essential oils of rosa

Research Article

Received: 10 October 2012 Revised: 8 December 2012 Accepted: 12 December 2012 Published online in Wiley Online Library

Rapid Commun. Mass Spectrom. 2013, 27, 591–602

Gas chromatography combined with mass spectrometry, flameionization detection and elemental analyzer/isotope ratio massspectrometry for characterizing and detecting the authenticity ofcommercial essential oils of Rosa damascena Mill.

Federica Pellati1*, Giulia Orlandini1, Katryna A. van Leeuwen2, Giulia Anesin1,2,Davide Bertelli1, Mauro Paolini2, Stefania Benvenuti1 and Federica Camin2**1Department of Life Sciences, University of Modena and Reggio Emilia, Via G. Campi 183, 41125 Modena, Italy2Food Quality and Nutrition Department, IASMA Research and Innovation Centre, Fondazione Edmund Mach, Via E. Mach 1,38010 San Michele all’Adige (TN), Italy

RATIONALE: The essential oil of Rosa damascena Mill. is known for its fine perfumery application, use in cosmetic pre-parations and for several pharmacological activities. Due to its high value, it can be easily adulterated with flavors orcheaper oils. This study is aimed at a detailed phytochemical characterization of commercial samples of R. damascenaessential oil and at their authenticity assessment.METHODS: Nineteen commercial samples of R. damascena essential oil of different geographic origin and an additionalauthentic one, directly extracted by hydro-distillation from fresh flowers, were considered. GC/MS andGC/FID techniqueswere applied for the phytochemical analysis of the samples. EA/IRMS (Elemental Analyzer/Isotope RatioMass Spectrome-try) and GC/C (Combustion)/IRMS were used to determine the d13C composition of bulk samples and of some specificcomponents.RESULTS:Citronellol (28.7–55.3%), geraniol (13.5–27.3%) and nonadecane (2.6–18.9%)were themain constituents of Bulgarianand Turkish essential oils, while those from Iran were characterized by a high level of aliphatic hydrocarbons (nonade-cane: 3.7–23.2%). The d13C values of bulk samples were between �28.1 and �26.9%, typical for C3 plants. The d13C valuesof specific components were in the usual range for natural aromatic substances from C3 plants, except for geranyl acetate,which displayed higher values (up to�18%). These unusual d13C values were explained by the addition of a natural cheaperoil from a C4 plant (Cymbopogon martinii, palmarosa), which was found to occur in most of the essential oils.CONCLUSION: GC/C/IRMS, in combination with GC/MS and GC/FID, can be considered as an effective and reliabletool for the authenticity control of R. damascena essential oil. Copyright © 2013 John Wiley & Sons, Ltd.

(wileyonlinelibrary.com) DOI: 10.1002/rcm.6489

Rosa damascena Mill. is one of the most important speciesbelonging to the genus Rosa (Rosaceae family).[1] This plant iscommonly called damask rose because it was originallybrought to Europe from Damascus.[1] At present, R. damascenais widely cultivated in many countries, such as Turkey,Bulgaria, Iran, India and Morocco.[2]

The techniques used to extract the essential oil from rose aremainly based on hydro- or steam distillation from fresh flow-ers[2] or on solvent extraction.[3] Due to its low content in thisspecies, rose essential oil is one of the most expensive in theworld market.[1] In this context, the chemical characterizationof this natural product is considered to be very important.

* Correspondence to: F. Pellati, Department of Life Sciences,University of Modena and Reggio Emilia, Via G. Campi183, 41125 Modena, Italy.E-mail: [email protected]

** Correspondence to: F. Camin, Food Quality and NutritionDepartment, IASMA Research and Innovation Centre,Fondazione Edmund Mach, Via E. Mach, 1, 38010 SanMichele all’Adige (TN), Italy.E-mail: [email protected]

Rapid Commun. Mass Spectrom. 2013, 27, 591–602

59

The essential oil of R. damascena is known for its fine perfum-ery application and use in cosmetic preparations,[1] includingperfumes, creams, soaps, and lotions. Several biological activitieshave been attributed to rose essential oil, including antimicro-bial,[4] antioxidant,[5] analgesic,[6] anti-inflammatory[6] and anti-spasmodic.[7] Several effects on the central nervous systemhave also been described,[8] such as a relaxing effect.[9]

The composition of rose essential oil is known to be verycomplex,[10] including monoterpene alcohols, in particularcitronellol, geraniol, nerol, as well as the aromatic phenylethylalcohol, and long-chain hydrocarbons, such as nonadecane,nonacedene, eicosane, and heneicosane.[2] From the sensorypoint of view, various minor constituents, such as rose oxides,significantly contribute to its characteristic aroma.[2]

Adulteration is a well-known and serious problem in essentialoils, especially for themost expensive ones, such as R. damascena.There are several flavoring compounds available that couldbe added to an essential oil to increase its fragrance. Theadulterants can be natural, obtained by physical, enzymatic ormicrobial processes from natural material, such as eugenol(from Syzygium aromaticum), geraniol (from Cymbopogon martinii(palmarosa) or Cymbopogon nardus (citronella)), geranyl acetate

Copyright © 2013 John Wiley & Sons, Ltd.

