Training Perfumery

Training the ABC's of Perfumery

The training of staff at TCFF, TCFF, (Thai-China Flavours and Fragrances Fra grances Industry Co. Ltd., Ayuthya, Ayuthya, Thailand) with guest post-graduate students and professors fro a!or universities in Thailand, on "The #erfuery $usiness". (under a grant fro the %ational &cience and Technology Technology 'evelopent Agency) This is paper was pulished in odified for in  #erfuer  Flavorist, Allured #ulications, *&A. + Address correspondence to &tephen . . 'owthwaite, #erfuersorld, email: "perfuer/#erfuers "perfuer/#erfuersorld.co" orld.co"

© Stephen V. Dowthwaite $ang0o0, Thailand.

Training the ABC's of Perfumery © Stephen V. Dowthwaite $ang0o0, Thailand.

"Letting a hundred flowers blossom and a hundred schools of thought contend is the policy" i Introduction

The training of any discipline is never an easy tas0. This is the case even when there are clearly defined rules and paraeters such as in language training. #erfuery is one language in which even the ost asic concepts have not yet een clearly defined, leaving us without definitive understandings of perfuistic ters. An individual1s understanding of odours, sells or aroas is usually the result of his culture, personal e2perience and environent rather than training. There also appears to e an underlying feeling that until the iological echanis of sell is fully understood then a unified language is perhaps ipossile or of liited use. 3owever, try descriing 1'eussy1s - L1Apres-idi d1un faune1 in ters of wavelength, pitch, aplitude or an 4oughs "&unflowers" using the 54$ colour syste. 6ur true understanding is not as greatly enhanced as we ight hope, in 0nowing and eing ale to easure the science of sound and and light o!ectively o!ectively.. I contend in the art, or is it science7 of perfuery that it is not necessary to wait for the coffin to e closed on the precise anatoy of olfaction and gustation  efore we can a0e good use of a language of sells. A language hoping to e useful should e relatively easy to grasp and learn, e fle2ile, offering the user a tool for deeper research and understanding. The seeds for the syste that I have een teaching since +8 are ased on old 0nowledge and follow the general principle laid out in 3aaran  5eier1s "$oo0 of #erfue" ii . ithout a ention of en9yl salicylate (orchid) or cis-:-he2enol (leafy-green), the oo0 showed us that the words and ters for ade;uate odour description already e2ist in the apanese represent the native languages of the students. hilst loo0ing for coon ground upon which to start it  ecae ;uic0ly apparent that every student was een ale to recite the
8

6dour descriptions can suffer daage fro graatical rules iposed y the
The approach adopted here is a pragatic one vi, orn of the desire to teach the art of  perfuery systeatically. It has een developed for its speed in training and it1s fle2iility in use. hilst we ay wait for a perfect understanding of the echanis of sell I suggest this wor0ale approach can lead to a reasonaly coplete odour classification syste for use in initial perfuery training and as a tool for ore advanced activities. This is erely one contending voice to descrie a hundred flowers. =ost students starting training are initially aware of only "li0e" and "disli0e", a few are ale to relate to specific food, flower and household sells ut within a few hours of using the syste ost can classify odours ;ualitatively and ;uantitatively. Step  ! "et's start at the #eginning with the $a%#%c$ 's ! The A-? syste originally egan as a way to avoid unnecessary duplication of wor0 y  perfuers. I have lost count of the nuer of ties I have een as0ed for a atch of a specific faous, ar0et-leading soap fragrance, and how any ties I have started a0ing the fro scratch. In the ore coon ethods of archiving saples, they are stored in a nuerical,  product or chronological syste. In these filing systes, siilar selling perfue saples are 0ept in copletely different locations.
The syste also gave perfuers ;uic0 access to siilar, previously copleted fragrances, allowing the to learn fro archived saples rather than continually re-inventing the sae wheel. It was then that raw aterials could also e grouped in a siilar fashion, so that for e2aple, woody selling aroa cheicals were 0ept with other woody selling aterials. As I ecae increasingly involved in training, I loo0ed for a siple noenclature that non-native
:

