FARMACIA, 2015, Vol. 63, 5
ASSESSMENT OF PERFUME INGREDIENTS WITH APHRODISIAC
POTENTIAL BY GAS CHROMATOGRAPHY-MASS SPECTROMETRY
IOANA GAVRIȘ1, EDE BODOKI1*, PHILIPPE VERITE2, RADU OPREAN1
1Department of Analytical Chemistry, Faculty of Pharmacy, University of Medicine and Pharmacy, 4, Louis Pasteur St.,
2Department of Analytical Chemistry, EA 4651 ABTE Aliments Bioprocédés, Toxicologie Environnement, Faculty of
Pharmacy, Université de Rouen, 22, Gambetta St., Rouen, France
*corresponding author: firstname.lastname@example.org
Manuscript received: July 2015
In the present study the composition of some of the best-selling perfumes of recent years has been evaluated, focusing on the
quantitative assessment of potential aphrodisiac ingredients of these fragrances, such as vanillin, ethylvanillin, trans- and cis-
methyl dihydrojasmonate, muskolactone and muscone. Seven samples of different brands (three women’s and four men’s
fragrances) were analysed by gas chromatography-mass spectrometry. All the other major and minor constituents were
assigned based on MS spectra library (Wiley) matching, followed by pattern analysis using chemometric data-mining (PCA).
Trans-methyl dihydrojasmonate was found in all analyzed fragrances, with the highest concentration found in a perfume for
women. Based on the obtained data, it seems that a common pattern of “most wanted recipe” among both men and women
consumers is to be observed, based on some of the classical scents of vanilla, jasmine and musk.
În studiul de faţă, a fost evaluată compoziţia unora dintre cele mai bine vândute parfumuri în ultimii ani, axându-ne pe
analiza cantitativă a ingredientelor cu potenţial afrodisiac cum ar fi vanilina, etilvanilina, trans-metildihidroiasomia, cis-
metildihidroiasomia, muscona şi muscolactona. Au fost analizate şapte parfumuri (trei pentru femei şi patru pentru bărbaţi)
prin cromatografie de gaze cuplată cu spectrometria de masă. Toţi compuşii au fost identificaţi, comparându-se spectrele de
masă înregistrate cu cele din baza de date Wiley. De asemenea, s-a studiat și comparat profilul complet de ingrediente al
probelor de parfum analizate prin analiză multivariată a datelor (PCA). Trans-metildihidroiasomia a fost găsită în toate
parfumurile analizate, având cea mai mare concentraţie într-un parfum pentru femei, sugerând faptul că dintre aromele
clasice de vanilie, iasomie şi mosc analizate, cea de iasomie este preferată atât de bărbați, cât și de femei.
Keywords: perfume, gas-chromatography, multivariate analysis, quantitative analysis
Studies performed in the last years have been
focused mainly on the analysis of perfumes in
terms of potentially allergenic substances and in a
lesser extent in terms of flavours that can attract
The scent marketing industry is an industry of
millions of dollars and still under a continuous
growth. Companies are interested in using those
flavours which consumers like and make them buy
the perfume . Nowadays it is a huge competition
among perfume producing companies and it is very
difficult to find the perfect formula to attract
consumers. Therefore a potential common aroma
pattern in their composition is sought, which could
assure the “winning” blend for men and women.
Moreover, a special attention was dedicated to
perfume ingredients with aphrodisiac potential,
because they could play a key role in the fragrance
choice and buying decision process. It is a well-
known fact that aphrodisiacs may influence your
mood, increase your sexual desire and attract the
opposite sex .
The aim of the present study was to analyse the
composition of some of the best-selling perfumes of
recent years, focusing on the quantitative
assessment by gas chromatography-mass
spectrometry of potential aphrodisiac compounds,
such as vanillin, ethylvanillin, trans- and cis-methyl
dihydrojasmonate, muskolactone and muscone,
being frequently used as perfume ingredients for
base and middle notes in perfumery such as
jasmine, vanilla and musk [7, 8].