1

F. Pellati et al.

592

(from Cymbopogon citratus (lemongrass)) and linalool (fromOcimum basilicum (basil)).[11] The flavoring adulterants can alsobe nature-identical, i.e. chemically identical to those present innature but obtained by chemical synthesis or isolated bychemical processes.[11]

In the case of R. damascena, the most frequently employedadulterants are natural or nature-identical, including citronel-lol, geraniol and geranyl acetate.[12] Other common adulter-ants are much cheaper essential oils rich in geraniol, suchas C. martinii essential oil, mainly produced in India.[13,14]

Adulterants are usually added at low percentages (5–8%) toavoid detection by commonly used analytical methods.[15]

Gas chromatography/mass spectrometry (GC/MS) is oneof the most reliable techniques for the determination of thecomposition of essential oils. Through GC/MS it is possibleto study the effects of different extraction methods andgeographic origin on the essential oil composition. Inaddition to conventional GC techniques, reliable analyticalmethods based on isotope ratio mass spectrometry (IRMS)and on enantioselective analysis for origin assessment andquality assurance of essential oils are of fundamental inter-est.[16] In particular, the determination of the d13C values oftarget components of natural products, by means of GC/C/IRMS, has become a rapid and convenient tool for authenti-city assessment,[17] allowing both the determination of geo-graphic origin and the evaluation of possible adulteration ofessential oils with synthetic or natural compounds. Analysisof the essential oils of orange,[18] lemon, lemon grass, lemonbalm, lemon gum, citronella,[19–21] coriander,[22] man-darin,[23,24] dill,[25] anise, fennel,[26] tarragon, basil, Mexicanmarigold, West Indian Bay tree fruit,[27] cinnamon (cassia),[28]

oregano, savory, thyme, fennel oil,[29] lavender,[30] neroli[31]

and bergamot[32] by GC/C/IRMS to determine authenticityhas been well documented. In the case of R. damascena,isotopic data of the essential oil have not been reported inthe peer-reviewed scientific literature, with the exception ofa technical paper concerning the d13C values of citronellol,nerol, geraniol and phenylethyl alcohol of two samples.[33]

Due to the high commercial value of rose extracts inperfumery and cosmetics, this study is aimed at a detailedphytochemical characterization of commercial samples of R.damascena essential oil of different geographic origin bymeans of GC/MS and GC/FID techniques and at theirauthenticity assessment using elemental analysis EA/IRMSand GC/C/IRMS. To the best of our knowledge, this is thefirst report of a comprehensive multi-component analysis ofrose essential oil from different countries by the above-citedtechniques. In particular, the d13C values of bulk R. damascenaessential oils and of some specific components are describedhere for the first time.

EXPERIMENTAL

Rose essential oil samples

Nineteen commercial samples of R. damascena essential oilwere kindly gifted by the manufacturers or purchased inItalian pharmacies and herbal shops in Winter 2009–Spring2010. Based on the label claims, these samples were groupedin agreement with their place of origin: eight samples werefrom Bulgaria (indicated in the text as BULG 01–08), eight fromTurkey (TURK 01–08) and three from Iran (IRAN 01–03). An

wileyonlinelibrary.com/journal/rcm Copyright © 2013 John Wi

additional authentic R. damascena essential oil, directlyextracted by hydro-distillation from fresh flowers in thelaboratory of Agronatura (Spigno Monferrato, Alessandria,Italy), was analyzed.

On the basis of what was reported in the labels, theBulgarian samples had been extracted by hydro-distillation,with the exception of samples BULG 01 and 04, whose extrac-tion method was not specified. Turkish samples labelled asTURK 01–03 and 05 were extracted by hydro-distillation; forsample TURK 08 the extraction procedure was not specified.The Iranian samples, indicated as IRAN 01–03, wereproduced by hydro-distillation: in particular, sample IRAN 02is a first rose oil, i.e. it was obtained by direct hydro-distillationof fresh rose petals, while sample IRAN 03 is a second rose oil,i.e. obtained by re-distillation of rose water.

Two additional samples of C. martinii (Roxb.) Wats. var.motia Burk. (Poaceae family) (palmarosa) essential oil werepurchased in local pharmacies in Fall 2010. These samplesoriginated from Nepal and India.

All samples were stored at low temperature (+4 �C), pro-tected from light and humidity, until required for chemicalanalysis.

Chemicals and solvents

All reference standards used for GC analysis were of chroma-tographic grade and were purchased from Sigma-Aldrich(Milan, Italy), Extrasynthese (Genay, France) and Roth(Karlsruhe, Germany). Chromatographic grade organic solventswere from Sigma-Aldrich and VWR (Milan, Italy).

Sample preparation for GC/MS and GC/FID analysis

The essential oils were diluted 1:2 (v/v) with n-hexanebefore GC/MS analysis. In the case of GC/FID, the sampleswere directly analyzed, without dilution. Three injectionswere performed for each sample.

GC/MS conditions

Analyses were performed on a 6890N gas chromatograph(Agilent Technologies, Waldbronn, Germany), coupled witha 5973 Network mass spectrometer (Agilent Technologies).Compounds were separated on a HP-5 MS cross-linkedpoly-5% diphenyl–95% dimethyl polysiloxane (30 m� 0.25mm i.d., 1.00 mm film thickness) capillary column (AgilentTechnologies). The column was maintained at 60 �C for6 min after the injection, then programmed at 3 �C/min to230 �C, at which temperature it was maintained for 7 min.The injection volume was 0.1 mL, with a split ratio 1:100.Helium was used as the carrier gas at a flow rate of0.7 mL/min. The injector, transfer line and ion-source tem-peratures were set at 250, 280 and 230 �C, respectively. MSdetection was performed with electron ionization (EI) at70 eV, operating in the full-scan acquisition mode in the m/zrange 30–350.