It was found that they could also e arranged roughly in ters of their appearance in the evaporation stages of a perfue. A-3 loosely represent notes apparent in top notes. I-& represents iddle notes T- represents the later stages of evaporation. 3owever, as in any two-diensional representation of a ulti-diensional property, it can never e copletely accurate at every level, and a little latitude is re;uested. %aes were chosen to e general rather than specific ("dairy" rather than utter or il0) to avoid having too any specific connotations ($utter with utyric esters, il0 with lactones etc.). To e too specific would e too create such a long list of groups that it would defeat the purpose. %aes were odified not only to fit conveniently in the A-? arrangeent, ut also to indicate the wider eaning of each group. The ter "aliphatic" has een chosen rather than "aldehydic" to represent the "fatty" characteristic of straight chain aldehydes, alcohols and fatty acids. 3ers are differentiated fro spices as cool or hot respectively to help students separate the as two distinct groups, rather than confusing oth as condient sells. >asin, =uguet (the French is used to stop the confusion with other lily species) and 5ose are 0ept as specific groups ecause, not only are these odours well-0nown and recognised, ut they also for the central thee in the clear a!ority of fragrances. "Bueen" of the 6rient was chosen to eody the concept of the rich, e2otic character of en9oin and other resins. &olvents and other low-odour additives, such as anti-o2idants were relegated to solvent with a "?" ("?olvent") to ensure that they could e conveniently grouped together at the end of a forula in line with the convention of forula writing.



S&ST(ATIC )D)*+ C"ASSI,ICATI)- +)*PS ,)+ P+,*(+& A%

C0assification

Common Description

1ey +eference (ateria0s

A

A"I%,AT%IC

Fatty, a2y, &oapy, Clean

Aliphatic Aldehydes, Alcohols

B

Berg % ICB+

Cooling,

=enthol, Caphor,
C

CIT+*S

&our, &harp, Citrus peel

Citral, 6range, Leon, Lie

D

DAI+&

=il0y, Crea, $utter, Cheese



DIB"

egetale, %ut, Fish, =eat

'iacetyl, $utyrates, Lactones, alerates Thia9oles, #yra9ines, &ulphides

,

,+*IT

&our, &weet fruits, &trawerry

Allyl caproate, erdo2



+-

Cut-grass, Leaves

cis-:-3e2enol, Triplal

2

2+B 3Coo04

Cool 3eraceous notes

Lavender, &age, (Terpene ased7)

I

I+IS

6rris, iolet

Ionones, =ethyl Ionone,

5

5AS(I-

Fruity, 6ily, %arcotic, >asin

3CA, $en9yl Acetate

1

1)-I,+  

#ine, #ineneedle

Terpineol, $ornyl Acetate

"

"I2T Chemica0 ,0ora0

Fresh light floral cheical

Linalool, ertene2, '=$C

( (**T

Lily of the alley, 4reen, Fresh 3ydro2y, Lilial, Lyral

-

-A+C)TIC

3eavy &weet Florals,

%arcissus, lang lang, Tuerose

)

)+C2ID

Aroatic, 'eep floral

&alicylates, $en9oates

P

P2-)"

#henol, =edicinal, 3oney

p-Cresol,
6

6ueen of the )+I-T

5esin, $alsa

$en9oin, Tolu, Terpenes

+

+)S

5ose 6tto, Asolute, 4eraniu

Citronellol, #
S

SPIC 32ot4

3ot Culinary, &pice

Clove, Cinnaon, Thye, (Cyclic7)

T

TA+ S()1

&o0e, Tar, $urnt

Cade, $irch Tar  

*

*rine ,aeca0 A-I(A"

Anial, Faecal, Leather

Civet, Castoreu, Aergris

V

VA-I""A

&weet
anillin, Couarin, 3eliotropin

7 7))D

ood, 6ily,

Cedar, &antal, etivert, #atchouli

8

8%rated (*S1 

&e2y, =us0, &ensual, &weet

=us0 Detone, 4ala2olide

&

A+T2& ()SS&

east, Fungal, =oss, =arine,

6a0oss, Calone

/

/)"V-TS

6dourless &olvents,

'<#, '#4, I##,
E

 Step 9 ! )dour profi0es of sing0e materia0s


#henyl ava Lavender 6il $ergaot 6il

6uantified description

5ose GH 4reen :GH oody EH
@

Step : ! +e0ati;e Impact % a measure of 'strength' 3aving learnt to classify the odour ;ualitatively the student continues the process to add a ;uantitative factor. That is, to evaluate the relative strength of one aroa cheical or essential oil to another. The e2pression 1ipact1 has een used in preference to 1strength1 ecause 1strength1 creates the ipression that a aterial with a low ipact has little effect on a copound.


