal (+ Citronellol) 95 91 80 57

Benzaldehyde 0 0 0 0

Benzaldehyde + Benzyl alcohol) 0 0 0 0

1 week

at RT at 37C

Phenylacetaldehyde 37 0

Phenylacetaldehyde-t- Phenylethylalcohol) 37 0

Another belief which was nvestigatedwas that concerningfixation'. It

has long been thought that the addition of high boiling materials educes

the lossof low boiling ingredients rom perfumes.Although this has been

well establishedor perfumesapplied to the skin it has never beenproved

for perfumes n products. t was possibleusing his methodof analysis o

show hat evenwhen he additionof a fixative HercolynD*) wasmade at a

level five times that of the perfumery material, no improvement n the

retentionof the material n the soapcould be observed.

The lack of detectable formation of hemiacetals and the ineffectiveness

of fixativesn soapare not surprising henoneconsidershe vastdifferences

in the rates of collisionbetween he moleculesof perfumery ngredients

themselves nd between he ingredientmolecules nd the soapbasemole-

cules.The number of collisionsbetween ndividual perfumery ngredient

moleculeswould be extremely ow in soapas comparedwith those n the

essencend therefore he effects f physicalor chemical nteractionbetween

suchmolecules ould be correspondinglymall.Any supposedlyeneficial

effectsof interactions hat take place in the essence efore ncorporation

and which are reversible,as in the casesof hemiacetal ormation, and the

physical interactions associatedwith 'fixatives', will be subsequently

minimized fter ncorporation f the essencen the productbase.

*Hercules Powder Co.


9/13


Table III

Material in laundry powder


12 weeks 24 weeks

at RT at 37C/ at RT at 37C/

70 % rh 70 % rh

Hydrocarbons

Diphenylmethane 62 24 72 33

Alcohols

Phenylethyl dimethyl carbinol 100 100 100 95

Caryophyllenealcohol 100 100 100 88

Dodecanol 100 100 100 70

Decanol 100 84 92 67

Anisyl alcohol 64 40 45 19

Borneol 77 31 37 0

Benzyl alcohol 65 18 31 0

Linalool 38 8 27 0

Esters

Musk oxalide 100 100 100 100

Benzyl benzoate 100 100 100 71

Gardocyclene 94 69 98 70

Ethyl cinnamate 59 67 63 38

Geranyl acetate 67 12 54 0

Citronellyl acetate 65 0 31 0

Benzyl acetate 18 0 13 0

Aldehydes

Hexylcinnamaldehyde 100 100 100 100

Anisaldehyde 30 14 17 13

Hydroxycitronellal 80 40 75 0

cis Citral, neral 40 0 0 0

trans Citral, geranial 40 0 0 0

Cinnamaldehyde 28 0 0 0

Phenylacetaldehyde 20 0 7 0

Decanal 14 0 0 0

Ketones

Versalide 100 100 100 100

Benzophenone 100 90 91 89

Celestolide 100 100 100 71

Methyl naphthyl ketone 100 100 100 71

-Ionone 86 26 78 4

Jasmalone 92 20 72 0

Acetophenone 0 0 0 0

Ethers

Methyl diphenyl ether 100 100 100 83

[-Naphthylethyl ether 93 76 97 66

Phenylethyln-butyl ether 44 15 38 0

Phenylethyl amyl ether 35 0 23 0

Amyl benzyl ether 40 0 12 0


10/13

334


Apart from the studyof the behaviourof ingredientsn soap base, a

similar detailed study has been carried out in laundry powders.The per-

fumery ingredientswere incorporated ndividually with diethyl phthalate

into a laundry powder (in this casePersil Automatic)at a level of 0.1.

The productswere henstoredn cartons or periodsof both 3 and 6 months

at both room temperatureand 37C/70 relative humidity. Again the

incorporationand storagewere so arranged as to enable the analysis o

be made n a period of 24 h. The chromatograms btained rom the stored

sampleswere comparedwith those rom fresh samples ept in glassars.

Table III shows the behaviour of some common perfumery ingredients

relative o diethylphthalate.*

One can see from the table that many perfumery ingredientsdo not

perform as well in packetsof laundry powder as in wrappedsoapbars.

Once again similar conclusions ere reached,namely that lower boiling

materialsare lost more readily than the higher boiling ones,and that there

are no obvious inks betweenchemicalstructuresand stability.

*The results are relative for it has been shown that there is loss of diethyl phthalate under the

most severe onditionsof 24 weeksat 37C and 70% rh amounting o about 20%.