GC/FID conditions

Analyses were carried out on a model 8610 gas chromatograph(DANI Instruments, Milan, Italy) with flame ionization detec-tion (FID). Compounds were separated on a HP-5 cross-linkedpoly-5% diphenyl–95% dimethyl polysiloxane (25 m� 0.2 mm

ley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2013, 27, 591–602

GC/MS, GC/FID and GC/C/IRMS analysis of Rosa damascena essential oil

59

i.d., 0.5 mm film thickness) capillary column (Agilent Technolo-gies). The temperature program and the chromatographicconditions were the same as described above. The injectionvolume was 0.2 mL, with a split ratio 1:20. Helium was usedas the carrier gas at a pressure of 1.5 bar at the column head.The injector and detector temperatures were set at 250 �C.A mixture of aliphatic hydrocarbons (C8–C25) in n-hexane

(Sigma) was injected under the above temperature programto calculate the linear retention index (LRI) of each compound.

Qualitative and semi-quantitative analysis

Compounds were identified by comparing the retentiontimes of the chromatographic peaks with those of authenticreference compounds run under the same conditions and bycomparing the LRIs with published data. Peak enrichmenton co-injection with authentic reference compounds wasalso carried out. Comparison of the MS fragmentationpattern of the target analytes with those of pure compoundswas also performed. A mass spectrum database search wasperformed using the National Institute of Standards andTechnology (NIST, Gaithersburg, MD, USA) mass spectraldatabase (version 2.0d, 2005).The percentage relative amount of individual components

was expressed as the percent peak area relative to the totalpeak area obtained by the GC/FID analysis. Semi-quantitativedata were acquired from the mean of three analyses.

Sample preparation for EA/IRMS and GC/C/IRMS analysis

Aliquots of 0.3 mg of bulk essential oils and pure commercialcompounds (used as reference standards) were weighed intin capsules (Säntis Analytical AG, Teufen, Switzerland) forEA/IRMS analysis. In the case of GC/C/IRMS analysis,essential oils were diluted 1:3 (v/v) with n-hexane beforeinjection. Each sample was measured at least twice inEA/IRMS and three times in GC/C/IRMS.

EA/IRMS conditions

The analyses were carried out using an elemental analyzer(Flash EA 1112, Thermo Scientific, Bremen, Germany),equipped with an autosampler (Finnigan AS 200, ThermoScientific) and interfaced through a ConFlo III dilutordevice (ThermoFinnigan, Bremen, Germany) with a DELTAV isotope ratio mass spectrometer Thermo Scientific).

GC/C/IRMS conditions

A model 6890A gas chromatograph (Agilent Technologies)equipped with an autosampler (GC-PAL, GC Analytics AG,Zwingen, Switzerland) and a DB-WAX capillary column (30 m0.32 mm i.d., 0.25 mm film thickness; Agilent Technologies)wasused. The injector temperaturewas set at 250 �Cand the tem-perature programwas as follows: the initial column temperaturewas 50 �C, held for 4 min; then increased to 160 �C at a rate of5 �C/min, to 180 �C at 2 �C/min, to 225.5 �C at 5 �C/min andto 250 �C at 10 �C/min, where it was held for 5 min. Heliumwas used as the carrier gas at a flow rate of 2 mL/min.The gas chromatograph was supplied with an oxidation

reactor composed of a 32 cm long alumina tube in whichthree thin (0.125 mm diameter) braided wires were placed:one nickel oxide, one copper oxide and one platinum. The

Copyright © 2013 JoRapid Commun. Mass Spectrom. 2013, 27, 591–602

tube was kept at a temperature of 940 �C. Water waseliminated by a water-removing trap, consisting of a NafionW

membrane. The GC/C system was interfaced with an isotoperatio mass spectrometer (DELTA V) through an open splitinterface.

13C/12C analysis of bulk rose essential oils and purecommercial compounds

The samples were quantitatively burnt to carbon dioxide andwater (the second one was removed using a magnesiumperchlorate filter) in the presence of oxygen and copper oxidein an elemental analyzer. The carbon dioxide was flushed intothe isotope ratio mass spectrometer, where the content of theisotopomers at m/z 44 (12C16O2) and 45 (13C16O2) wasdetermined.[34]

The isotopic values were calculated by comparing the ratio ofthe sample with that of a working standard (oil) calibratedagainst international reference materials: fuel oil NBS-22(International Atomic Energy Agency (IAEA), Vienna, Austria)and sugar IAEA-CH-6 (IAEA). The isotopic ratio values of theaforementioned reference materials and, therefore, also ofthe samples were expressed in d% versus the internationalstandard V-PDB (Vienna-Pee Dee Belemnite), according to thefollowing equation:

13C12C

� �sample� 13C

12C

� �std

13C12C

� �std

� 1000

The uncertainty of measurement, calculated as 2� standarddeviation (SD) of the intra-laboratory reproducibility, was0.3%.

13C/12C analysis of specific compounds in rose essential oils

To analyze the main constituents present in R. damascenaessential oil (citronellol, nerol, geraniol), 1 mL of samplesolution was injected using a 10 mL Hamilton syringe in thesplit mode (1:75). In the essential oils, the nonadecane peakwas not completely resolved and, consequently, it was notconsidered for the analysis, with the exception of the Iraniansamples, in which this compound was determined. After thechromatographic separation, each compound was quantita-tively burnt to carbon dioxide and water in the oxidationreactor. The solvent was vented with a back-flush valveduring the first 200 s of analysis to avoid damage andinactivation of the oxidation column. Carbon dioxide, carriedby a helium stream, was then introduced into the ion sourceof the isotope ratio mass spectrometer for the measurementof d13C values.

The lower concentration compounds were analyzed usinga 1:25 split ratio with 1 mL of solution injected and ventingthe main compounds with a back-flush valve. Under theseconditions, satisfactory results were obtained for geranylacetate, heneicosane and farnesol. The other compounds werenot completely separated or their concentration was too lowto be analyzed (<0.5% by GC/MS or GC/FID analysis).