Step < ! )dour 0ife

It is not long efore the student notices that a aterials 1relative ipact1 changes over tie, alost invarialy ecoing wea0er. %ow the student applies another ;uantification in the for of the odour life of the aterial. 6dour life is deterined on the selling strip (thinner chroatography 1paper1 gives ore consistent results) to the point at which the aterial ecoes wea0 and uncharacteristic of itself. *nderstandaly the results of this type of e2aination are very dependent on the aount dipped, aient teperature, huidity, air-flow and testers1 differences and e2perience. 'espite this even in poorly controlled conditions the student is ale to produce a set of coparative values that are worthwhile easures of raw aterials relative lotter lives. A aterial can further e evaluated y the 1volue1 of odour that it produces over its evaporation. 6dour 1volue1 could e defined as a saple1s ipact ultiplied y its lotter life. Impact = hours 0ife on the sme00ing strip. (ateria0

Impact

"ife

Vo0ume

Linalool #
+GG +E EEG :G

+ 88 + GG

+GG ::G EEG +8GGG

Step > % Synaesthesia % the #0ending of the senses

There is a theory that at irth our senses are inter-lin0ed. As aies when we heard sounds we saw colours or patterns, when we touched we e2perienced sells. As we developed, the theory goes, we learnt to differentiate the sensual e2periences and copartentalise each one into a definite single sensation. There are a handful of people, however, who have aintained soe of these lin0s etween apparently 1disparate1 senses into adulthood. This helps to e2plain


Synaesthesia responses to odour

A%

Co0our

C0assification

Synaesthesia response

A

A"I%,AT%IC

&ilver-hite

running-water, neutral, lead-guitar, ird-

B

Berg % ICB+

Frosty-hite

cue, cooler, ell, dru, clear 

C

CIT+*S

6range ellow

sall ell, flute, sour, 1999......1, sunshine

D

DAI+&

Crea

stic0y, oily, creay, sooth



DIB"

ellow-$rown

rule, rain-falling, alond-shape

,

,+*IT

$right 5ed

Denny-4, iddle 1C1, elliptical



+-

$right 4reen

sharp, leaves, ower, chal0-on-lac0oard

2

2+B 3Coo04

'ar0 green

roantic, traditional, ice-s0ating, rela2ing

I

I+IS

#urple

sad, old-house, crea0y-wood, dull, serious

5

5AS(I-

ellowish-hite

garland, classic, !a99, !olly, sooth

1

1)-I,+  

Light Dha0i-

ells, 2as, wind-lowing, velvet, ountain

"

"I2T Chemica0 ,0ora0

#astel ellow

slow-usic, happy, lovely, piano, harsh

( (**T

#ale 4reen

light-usic, garden, tender, soft-wind-lowing

-

-A+C)TIC

'0

cigarette, headache, deep-sleep, train, ites nose

)

)+C2ID

#urple

elegant, lues, !a99, a-fine-day

P

P2-)"

'ar0 4ray

to2ic, "po0..po0", fi99y, Indian-dance, faric

6

6ueen of the )+I-T

'eep 6range

Indian-song, u99ing, eauty, rough

+

+)S

#in0

love, water-fall, !a99-violin, classic-song

S

SPIC 32ot4

5edJ6range

ig-ell, arching, e2citing, sha0ing

T

TA+ S()1

$lac0

--;, so0e, ig-dru, isty, >ohn-'enver  

*

*rine%,aeca0 A-I(A"

'irty

dirty, 9oo, sniffing-sound, hairy, can1t-see, sour 

V

VA-I""A

Light $rown

laughing-children, ay-chatter, ca0e, ees-

7 7))D

$rown

ar0, log, dru, hard, water-fall

8

8%rated (*S1 

4ray-hite

forest-sounds, running-deer, cuc0oo, earth,

&

A+T2& ()SS&

4reen-$lac0

udding, oss, peace-in-the-forest, oist

/

/)"V-TS

Colourless

water, water-fall, clear, harsh, tu..tu



"et's picture the sme00

The results of the colour - sell relationships are then applied to drawing a graphical representation of a fragrance1s odour &o instead of displaying a description of an odour in a written fro it can e displayed for very ;uic0 understanding in the for of a pie graph. (Colours represented here ay not e optiu due to restrictions of printing or graphics capailities) For the e2aples aove