Materials and Methods
Materials, reagents and chemicals
Seven samples of the best-selling perfumes of the
global market reported for the timeframe of 2010-
2014 by several beauty market research companies
[9-12], from different brands (three women “eau de
perfume” and four men “eau de toilette”) were
obtained from local perfumeries and stored at room
FARMACIA, 2015, Vol. 63, 5
temperature until analysis. For reasons of
confidentiality, the brand names and manufacturers
are not shown. They are noted as follows: no.1w,
no.2w and no.3w for women fragrances and no.4m,
no.5m, no.6m and no.7m for men fragrances.
Vanillin (99%), mixture of cis- and trans- methyl
dihydrojasmonate (≥96%) were obtained from
Sigma Aldrich (Germany), ethyl vanillin (98%),
muskolactone (98%) were obtained from Alfa
Aesar (USA) and muscone from Santa Cruz
Biotechnology (USA). HPLC grade methanol was
used as solvent throughout the analyses. The
internal standard 6-methoxy-8-nitroquinoline (99%)
was purchased from Acros Organics (Belgium).
Perfumes were diluted with methanol in a ratio of
1:5, 1:50, 1:200 and stored at room temperature
Individual stock solutions of each compound were
prepared in methanol at a concentration of 1
mg/mL. The internal standard solution (0.5 mg/mL)
was also prepared in methanol.
The GC–MS analysis was performed using an
Agilent 7890A (GC system), Agilent 5975C inert
XL EI/CI MSD with triple axis detector and an
Agilent 7693 autosampler from Agilent
Technologies. Separation was carried out on a
Macherey-Nagel Optima 5 MS capillary column
(60 m × 250 µm, 0.25 µm film thickness). Helium
(purity 99.999%) was employed as carrier gas at a
constant column flow of 1.0 mL/min. The GC oven
temperature was programmed from 60ºC to 280ºC
at 3ºC/min, and held 20 min at 280ºC. Splitless
mode of 1 µL sample volume was used for
injection. Electronic impact at 70 eV was used as
ionization mode. The MS spectra were recorded in
scan mode on the range of 50 to 800 a.m.u.
The analytes were identified by comparison of their
retention times and mass spectra with those of
authentic standards. The identification was also
performed by comparing the obtained mass spectra
of relevant chromatographic peaks with
corresponding spectra from the Wiley MS library.
Multivariate data analysis was performed by using
Simca v.13.0.3 (Umetrics, Sweden).
The internal standard method was used for the
From stock standard solutions, calibration solutions
were prepared in methanol at individual
concentrations of 0.1, 0.075, 0.05, 0.025, 0.0125,
0.005, 0.0025 and 0.0005 mg/mL of each
compound. To each of them 0.25 mg/ml internal
standard was added.
To each analysed perfume sample 0.25 mg/mL
internal standard was added.
Results and Discussion
The identification of the perfume ingredients
(Figure 1 A,B) was performed by comparing the
obtained electron ionization mass spectra of
relevant chromatographic peaks (Figure 1 C) with
corresponding spectra from the Wiley MS library.
The high amount of data obtained in the perfume
profiling required specific data mining tools,
successfully applied in solving various basic
analytical problems such as data overview,
classification and/or discrimination and
multivariate regression modelling [13, 14].
Therefore, multivariate data analysis based on
projection methods was performed involving the
entire set of compounds identified by MS library
matching. Considering the nature of the available
data (compounds – as variables, relative peak area
fractions – as X-data) principal component analysis
(PCA) was employed to identify any grouping or
trends with respect of perfume composition. PCA,
besides providing a first overview of the entire
dataset allows extracting and displaying the
systematic variation in the data matrix X. No
outliers were detected in the obtained models.