The peak of each constituent was identified by previouslyanalyzing pure commercial reference compounds under thesame experimental conditions. To calculate the d13C values,a reference mixture composed of pure compounds was used,

wileyonlinelibrary.com/journal/rcmhn Wiley & Sons, Ltd.

3

F. Pellati et al.

594

for which the d13C values had previously been measuredusing EA/IRMS. The analytical procedure followed a specificsequence: re-oxidation of the oxidation reactor through a flowof oxygen; injection of n-hexane (used in reference mix) tocheck for purity; four injections of the reference mixture (thefirst one was not considered); two samples (three injectionsof each); and three injections of the reference mixture.The instrumental data for each constituent were corrected

on the basis of the difference existing between the d13C valueof the pure compound in GC/C/IRMS (mean of the sixresults, three before and three after the samples) and that inEA/IRMS.The SD obtained from three repeated analyses of one sam-

ple was ≤0.2% for citronellol, ≤0.4% for nerol and geraniol,≤0.5% for geranyl acetate and≤1% for the other compounds.

Statistical analysis

Analysis of variance (ANOVA) was used to evaluate thestatistical significance of the measured differences betweenessential oils of different origin. To evaluate the most impor-tant variables that discriminate between origins, a post hoctest was performed using the Tukey ’Honestly SignificantDifference’ (HSD) test. For all these tests, the P level was setat 0.05. The statistical analysis was performed using Statistica6 for Windows (StatSoftW Italia, Vigonza, Italy).

RESULTS AND DISCUSSION

Chemical composition of rose essential oils

Figure 1 shows a representative GC/FID chromatogramobtained by the analysis of a R. damascena essential oil fromBulgaria (BULG 01). A total of 39 compounds were identifiedin this sample, representing 98.9% of the overall essential oilcomposition.Table 1 shows the chemical composition of the Bulgarian rose

essential oils analyzed, all of them having been extracted byhydro-distillation. All these samples displayed a commonphytochemical profile, based on the presence of citronellol(28.7–53.1%), geraniol (13.5–25.0%) and nonadecane (2.6–18.9%).

Figure 1. GC/FID chromatogram obtained by the an(BULG 01). For peak identification, see Table 1.


Other constituents were found to be nerol (1.7–3.3%), nonade-cene (0.8–4.8%) and phenylethyl alcohol (0.3–3.4%). cis- andtrans-Rose oxides were found at low levels (0.1–0.5%). Thesesubstances are very important for the characteristic rose scentand exhibit extremely low threshold values. The describedcomposition was found to be in good agreement with pub-lished results where Gochev et al.[35] analyzed the chemicalcomposition of a historical rose oil from Bulgaria and deter-mined citronellol (23.4%), geraniol (19.0%), nonadecane(11.9%) and nerol (7.5%) as the main constituents.

Table 2 shows the composition of Turkish rose essential oilsextracted by hydro-distillation. The rose essential oils describedin Table 2 have a chemical composition closely related to thoseof the above described samples of Bulgarian origin: the mainconstituents were found to be citronellol (41.7–55.3%), geraniol(15.8–27.3%) and nonadecane (3.8–7.1%). These results are alsoin good agreement with those of hydro-distilled essential oilsfrom Turkey previously described: Bayrak and Akgül[36]

reported the presence of citronellol in the range 24.5–42.9%,2.1–18.0% for geraniol and 6.4–19.0% for nonadecane. In theessential oils analyzed by Chalchat and Özcan[37] the levelsof citronellol, geraniol and nonadecane were 43.0–48.0%, 12.2–19.9% and 9.8–11.0%, respectively.

Table 3 shows the chemical composition of Iranian roseessential oils which were extracted by hydro-distillation.The samples labeled as IRAN 01 and 02 showed a chemicalcomposition based on high levels of aliphatic hydrocarbons,such as nonadecane (16.1–23.2%), nonadecene (5.3–6.9%),heneicosane (4.3–6.9%) and heptadecane (3.7–4.2%), andquite low relative percentages of monoterpene alcohols, suchas citronellol (25.2–25.5%) and geraniol (8.3–9.6%). Thiscomposition was found to be in agreement with the resultsobtained by the analysis of a historical rose oil sample fromIran,[2] which contained citronellol (25.1%), nonadecane(13.4%), geraniol (11.8%), nonadecene (6.9%), heneicosane(6.2%) and heptadecane (3.9%) as the main constituents.Other workers have reported similar compositions for Iranianessential oils.[38,39] Sample IRAN 03 displayed a chromato-graphic profile very similar to those oils of Bulgarian andTurkish origin, having citronellol (45.9%) and geraniol(28.5%) as the main constituents and a low percentage of non-adecane (3.7%). This sample has a chemical composition in