Phenyl Ethyl Alcohol

30% Rose Green 70%

Bergamot Oil 2% Citrus 40%

Lt. Floral

58%

Herb

Lavender Oil

5% Lt. Floral 30% 60%

Ice Sice

+G

Step ? ! 'Crac@ing' formu0as with the system The A-? classification is then applied to siple florals and cople2es as they are introduced and later developed to 1crac0 the code1 in lists of cheicals and naturals found in forulas and 4C data output to produce odour descriptions of the copounds. To deonstrate the power of the syste spend a few inutes loo0ing at the forula  elow and try to create an ipression of the odour type. <2perienced perfuers will e failiar with this type of e2ercise ut it is y no eans easy and the outcoe is usually ore of a guess. 4eran l acetate :. ACA @. $en9 l acetate :.E $en9 l ut rate 8. $er aot 6il . $isaolene +.@ $our geonal 8.+ $rost rol, etaG.@ Calone G.+ Cedro2 de 8.: Citral G.88 Citronellyl acetate 8.E 3 dro2 citronellal . Civettone +.@ 'ecanal G.@@ 'ih dro couarin +.: 'ietol +.
++

 %ow in the sae forula the raw aterials have the principal classification ar0ed in the left hand colun and the forula has een sorted A - ?. Try to repeat the e2erciseM

A A A A C C C C I I  J  J  J  J L L L M M M N O O O O R R R R R R U U U V V V W W W     ! 

!.2 0.47 0.66 0.28 4.7 0.22 !.$ !.$ !.8 0.3! 3.5 2.$ 6.8 !.6 0.7$ 4.7 7.0!$ 3.4 2.! 0.$3 4.7 !.3 0.83 6 0.68 3.5 ! 3.8 4.7 2.6 2.5 0.43 0.23 0.28 0.28 !.3 0.00! 4.6 !.6 2.3 3.5 !.6 !.! 3.8 0.!$

Lauric Aldehyde *ndecanal 'ecanal +G-undecen-+-al $er aot 6il Citral 'ietol  %erol Ionone 6rivone $en9 l acetate $en9yl utyrate ACA 3CA 3edione  %erolidol Isophytol Linalool $ourgeonal Lilial 3ydro2ycitronellal lan lan 6il Isout l salic late #henyl ethyl salicylate $rostyrol, eta$en9oin &ia 5esinoid 4eranyl utyrate 4eranyl acetate #
+8

"# 2.6!

C# 8.72 I# 2.!!

%# !5.5$

L#!5.!!

&# 7.73 '# !.3

(#

R#

)# 0.7$

*#

+# 7.4

,# 6.5 -# 0.!$

,ragrance V % Did you 'crac@' it yet 

For the e2perienced perfuers the ist has proaly cleared. It can ;uic0ly e seen this is an aldehydic floral with a strong !asin, rose, orchid and uguet heart and a sweet vanillic, anialic, us0 ac0ground. #erhaps a odern interpretation of a Chanel style fragrance. This sae data can e represented in 4raph for for even easier appreciation. #lease note that the graph is %6T ad!usted for the ipact of individual aterials - an ad!usted graph is shown at the end of the article.

Frarance /*/  Co1oun % o i1act  2n classiiers

&us9 7% "liatic 3% Citrus $% +oo 7% Iris 2%

*anilla 6% "ni1al !%   as1in !5%

Rose !4%

Lt. Floral !5%

(rci !!% 'arcotic!%&uuet 8%

+:

Step  ! Su#%C0assifiers

The ne2t step is to ascrie a su-classifier to the aterials odour. There are any thousands of odours ut over-classification leads to a cople2ity that does not act as an aid ut rather a hindrance. It is e;ually clear that 8@ classifications is hardly sufficient to descrie the thousands of odours that we use in perfuery. Cedarwood and &andalwood ay oth e oody  ut can hardly e said that they possess close odour profiles. For this reason a second letter is applied to indicate, for e2aple, which type of woody note we are tal0ing aout. This gives us 8@ (A-?) 2 8@ (a-9) K @@ theoretical classifications. In practice we find that with the e2ception of F-Fruity notes usually only E or so su-classes are necessary for each classification. The su-classification is used less rigidly. In the aove forula sorting e2ercise the su-classification, if added, would ensure that siilar odour aterials appear ne2t to each other. For e2aple, etivert is listed on ad!acent lines to etiverol, ertofi2, Iso < &uper and Dephalis or that Allyl Caproate is listed with Allyl Cyclo 3e2yl #ropionate and Allyl Ayl 4lycollate. This ay prove particularly useful for flavourists in adapting the syste for their use. The suclassifiers listed here are ased priarily on the secondary odour note. eg. $lac0currant is a fruity( ,) note with a inty( #) ac0ground so its classification  ecoes ",#". This is always done to give the est sorting of aterials and in soe cases the rule is odified to acheive this.