PCA performed on the available observations (N =
7, seven perfume samples) considering all the
variables (K = 294) with the unit variance scaled
and mean centred X-data gave a two-component
model (Figure 2A), which explained 47.9% of the
variation (R2X = 0.479). Based on the numerous
unique components identified in each of the
analysed perfumes, this PCA-X model
distinguishes a certain grouping of the samples (2w,
3w, 4w, 6m and 7m), whereas samples 1w and 5m
seem to possess a more distinctive ingredient
Loading scatter plot of PCA-X model without
single value variables, coloured according to
relative peak area fractions of ingredients of sample
1w (C.), coloured according to relative peak area
fractions of ingredients of sample 4m (D.).
FARMACIA, 2015, Vol. 63, 5
Gas chromatograms of perfume samples: no.1w (A.), no.4m (B.), quantified components - (1) vanillin, (2) trans-
methyldihydrojasmonate, (3) cis-methyldihydrojasmonate, (4) muskolactone, (5) muscone. Electron ionization
mass spectra of muskolactone (C.)
Score scatter 3D plots using principal component analysis considering all variables (scaling base weight:
centring and scaled to unit variance) (A.), excluding single value variables (scaling base weight: no centring,
scaled to pareto variance)
FARMACIA, 2015, Vol. 63, 5
As expected, by excluding the single value
variables (the unique components of the analysed
samples) the resulting PCA-X two-component
model (K = 81, with scaling base weight - no
centring, scaled to pareto variance) accounted for
almost 65% of the variation in the X-matrix (R2X =
0.646). In this case a tighter sample grouping was
observed (except sample 1w), indicating certain
similarities based on the shared fragrance
ingredients and their relative ratio (Figure 2 B). The
first component, accounting for almost 50%,
captures the overall variation within the entire
dataset. The loadings scatter plot (Figure 2 C and
D), expressing the dominating correlation structure
of the X-matrix, indicates how the X-variables vary
in relation to each other, which one provide similar
information, but also which ones are not well
explained by the model. As it can be seen, the
variables bearing the highest loadings in component
one are trans- methyl dihydrojasmonate and 6-
methoxy 8-nitroquinoline (internal standard),
whereas linalool holds the highest loading in
component two. The sample 1w seems to be
somewhat different in composition compared to the
other samples, and by comparing the loadings plot
coloured according to the relative peak area
fractions of ingredients for sample 1w (Figure 2 C)
and 4m (Figure 2 D) the variables responsible for
this distinction may be pointed out. Therefore,
some of the compounds bearing higher loadings
(e.g. linalool, dihydromethyljasmonate, 6-methyl γ-
ionone) significantly differ in relative content for
sample 1w in comparison with the other analysed
Even if at a first organoleptic evaluation of
perfumes considerable differences may be noted,
the performed multivariate analysis indicates that
the analysed perfume samples, based on a
qualitative and quantitative level, tend to share a
common pattern of ingredients (except samples 1w
and 5m), whereas considering the relative content
of the common ingredients through the exclusion of
the unique fragrance constituents (X = 213), only
sample 1w continues to differ somewhat from the
other perfumes. Therefore, even though perfumes
of completely different origin were considered, in
principle they tend to own considerable similitudes
in terms of the nature and relative percent of
constituents. Of course, more relevant conclusions
may be drawn by performing similar data mining
on a more extensive set of observations (perfume
Seven samples of different brands (three women’s
and four men’s fragrance) were analysed by gas
chromatography-mass spectrometry using internal
The chromatographic conditions were optimized to
achieve the most efficient separation of the samples,
with a special attention for the six reference compounds.
The content of vanillin, ethylvanillin, trans- and cis-
methyl dihydrojasmonate, muskolactone and
muscone in the analysed samples are shown in
Quantitative evaluation of selected perfume ingredients
*w- women’s perfume; **m- men’s perfume
Perfume no.3w is the sweetest perfume amongst the
investigated samples; most probably because it has the
highest concentration of ethylvanillin responsible
for the vanilla scent as a base note.