alysis of a R. damascena essential oil from Bulgaria


Table

1.Volatile

arom

acompo

nentsiden

tified

inBulga

rian

R.d

amascena

essentialo

ilsby

GC

analysisa

Peak

numbe

rCom

poun

db

LRI

BULG

01c

BULG

02c

BULG

03c

BULG

04c

BULG

05c

BULG

06c

BULG

07c

BULG

08c

Iden

tification

metho

dd

1Hexan

ol866

0.3�0.1

0.2�0.1

0.3e

0.2e

0.4�0.1

-0.1e

0.1e

a,b,c,d

2Hep

tana

l901

0.1e

tr0.1e

0.1e

0.1e

-0.1e

0.3�0.1

b,d

3a-Pine

ne935

1.1�0.2

0.6e

0.5e

0.7�0.1

0.6e

-0.6�0.1

0.3e

a,b,c,d

4Ben

zaldeh

yde

963

0.1e

-tr

tr-

--

-b,d

5Sa

bine

ne975

0.1e

trtr

tr0.1e

-tr

-a,b,c,d

6b-Pine

ne979

0.3�0.1

0.1e

0.1e

0.2e

0.2e

-0.1e

0.1e

a,b,c,d

7Myrcene

990

0.5�0.1

0.2�0.1

0.3�0.1

0.4e

0.3e

-0.3�0.1

0.2�0.1

a,b,c,d

8Lim

onen

e1030

0.1e

tr0.1e

0.1e

--

0.1e

-a,b,c,d

9g-Terpinen

e1060

tr-

trtr

--

tr-

a,b,c,d

10Terpinolen

e1091

0.1e

trtr

tr-

-tr

tra,b,c,d

11Linaloo

l1102

1.8�0.2

1.8�0.1

2.9e

1.6�0.1

1.2�0.1

0.7�0.1

2.3�0.1

0.7�0.1

a,b,c,d

12cis-Roseox

ide

1113

0.5�0.1

0.3e

0.5e

0.4e

0.5e

0.2e

0.4e

0.2e

b,d

13Ph

enylethy

lalcoh

ol1120

1.1�0.1

1.3�0.1

1.0e

2.5�0.2

0.3e

1.2�0.1

3.4�0.1

0.4�0.1

a,b,c,d

14trans-Roseox

ide

1130

0.2e

0.2�0.1

0.2e

0.2e

0.2e

0.1e

0.2e

0.1e

b,d

15Terpinen

-4-ol

1184

0.4�0.1

0.5�0.1

0.5e

0.4e

0.4e

0.2�0.1

0.3e

0.2e

a,b,c,d

16a-Terpineo

l1198

0.4e

0.5�0.1

0.6e

0.4e

0.3e

0.2�0.1

0.5e

0.1e

a,b,c,d

17Citrone

llol

1238

49.5�0.9

45.8�1.0

53.1�0.3

47.1�0.1

51.9�0.9

37.4�0.4

44.1�1.1

28.7�0.1

a,b,c,d

18Nerol

1239

1.7�0.3

2.8�0.2

1.9�0.1

2.1�0.5

2.8�0.2

3.3�0.7

3.1�0.3

2.2�0.1

a,b,c,d

19Neral

1247

0.8�0.5

0.8e

0.2e

0.8�0.1

1.0�0.1

0.9�0.1

0.8�0.1

0.7�0.1

b,d

20Geran

iol

1263

20.1�0.7

25.0�0.5

23.3�0.1

23.6�0.2

22.3�0.5

20.5�0.5

24.0�0.4

13.5�0.1

a,b,c,d

21Geran

ial

1276

1.5�0.1

1.5�0.1

1.3�0.1

1.4e

1.3�0.1

1.2�0.1

1.0e

0.8e

a,b,c,d

22Methy

lgeran

ate

1323

0.1e

tr0.1e

0.1e

0.1e

-tr

trb,d

23Citrone

llyla

cetate

1353

0.5e

0.4e

0.5e

0.6�0.1

0.6�0.1

0.6�0.1

0.4e

0.5�0.1

a,b,c,d

24Eug

enol

1367

1.2�0.1

1.3�0.1

1.1�0.1

1.2e

0.5e

1.6�0.1

1.4�0.1

0.6�0.1

a,b,c,d

25Geran

ylacetate

1384

1.0e

0.8�0.1

0.7e

1.1�0.1

0.9e

1.5�0.1

1.3e

0.8e

a,b,c,d

26Methy

leug

enol

1408

1.6�0.1

1.7e

1.4�0.1

1.7�0.1

1.9�0.1

1.6�0.2

1.4�0.1

1.2�0.1

a,b,c,d

27b-Caryo

phyllene

1435

0.6�0.1

0.5e

0.6e

0.6e

0.6e

0.4�0.1

0.4e

0.8e

a,b,c,d

28a-Gua

iene

1447

0.4e

0.3e

0.3e

0.4e

0.4e

0.3�0.1

0.2e

0.7�0.1

b,d

29a-Hum

ulen

e1467

0.3e

0.3e

0.2e

0.3e

0.2�0.1

0.3�0.2

0.2e

0.5e

a,b,c,d

(Contin

ues)


wileyonlinelibrary.com/journal/rcmCopyright © 2013 John Wiley & Sons, Ltd.Rapid Commun. Mass Spectrom. 2013, 27, 591–602

595

Table

1.(C

ontinu

ed)