+

Su# C0assifications 3a %  0ower case4

(ain C0assification A

A"I%,AT%IC

Aa%Aliphatic Alcohols, Ac%Aldehydes C-+G, Ar-Aldehydes C++-

B

Berg % ICB+

Bh%
C

CIT+*S

Ca%6range, Cc%Leon, C@%Lie, C0%$ergaot, Cn%=andarin, CE% %eroli, Cw%#etitgrain

D

DAI+&

D#%$utter =il0, Dd%Lactone D;%Coconut



DIB"

g%&ulphur, h%egetale, p%Coffee, t%=eat, u%Fish, w% %ut

,

,+*IT

,a%Apple, ,#%$lac0currant, ,d%#each, Apricot, ,e%'urian, ,g%=elon, ,i%5asperry, ,E%$anana, ,o%Cherry, ,n%4rape, ,p%#ineapple, ,;%&trawerry



+-

g%Fresh-grassy, h%4alanu, n%#henylAcetAldehyde

2

2+B 3Coo04

20%Lavender, 2#%&age

I

I+IS

If%Ionone, Iw%=ethylionone, Ia%6rris, Ig%iolet

5

5AS(I-

5f% Fruity-!asin, 5n% >asin-asolute

1

1)-I,+  

10- #ine, 1h- #ineneedle

"

"I2T Chemica0 ,0ora0

"w%Linalool, "#%cool cyclic esters, "r%'=$C, "w%ertene2

( (**T

(g%4reen-fresh, (0%Flowery

-

-A+C)TIC

-E%lang, -n% %arcissus, -o%Tuerose

)

)+C2ID

)o%&alicylates, )n-$en9oates, )o%Aroatic notes, )n%6rchid

P

P2-)"

Pp%=edicinal, P;%3oney

6

6ueen of the )+I-T

6;%$alsa, 6n%5esin

+

+)S

+0%6tto, +r%Asolute, +0%Flowery

S

SPIC 32ot4

S#%Anise, Sp%Clove, So%Cinnaon, Sh%3ot "3ers"(Thye)

T

TA+

Tt%&o0e

*

*rine%,aeca0 A-I(A"

*u%Civet-Faecal, *w%Aergris, *p%Castoreu-Leather 

V

VA-I""A

V;%anillin, Vt%CaraelJFuranones, Vh%Couarin

7 7))D

7#%#atchouli, 7F%&andal, 7w%Cedar, 7y%etiver

8

8%rated (*S1 

8=%=acro  #olycylcic =us0s, 8o% %itro

&

()SS& A+T2&

&h%=oss, &p%&eaweed, &g%69one, &y%=ushroo

/

/)"V-TS

/%'<#, /o%$en9yl en9oate

+E

Step G ! ,uy 0ogic she0f arrangement

The student now applies the syste to his arrangeent of his perfuery organJwor0 ench. 5aw =aterials ottles are laeled with coloured laels to represent each classification group A-?. The students perfuery-organ is arranged according to the alphaetical classification code. 'uring the copounding process the student is encouraged to investigate the ad!acent  ottles as alternatives to the indicated aterial in the forulation. This "Fu99y logic" vii syste helps to uild a uch wider 0nowledge of raw aterials in a shorter tie than if the aterials are arranged according to their identity or trade nae. Coloured laels assist in ;uic0 replaceent of the aterial to it1s correct position on the shelves. The student is advised to change the laels if necessary in light of additional e2perience ut reeering to change this on his coputer inde2 as well so that oth records correspond. This ensures that if the student is loo0ing for a specific aterial, that he has forgotten the odour of, then he can find it1s reference ;uic0ly.