Trans-methyl dihydrojasmonate was found in the
majority of analysed fragrances, based on which
perfume no.3w is distinguished with a very high
concentration of this ingredient. It is also interesting
to note that ingredients defining the jasmine scent
are also present in men’s fragrance, although being
considered a feminine scent.
As expected, the sweet and animal-based fragrance
of musk is to be found in highest concentrations in
FARMACIA, 2015, Vol. 63, 5
The small concentration of muskolactone and
muscone could be explained by the fact that they
are used as base notes for the musk flavour but also
as fixatives increasing the effectiveness of other
ingredients being called “exalting fixatives” [7, 13].
In the present study the composition of some of the
best-selling perfumes has been evaluated, focusing
on the quantitative assessment of potential
aphrodisiac ingredients in these fragrances, such as
vanillin, ethylvanillin, trans- and cis-methyl
dihydrojasmonate, muskolactone and muscone.
Trans-methyl dihydrojasmonate was found in all
analysed fragrances, with the highest concentration
in a perfume for women, suggesting in a way that
the “winning” blend is the classic aroma of jasmine,
both for men and women.
Even though perfumes of completely different
origin were considered, in principle they tend to
own considerable similitudes in terms of the nature
and relative percent of constituents.
All the calculations were performed in the
DataCenter of the INCDTIM. Authors are sincerely
thankful to Dr. Cristian Morari for the helpful
discussions. This paper was published under the
frame of European Social Found, Human Resources
Development Operational Programme 2007-2013,
project no. POSDRU/159/1.5/S/136893.
1. López-Nogueroles M., Chisvert A., Salvador A.,
Determination of atranol and chloroatranol in
perfumes using simultaneous derivatization and
dispersive liquid–liquid microextraction followed
by gas chromatography–mass spectrometry.
Analytica Chimica Acta, 2014; 826: 28-34.
2. Pérez-Fernández V., González M.J., García M.A.,
Marina M.L., Separation of phthalates by
cyclodextrin modified micellar electrokinetic
chromatography: Quantitation in perfumes.
Analytica Chimica Acta, 2013; 782: 67-74.
3. Mondello L., Casilli A., Tranchida P.Q., Dugo G.,
Dugo P., Comprehensive two-dimensional gas
chromatography in combination with rapid
scanning quadrupole mass spectrometry in perfume
analysis. Journal of Chromatography A, 2005;
4. Minematsu S., Guang-Shan Xuan, Xing-Zheng Wu,
Determination of vanillin in vanilla perfumes and
air by capillary electrophoresis. Journal of
Environmental Sciences, 2013, 25: S8-14.
5. Teixeira M.A., Rodriguez O., Gomes P., Mata V.,
Rodrigues A., Perfume Engineering, Butterworth
Heinemann, Oxford, 2013.
6. Melnyk J.P., Marcone M.F., Aphrodisiacs from
plant and animal sources - A review of current
scientific literature. Food Research International,
2011; 44: 840-850.
7. Miller R., Miller I., The Magical and Ritual use of
Perfumes, Destiny Books, Vermont, 1990.
8. Calkin R.R., Jellinek J.S., Perfumery: Practice and
Principles, John Wiley & Sons, Inc., Hoboken, 1994.
13. Gomes C.L., Lima A.C.A., Cândido M.C.L., Silva
A.B.R., Loiola A.R., Nascimento R.F., Multivariate
Analysis of Perfumes by Ultraviolet Spectrophotometry.
Journal of the Brazilian Chemical Society, 2015;
14. Iacob B.C., Tiuca I., Bodoki E., Oprean R.,
Multivariate calibration and modeling of UV-VIS
spectra of guest-host complexes for the
determination of the enantiomeric ratio of
propranolol. Farmacia, 2013; 61(1): 79-87.
15. Sell C., The Chemistry of Fragrances: From
Perfumer to Consumer, The Royal Society of
Chemistry, Cambridge, 2006.