Peak

numbe

rCom

poun

db

LRI

BULG

01c

BULG

02c

BULG

03c

BULG

04c

BULG

05c

BULG

06c

BULG

07c

BULG

08c

Iden

tification

metho

dd

30Germacrene

D1496

1.1�0.1

0.8e

0.7�0.1

0.9�0.1

0.9e

1.0e

0.6e

1.2�0.1

b,d

31Pe

ntad

ecan

e1511

0.2e

0.1e

0.2e

0.2e

0.1e

0.1e

0.1e

0.3e

b,d

32d-Gua

iene

1520

0.3e

0.3e

0.2�0.1

0.3e

0.3e

0.3e

0.2e

0.5e

b,d

33Hep

tadecan

e1696

1.1�0.2

1.0�0.1

0.6e

1.0�0.1

0.8�0.1

1.6�0.1

1.1�0.1

2.1�0.1

a,b,c,d

34Fa

rnesol

1728

1.0�0.2

1.4�0.1

0.9e

1.2�0.1

1.0�0.1

1.2�0.2

0.7�0.1

2.2�0.1

a,b,c,d

35Octad

ecan

e1796

0.1e

0.1e

0.1e

0.1e

0.1e

0.2e

0.1e

-b,d

36Non

adecen

e1874

1.6�0.4

1.5�0.3

0.8�0.1

1.3�0.1

1.3�0.3

3.3�0.4

1.3�0.1

4.8�0.1

b,d

37Non

adecan

e1898

5.3�1.0

5.0�0.8

2.6�0.1

4.0�0.1

4.0�1.0

10.6�0.8

5.8�0.6

18.9�0.1

a,b,c,d

38Eicosan

e1995

0.4�0.2

0.3�0.1

0.2e

0.2�0.1

0.3�0.1

0.9�0.1

0.3e

1.9e

b,d

39Hen

eicosane

2110

1.5�0.6

1.4�0.3

0.7�0.2

0.8�0.2

1.3�0.5

3.9�0.4

1.0�0.1

9.3�0.1

a,b,c,d

Total

98.9�0.3

99.0�0.2

99.1�0.2

98.2�0.6

99.0�0.6

95.4�1.9

98.1�0.4

94.9�0.2

a GC

cond

itions

asin

Exp

erim

entalsection

.bCom

poun

dsarelistedin

order

ofelutiontime.

c Dataareexpressedas

mean(n

=3)

of%

relative

peak

area

values

�SD

.da:

retentiontime;

b:LR

I;c:pe

aken

rich

men

t;d:m

asssp

ectrum

.e SD

<0.05.

Table

2.Volatile

arom

acompo

nentsiden

tified

inTu

rkishR.d

amascena

essentialo

ilsby

GC

analysisa

Peak

numbe

rCom

poun

db

LRI

TURK01

cTURK02

cTURK03

cTURK04

cTURK05

cTURK06

cTURK07

cTURK08

cIden

tification

metho

dd

1Hexan

ol866

0.4e

0.2e

0.3e

0.3�0.1

0.3e

0.3e

0.4�0.1

-a,b,c,d

2Hep

tana

l901

tr-

0.1e

0.2e

0.2�0.1

0.2e

tr-

b,d

3a-Pine

ne935

0.5e

2.0e

1.3e

1.4�0.2

1.8�0.1

1.0e

0.6�0.1

-a,b,c,d

5Ben

zaldeh

yde

963

tr-

--

--

0.1e

-b,d

4Sa

bine

ne975

tr0.2e

0.1e

0.2�0.1

0.2e

0.1e

tr-

a,b,c,d

6b-Pine

ne979

0.1e

0.4e

0.3e

0.4�0.1

0.4e

0.2e

0.2�0.1

-a,b,c,d

7Myrcene

990

0.2�0.1

1.0e

0.6e

0.7�0.1

0.8�0.1

0.5e

0.2e

-a,b,c,d

8Lim

onen

e1030

tr0.1e

0.1e

0.1e

0.1e

0.1e

tr-

a,b,c,d

9g-Terpinen

e1060

-0.1e

tr0.1e

trtr

--

a,b,c,d

10Terpinolen

e1091

-tr

tr-

trtr

tr-

a,b,c,d

11Linaloo

l1102

1.6e

2.8�0.1

0.7e

1.1�0.1

0.9�0.1

0.9�0.1

1.5�0.2

0.6�0.2

a,b,c,d

12cis-Roseox

ide

1113

0.5e

0.3e

0.4e

0.4e

0.3e

0.3e

0.5e

-b,d

13Ph

enylethy

lalcoh

ol1120

2.4�0.1

1.7e

1.9e

2.4�0.2

1.6�0.1

2.2�0.1

2.1�0.2

2.0�0.4

a,b,c,d

14trans-Roseox

ide

1130

0.3e

0.2�0.1

0.2e

0.2e

0.2e

0.2e

0.3�0.1

-b,d

15Terpinen

-4-ol

1184

0.5e

0.6e

0.5e

0.6e

0.5e

0.5e

0.4�0.1

0.2�0.1

a,b,c,d

(Contin

ues)

F. Pellati et al.

wileyonlinelibrary.com/journal/rcm Copyright © 2013 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2013, 27, 591–602

596

Table

2.(C

ontinu

ed)