+@

Step H ! )dour description and o0factory ana0ysis % 'odour%prints' and odour profi0es

As their copetency increases the syste is used for odour description and producing an 1odour-print1. I use the ter 1odour-print1 as it ore accurately descries the process than a  profile. It is ased on the analogy of fingerprint analysis used y the police who use +@ points of reference (whorls, rea0s and ends in lines, etc.) that appear to e uni;ue for the pattern in ;uestion. They do not descrie the pattern of the whole fingerprint only the +@ points of reference. In the fragrance odour-print then we go through a process of loo0ing at the 8@ attriutes ( A-? ) in the odour of a fragrance and ;uantify each attriute with a proportionate value. A true profile would re;uire that we ta0e oth the ain classification and the suclassification, which is covered in the second stage of our analysis. The student is supplied with a saple of the fragrance on a selling strip and as0ed to systeatically go through the 8@ classifications, A to ?. The student deterines if the fragrance has A - Ali-Fatic notes, if %o then he oves on to $ if es then he gives a proportionate value. #ercentages can e calculated at the end of the process. The principle with this process that I stress with students is  it is often as useful to deterine "what the fragrance is not" as it is as to "0now what it is" . Fre;uently, when selling a saple for the first tie, particularly if we have no nae or visual pointers then a student or perfuer1s reaction can e "no idea what it sells of" or "I 0now it ut it1s on the tip of y tongue". $y systeatically going through the A-? then we can rea0 this 1ental loc01 relatively easily. In addition to deterining for e2aple, that a specific fragrance is fruity and green, we also have eliinated the presence of ali-fat-ic or spice notes. 6f course these deterinations can  e corrected at any stage in the process ut we ;uic0ly otain a starting point to ove off fro. e go on to uild a fairly accurate description of a saple1s odour systeatically. The process ay go soething li0e this for soething li0e this for "Fragrance N" M (a 1L1air du teps1 style fragrance)

+

The odour%print of ,ragrance '8' 

A-? A $ C ' < F 4 3 I > D L =  % 6 # B 5 & T *   N  ?

=ain Classification Ali-fat-ic Ice (Ice$erg) Citrus 'airy asin Donifer Lt. Che. Floral =uguet %arcotic 6rchid #henol 6riental (Bueen of the 6rient) 5ose & pice &o0e (Tar) Anial (*rine-Faecal) anilla oody =us0 =ossy (
Is this note present or not7 %o %o es %o %o es es %o es %o %o es es es es es es


es es %o es es es es es %o

8G +8 8 8 : E + -

+

G.E G.E : +E +E 8 +G + :

A second pass (A-?) of the positive scores identifies specifically which su-classification applies and supplies ore detail to turn our odour-print into a description of the fragrance

A%/

7hat Su#%c0assification 

C F 4 I L =  % 6 # B 5 & *   N 

$ergaot #each Flowery 4reen =ethyl ionone $ois de 5oseJ$ergaot Light Floral Flowery lang &alicylates 3oney $en9oin 5ose de =ai type CloveJCarnation Civet, aergris anilla etivert %itrous0 6a0oss

Coining data fro the odour-print scores and the su-classifications the following description is gleaned M $,ragrance 8$  A heart (main) of Rose de Mai, Carnation, Light Floral Muguet with notes of Spice, rchid   A top note of !ergamot and nuances("#$) of Flowery %reen and &each notes  !lended with !ois de Rose, Mus', ris, lang,  Fi*ed with !en+oin, anilla, Am-ergris, Ci.et and Moss *sing the odour%print and description to produce a fu00 odour profi0e

The student ared with this description then sits at his perfuery-organJwor0ench and oves through the aterials arranged A-?. 3e chooses those that he thin0s fit est into each classification of his proposed fragrance. 3e will proaly choose for any of his classifications ore than + aterial to acheive a etter atch to the target sell in the classification. In this event the student calculates the proportion of each aterial to a0e up that classification, without reference to its relative ipact. In the e2aple aove 8GH of the sell of 5ose is re;uired in the copound and the su-classification deterines that it should e a 5ose de =ai type note. hen the student sits at his perfuery organ he perhaps decides that the following three aterials are necessary in these head-space proportions to achieve this particular note re;uired. +ose de (ai type note (total 8G) !

:en;l
+

!0 $ !

An aside ! The proportions in the head%space of a compound are not the same as the formu0a proportions

At this stage it is good to ephasise that the student1s description applies to the sell ("head-space") of the fragrance and not the percentage coposition in the forula. The  percentage of fragrance aterials necessary in the forula to achieve this "head-space" are deterined y factors of vapour pressure, attraction etween the olecules within the fragrance itself and the sustrate that is eing evaluated on (e.g. selling lotter). The easure that we use in this syste to replace these factors is the relative ipact of the aterial (which ta0es in these aspects in a generalised for). In essence the proportion necessary in a copound to produce lets say +GH of a low ipact aterial in the head-space will need to e uch higher than +GH. In contrast a high ipact aterial it would need to e consideraly less than +GH in the copound. The following siplified forula is offered to appro2iate the ad!ustent that is necessary for each aterial to ensure that the proportion in the copound is aout right to give the target head-space proportion re;uired. Proportion in CompoundJ K Target 2eadspaceJ = 3LL M +e0ati;e Impact4