Table IV



12 weeks 24 weeks


70 % rh 70 % rh

n-Decanal 14 0 0 0

n-Decanal (-t- n-Decanol) 15 0 0 0

n-Undecenal 50 0 29 0

n-Undecenal (-t- n-Undecenol) 41 0 22 0

n-Dodecanal 83 0 74 0

n-Dodecanal (-t- n-Dodecanol) 80 0 68 0

Linalyl acetate 32 0 12 0

Linalyl acetate -t- Hercolyn D) 30 0 13 0

Citronellol 100 74 74 30

Citronellol (- Hercolyn D) 94 61 74 33

Cyclamen aldehyde 81 0 59 0

Cyclamenaldehyde -t- Hercolyn D) 80 0 49 0

Phenylethylamyl ether 35 0 23 0

Phenylethylamyl ether (- Hercolyn D 45 0 25 0

Jasmacyclene 78 14 59 11

Jasmacyclene-t- Hercolyn D) 85 18 59 14


11/13


Hemiacetal formation from aldehydes and the fixative effects of

HercolynD were investigatedor ingredientsn laundry powder and once

again these effects were not demonstrable,as shown by the results in

Table IV.

These resultsshowed hat the majority of perfumerymaterials do not

performwell in laundrypowders tored n cartons, hereforea second est

was undertaken n order to differentiate etween he lossby evaporation

and that by chemical nstability. To achieve this objective the product

containing he ingredientwas stored n glass ars. The resultsof this test

are given in Table V, which dearly demonstrateshat in the caseof this

particular laundry powder he lossof ingredients bserved n packetswas

causedmainly by evaporation.

Table V


o material incorporated present after

12 weeks 24 weeks


70 % rh 70 % rh

Linalool Carton 38 8 27 0

Glass 100 100 100 98

Phenylethylalcohol Carton 66 66 52 17

Glass 100 97 93 80

Benzyl acetate Carton 18 0 13 0

Glass 100 83 75 50

Citronellyl acetate Carton 65 0 31 0

Glass 100 100 100 86

Linalyl acetate Carton 32 0 12 0

Glass 100 100 100 100

Citral (cis and trans) Carton 40 0 0 0

Glass 100 100 100 92

Dihydrojasmone Carton 92 20 72 0

Glass 100 79 83 37

a-Ionone Carton 86 26 78 4

Glass 100 100 100 93

Phenylacetaldehyde Carton 20 0 7 0

Glass 100 100 100 100

Amyl benzyl ether Carton 40 0 12 0

Glass 100 100 100 75

Phenylethylamyl ether Carton 35 0 23 0

Glass 100 100 100 93

Phenylethyln-butyl ether Carton 44 15 38 0

Glass 100 100 100 100


12/13

336


APPLICATION OF RESULTS

Clearly the fact that evaporations a major causeof perfumery ngre-

dient oss rom laundrypowderswarrantsnvestigationsnto the efficiency

of variouspackagingmaterials.The loss of perfumecan be as much as

three imeshigher n the leastsuitable ype of packet han someof the best

packets sedcommercially, nd the glcmethodcanbe of use n discovering

the bestpackage ompromisen termsof effectivenessnd cost.

Anotherareawhere he glc echnique as oundapplications in market

research. he ability to analyse omplete erfumes irectly rom a product,

which is possibleby attaching he glc to a mass spectrometer, ermits

studyof perfumerychemicals sed n relation o market trendsand fashions.

This type of information s of immense alue o a companyn a competitive

situation.

Perhapshe mostvaluable esultof this study,however, s that it is now

possible o create perfumeswhich are chemically stable and relatively

unaffected y evaporation, venwith completely ewproductbase ormula-

tions. In thesecases here are no historicaldata for the perfumerso call

upon. More detailed investigations ave shown that the materials n a

complexperfumemixture behave n the sameway as when ngredients re

incorporated ndividually. Therefore t is possible o reduce the total

numberof analysesequiredby screening large numberof ingredients s

a seriesof mixtures, hereby acquiringexpert knowledge n a relatively

shortperiod.

OTHER BASE APPLICATIONS

All of the work describedso far has been concernedwith two major

products, .e. soapsand laundry powders.However, the techniquehas

widerapplication nd t hasbeengenerally mployedn our laboratoriesn

the analysisof perfumesand flayours n other bases.For example, t is

possible o study he behaviourof flayours n toothpastebases,perfumes

in haircreams nd in talcumpowders, nd the technique asevenbeenused

as a method of analysing he essentialoil of lavandin directly from a single

floret. The technique an in fact be usedwhenever n analysiss required

of any volatilematerial present n a relativelynon-volatilebase.

(Received:30th January1973)


13/13

BEHAVIOUR OF PERFUMERY NGREDIENTS IN PRODUCTS 337

REFERENCES

(1) Stiras, . andDemeilliers, . Studyby gaschromatographyt'odorant apour mitted y a

perfumedcake of soap. Recherches 4 33 (1964).

(2) Dervichian,D. Role de la structuremoleculaire ans a fixationdesparrums ans e savon.

La France et sesParrums324 (August 1961).

(3) Shiftan,E. and Feinsilver,M. Practical esearch f the art of perfumery.Ann. N.Y. Acad.

Sci. 116 692 (1964).

behaviour of perfumery ingredients in products

Documents