Peak

numbe

rCom

poun

db

LRI

TURK01

cTURK02

cTURK03

cTURK04

cTURK05

cTURK06

cTURK07

cTURK08

cIden

tification

metho

dd

16a-Terpineo

l1198

0.3e

0.7�0.1

0.2e

0.2e

0.2e

0.2e

0.3e

0.2e

a,b,c,d

17Citrone

llol

1238

55.3�0.8

42.7�0.9

50.1e

48.3�1.5

41.7�1.0

47.4�1.2

52.1�1.0

49.2�1.2

a,b,c,d

18Nerol

1239

1.9�0.2

2.8�0.6

1.4e

2.6�0.5

2.6�0.2

2.5�0.4

1.7�0.5

2.4�0.1

a,b,c,d

19Neral

1247

0.6�0.1

0.5�0.1

0.1e

0.6�0.1

0.5e

0.7�0.1

0.6�0.1

0.3�0.1

b,d

20Geran

iol

1263

17.0�0.4

27.3�0.4

19.0e

19.8�1.2

23.6�0.8

20.5�0.9

15.8�1.3

22.0�0.9

a,b,c,d

21Geran

ial

1276

1.0�0.3

0.8e

0.9�0.1

0.9e

0.8�0.1

0.9e

1.2�0.2

0.8e

a,b,c,d

22Methy

lgeran

ate

1323

0.1e

0.1e

0.1e

-tr

tr0.1e

-b,d

23Citrone

llyla

cetate

1353

0.8e

0.6e

0.8e

0.8�0.1

0.7e

0.8e

0.8�0.1

0.5�0.1

a,b,c,d

24Eug

enol

1367

0.4e

1.0e

1.7e

1.3�0.1

0.8e

1.4�0.1

0.4�0.1

0.4�0.2

a,b,c,d

25Geran

ylacetate

1384

1.3e

1.6e

1.5e

2.1�0.1

1.8�0.1

2.1�0.1

1.3�0.1

0.9�0.1

a,b,c,d

26Methy

leug

enol

1408

2.7�0.1

1.6�0.1

2.6e

2.4�0.1

1.7�0.1

2.4�0.1

2.7�0.1

1.8�0.2

a,b,c,d

27b-Caryo

phyllene

1435

0.4e

0.5e

0.5e

0.5e

0.6e

0.5e

0.4e

0.2e

a,b,c,d

28a-Gua

iene

1447

0.3e

0.3e

0.4e

0.5�0.1

0.5e

0.5e

0.3e

0.2�0.1

b,d

29a-Hum

ulen

e1467

0.2�0.1

0.3e

0.3e

0.3e

0.4e

0.3e

0.2e

0.1e

a,b,c,d

30Germacrene

D1496

0.6e

0.6e

0.8e

1.1�0.2

1.1�0.1

1.0e

0.7�0.1

0.4�0.1

b,d

31Pe

ntad

ecan

e1511

0.1e

0.1e

0.2�0.1

0.2e

0.2�0.1

0.2e

0.1e

trb,d

32d-Gua

iene

1520

0.1e

0.2e

0.3�0.1

0.3�0.1

0.3�0.1

0.3�0.1

0.2e

-b,d

33Hep

tadecan

e1696

1.0�0.1

1.0�0.1

1.0e

0.7�0.2

1.0�0.2

0.8�0.2

1.3�0.3

0.8�0.3

a,b,c,d

34Fa

rnesol

1728

0.4�0.1

0.7e

0.6e

0.5�0.2

0.9�0.1

0.6�0.1

0.5�0.1

1.1�0.3

a,b,c,d

35Octad

ecan

e1796

0.1e

0.1e

0.1e

0.1e

0.1e

0.1e

0.1e

0.1e

b,d

36Non

adecen

e1874

1.5�0.2

1.1�0.2

1.6e

1.2�0.6

1.6�0.4

1.5�0.6

2.0�0.6

2.3�0.2

b,d

37Non

adecan

e1898

4.3�0.3

3.8�0.5

5.4�0.1

4.2�1.0

6.6�1.0

4.1�1.2

7.1�0.7

7.1�0.4

a,b,c,d

38Eicosan

e1995

0.3e

0.3�0.1

0.5�0.1

0.3�0.2

0.6�0.1

0.4�0.2

0.5�0.2

0.5�0.2

b,d

39Hen

eicosane

2110

1.0e

0.8�0.1

1.8e

1.4�0.8

2.6�0.4

1.9�1.0

1.8�0.8

2.6�0.7

a,b,c,d

Total

98.2�0.6

99.1e

98.3�0.1

98.6�0.9

98.2�0.4

97.8�0.5

98.4�0.2

96.6�0.1

a GC

cond

itions

asin

Exp

erim

entalsection

.bCom

poun

dsarelistedin

order

ofelutiontime.

c Dataareexpressedas

mean(n

=3)

of%

relative

peak

area

values

�SD

.da:

retentiontime;

b:LR

I;c:pe

aken

rich

men

t;d:m

asssp

ectrum

.e SD

<0.05.


wileyonlinelibrary.com/journal/rcmCopyright © 2013 John Wiley & Sons, Ltd.Rapid Commun. Mass Spectrom. 2013, 27, 591–602

597

Table 3. Volatile aroma components identified in Iranian R. damascena essential oils by GC analysisa

Peak number Compoundb LRI IRAN 01c IRAN 02c IRAN 03cIdentificationmethodd

1 Hexanol 866 - - 0.1e a,b,c,d2 a-Pinene 935 2.0� 0.2 5.9� 0.3 0.5e a,b,c,d3 Sabinene 975 0.2� 0.1 0.5� 0.1 tr a,b,c,d4 b-Pinene 979 0.5� 0.2 1.1� 0.2 0.1e a,b,c,d5 Myrcene 990 0.8� 0.1 2.3� 0.2 0.3e a,b,c,d6 Limonene 1030 0.1e 0.3� 0.1 0.1e a,b,c,d7 g-Terpinene 1060 0.1e 0.3� 0.1 tr a,b,c,d8 Terpinolene 1091 tr 0.1e tr a,b,c,d9 Linalool 1102 1.8e 0.5� 0.1 2.6e a,b,c,d10 cis-Rose oxide 1113 0.3� 0.2 0.2� 0.1 0.2e b,d11 Phenylethyl