e.g. +) I want to add +:H of 3ydro2ycitronellal to the "head-space". 5elative Ipact K 8E #roportion in CopoundH K +: 2 (+GG J 8E) K E8 To get +:H of the sell of 3ydro2ycitronellal in the "sell" I need to use E8H in the forula e.g. 8) I want to add G.EH of #henylacetaldehyde EGH to the "head-space". 5elative Ipact K8:G #roportion in CopoundH K G.E 2 (+GG J 8EG) K G.88 To get G.EH of the sell of #henyl acetaldehyde in the "sell" I need to use G.88H in the forula

8G

The reconstruction of ,ragrance 8 !

2 A e % a / d % S p a c e

Su#%C0ass

+aw materia0 choice at the students perfumery organ

+(. Prop

+(. Impac t

+(. Prop = 3LLMImpact4

J

C E F G. E 4 G. E I :

$ergaot #each

$ergaot 6il &icilian Aldehyde C+

E G.E

++G GG

.EE G.+:

8.@ G.G

Flowery 4reen

#henyl Acet Aldehyde EGH

G.E

8:G

G.88

G.+8

=ethyl Ionone

L

$ois de 5ose

Iso-a-=ethyl Ionone 6rris Concrete $ois de 5ose Artificial

8. G.8 +E

G G +GG

:.++ G.E +E

+.@ G.8 .+G

Light Flowery

3ydro2ycitronellal

+:

8E

E8

8.G

lang &alicylates

Cyclaen Aldehyde lang lang oil <2tra $en9yl &alicylate

8 8 

G +8G 8E

8.88 +.@ :8

+.8G G.G +.8

3oney $en9oin Flowery type

#henyl
8 + : +G

EG @E @G @G

 +.E E +@.@

8.+@ G.: 8.G .GG

Clove

5hodinol 5ose Asolute Clove 6il

 + @

+GG +8G 8GG

 G.: :

.@ G.E +.@8

Acetyl iso eugenol Civet Artificial Aergris $ase 3eliotropin etiveryl Acetate =us0 Detone Celestolide ugoslavian 6a0oss As.

@ + + 8 :  + +

G EGG 8E +GG G EG +E G

@.@ G.8  8 :.E  @.@ 8.E

:.@G G.++ 8.+@ +.G 8.G8 .:8 :.@G +.:E

G>.9:

LL

+ E = + E  % 6

# B 5

&

8 + G + : 8 G

+ 8

 8  : N E

Carnation Civet Aergris anilla etivert %itro=us0



6a0oss

*

8

+

T)TA"S

3ead-&pace K proportion in sell 5= K raw aterial 8+

#rop K proportion

88

Step L ! Creati;e use of the system

Is the syste then only a tool for atching7 %o, while it has een easiest to deonstrate the principles against a specific saple sell that can e analysed e;ually y each student or oserver, the true strength is in the opportunities that it opens for creative wor0. hat is iportant is that the student uilds his inspiration for a fragrance into a concept that he can test his idea against. hen one atches a fragrance the goal is defined clearly  ecause the result can e tested y coparison of the original saple and the at ched saple y any third party. It is all too easy in a creative pro!ect to say that the first saple ade is 1nice1 and use this as !ustification to conclude the pro!ect. This is why students are re;uired to define their o!ectives for their creative fragrance pro!ect concretely, either visually, in the for of a  phrase, diagra, picture or even usical score. The description ust include in written for or  y inference the eotional response e2pected fro the fragrance at its copletion. The point is not for the result to e !udged y a critic ut rather that the student does not alter the eory of his goal to fit the result he has otained. If we accept what we a0e as creative ecause we li0e the result, or our colleagues li0e the sell, then this reduces the act of creation to an act chance. 3ow long will it ta0e +,GGG  perfuers on a +,GGG alances to reproduce "Chanel %o.E" 7 5ather if the act of creation is a discipline then we should e ale to visualise the result that we re;uire in such a anner that we can sell it, in our ind1s eye, efore the first aterial goes into the ottle. The process of atching, I encourage, e viewed as a ethod of learning procedures, accords and techni;ues rather than an end in itself.  Surely, if we follow the leaders then we will always be behind them.

There are any ethods (perhaps ore than there are perfuers) to turning an inspiration into a forulation. 3ere I shall suarise one ethod that I have found students can ;uite ;uic0ly relate to. It uses the "=ind =ap" syste developed y Tony $u9an viii. &tudents choose a su!ect, which can e an occasion, a place, a person, a landscape, an event or any su!ect that can e visualised y that person in detail. They draw a circle +G c1s in diaeter on a piece of paper, A: or larger. In the circle they draw a picture that represents the essence of their su!ect. For e2aple if they had chosen "an evening in the city of #aris" the central picture ight e the
inforation aout the su!ect and e;ually iportantly the 1feel1 or atospheric attriutes of the su!ect. This point is often where students who had approached their training in perfuery as a scientific e2ercise suddenly see the artistic e2pression that is possile within perfuery. A 1crystallisation1 of thought can occur and I have seen students ecoe ;uite eotional at this  point. &tudents then give a proportionate value to each grouping that they have circled. Changes to any of the ranches can e ade as necessary if the student feels that the concept is oving too far away fro his original intention. *sing the circles1 data as a guide, the student writes a concise sentence or short paragraph including the ain features of the fragrance and it1s eotional essage that it should convey upon its copletion.

8

An ;ening in the City of Paris

The student then applies the syste A-? as outlined fro &tep  and uilds his first forulation. hilst the aove e2aple is intentioned to produce a 1fragrance of the heart1, the ethod is easily adapted to coercial applications and can include concepts such as "cover the ase sell", deographics1s inforation, age, class, pricing, etc. Those interested in the su!ect are directed to Tony $u9an1s wor0 "The =ind =ap $oo0", $$C +:.

8E

raphic )dour Design

Adding the diension of ;uantification to a syste of classification allows the possiility to produce graphic odour-prints and full profiles of current forulations. Listing forulations in their groups A-? then applying the factors of ipact gives us the opportunity to predict the fragrance1s odour and display it in graph for. 6vious oversights such as the inclusion of +H 'iethyl &ulphide in the forulation ;uic0ly ecoe apparent and the forulation can e corrected efore copounding.

"ragrance #V# $ %ead$&'ace Odo(r )inc*Im'act + ,nd Cla--i.ier-/


+oo; 5%

*anilla 5%

"ni1al 3% Citrus

%$Fruit 3%

Rose !5%

Green 0.3% Iris 4%

(riental 2%   as1in !2%

(rci 5% 'arcotic 4% &uuet 5%

8@

Lt. Floral 8%

Also if the odour-print and profiling ethod can successfully predict a copound1s odour consistently then it is theoretically possile fro a precise enough description to produce a copound forula autoatically, using the very sae steps that the students have een following with the aid of coputer algoriths - 4raphic 6dour 'esign. Fragrance , was a coputer generated forula using the 4raphic 6dour 'esign i9ard in " The #erfuer1s or0oo0" i2 fro a foratted description. The forula was not odified (y huans) for the purpose of this presentation.

8

i5 eferences! =y than0s to the "students" that helped in the preparation of this article especially  'r. #rateungsri &inchaisri, &irinan Taptiteace, $angon #raditphet, &upi Tipprasongporn, &asitorn Dietthana0orn Address correspondence and en;uiries aout !oining perfuery training courses toM &tephen . 'owthwaite, #erfuersorld. Ltd.. www.#erfuersorld.co

=ao Tse-Tung , =ay +E@

ii iii

"3  51s $oo0 of #erfue ol. I" >ohnson, London +E "3  51s Fragrance 4uide  Feinine %otes, =asculine %otes " ean-%oel >auert, Claude Tapiero and >ean-Christopher 'ore (Article  #erfuer  Flavorist =ayJ>une +E)

iv

"6dour 'escription and 6dour Classification" 3arper, $ate-&ith  Land, Churchill, London +@

v vi vii viii

"6dour #references" 5 =oncreiff, Leonard 3ill, +@G "Critical Thin0ing" Francis atanae 'auer, 62ford *niversity #ress % + "Fu99y Thin0ing  The new science of Fu99y Logic" $art Dos0o, 3yperion % +: "The =ind =ap $oo0  5adiant Thin0ing" Tony and $arry $u9an, $$C London, +:

i2 "The #erfuer1s or0oo0" &tephen . 'owthwaite, &oontorn Dietthana0orn. $ang0o0 + indows ased perfuer1s raw aterial dataase and forulation anageent progra. Free download at www.perfuersworld.co.