alcohol1120 1.0� 0.3 0.5� 0.2 1.6� 0.1 a,b,c,d

12 trans-Rose oxide 1130 0.1e 0.1e 0.1e b,d13 Terpinen-4-ol 1184 0.2e 0.1e 0.5e a,b,c,d14 a-Terpineol 1198 0.4e 0.1e 0.6e a,b,c,d15 Citronellol 1238 25.5� 0.6 25.2� 0.2 45.9� 0.5 a,b,c,d16 Nerol 1239 0.9� 0.2 0.9� 0.1 2.4� 0.3 a,b,c,d17 Neral 1247 0.1e 0.1e 0.5e b,d18 Geraniol 1263 8.3� 0.3 9.6� 0.2 28.5� 0.1 a,b,c,d19 Geranial 1276 0.2e 0.2� 0.1 0.6� 0.1 a,b,c,d20 Methyl geranate 1323 tr 0.3e - b,d21 Citronellyl acetate 1353 0.8e 1.3� 0.2 0.5e a,b,c,d22 Eugenol 1367 0.6e 0.5� 0.1 1.2� 0.1 a,b,c,d23 Geranyl acetate 1384 1.4� 0.1 3.3� 0.3 1.7e a,b,c,d24 Methyl eugenol 1408 0.8e 0.4� 0.1 1.3� 0.1 a,b,c,d25 b-Caryophyllene 1435 0.7e 1.1� 0.1 0.3e a,b,c,d26 a-Guaiene 1447 0.8e 1.4� 0.2 0.4e b,d27 a-Humulene 1467 0.6� 0.1 1.0� 0.1 0.3e a,b,c,d28 Germacrene D 1496 2.8� 0.1 4.2� 0.2 1.3e b,d29 Pentadecane 1511 0.4e 0.7� 0.1 0.2e b,d30 d-Guaiene 1520 0.6e 1.0� 0.1 0.3e b,d31 Heptadecane 1696 4.2� 0.1 3.7� 0.1 1.0� 0.1 a,b,c,d32 Farnesol 1728 0.8e 1.1� 0.2 0.4� 0.1 a,b,c,d33 Octadecane 1796 0.4e 0.3� 0.1 0.1e b,d34 Nonadecene 1874 6.9� 0.1 5.3� 0.9 1.1e b,d35 Nonadecane 1898 23.2� 0.4 16.1� 0.4 3.7� 0.1 a,b,c,d36 Eicosane 1995 1.7� 0.1 1.3� 0.5 0.2e b,d37 Heneicosane 2110 6.9� 0.1 4.3� 0.4 0.6� 0.1 a,b,c,d

Total 96.1�1.5 95.1� 0.2 99.3e

aGC conditions as in Experimental section.bCompounds are listed in order of elution time.cData are expressed as mean (n=3) of % relative peak area values� SD.da: retention time; b: LRI; c: peak enrichment; d: mass spectrum.eSD <0.05.

F. Pellati et al.

598
agreement with that of a so-called Iranian ’second rose oil’,[39]
which was produced by re-distillation of rose water, andcontained 53.4% of citronellol, 22.7% of geraniol and 3.4% ofnonadecane.A statistical analysis was performed on the GC data of

the Bulgarian, Turkish and Iranian essential oils and theresults are shown in Table 4. Considering the data reportedin Tables 1–3, some minor compounds, such as benzaldehyde,g-terpinene, terpinolene, heptanal and methyl geranate,were excluded from the statistical analysis. It is evidentfrom the data of Table 4 that all the volatile compoundspresent a very high variability. Significant differences wereobserved for 20 of the 35 considered compounds andthese differences are more frequent for compounds with


high retention times. The HSD test did not allow identifica-tion of any particular trend in the samples related to thegeographic origin of the essential oils.

d13C composition of rose essential oils

The d13C values of bulk samples of R. damascena essentialoil ranged between �28.1 and �26.9%, with a meanvalue of �27.5% and a SD of 0.4%. No significant differences(P <0.05) are evident between geographic origins. Thesevalues are typical for C3 plants, such as R. damascena,that assimilate atmospheric carbon dioxide according to theso-called ’C3 pathway’ described by Calvin and are character-ized by d13C values between �34 and �22%.[40] Plants


Table 4. Statistical analysis of Bulgarian, Turkish and Iranian essential oils

Compound Overall mean Bulgarian Turkish Iranian ANOVA Pa HSD testb

Hexanol 0.20� 0.14 0.20� 0.12 0.28� 0.12 0.03� 0.06 <0.05 AB A Ba-Pinene 1.13� 1.31 0.56� 0.32 1.07� 0.68 2.81� 2.80 <0.05 A AB BSabinene 0.08� 0.12 0.02� 0.05 0.10� 0.09 0.21� 0.24 NS A A Ab-Pinene 0.25� 0.24 0.14� 0.08 0.24� 0.15 0.57� 0.48 <0.05 A AB BMyrcene 0.51� 0.50 0.28� 0.14 0.50� 0.35 1.11�1.03 <0.05 A A ALimonene 0.07� 0.07 0.05� 0.05 0.06� 0.05 0.16� 0.10 NS A A ALinalool 1.48� 0.76 1.62� 0.75 1.28� 0.72 1.64� 1.04 NS A A Acis-Rose oxide 0.34� 0.14 0.37� 0.12 0.34� 0.16 0.23� 0.09 NS A A APhenylethyl alcohol 1.61� 0.81 1.41�1.05 2.03� 0.30 1.01� 0.53 NS A A Atrans-Rose oxide 0.17� 0.07 0.17� 0.04 0.19� 0.09 0.10c NS A A ATerpinen-4-ol 0.40� 0.14 0.35� 0.11 0.48� 0.16 0.27� 0.21 <0.05 A A Aa-Terpineol 0.34� 0.18 0.37� 0.17 0.29� 0.18 0.37� 0.25 NS A A ACitronellol 44.2� 9.00 44.7� 8.11 48.3� 4.52 32.2� 11.8 <0.05 AB A BNerol 2.20� 0.68 2.48� 0.59 2.23� 0.49 1.41� 0.88 NS A A ANeral 0.57� 0.28 0.77� 0.25 0.50� 0.17 0.24� 0.22 <0.05 A AB BGeraniol 20.2� 5.42 21.5� 3.66 20.6� 3.70 15.5� 11.3 NS A A AGeranial 0.96� 0.37 1.24� 0.25 0.91� 0.14 0.36� 0.24 <0.05 A B CMethyl geranate 0.06� 0.08 0.05� 0.05 0.05� 0.05 0.10� 0.17 NS A A ACitronellyl acetate 0.66� 0.22 0.51� 0.08 0.73� 0.13 0.87� 0.40 <0.05 A B BEugenol 0.99� 0.44 1.13� 0.38 0.92� 0.51 0.78� 0.40 NS A A AGeranyl acetate 1.43� 0.62 1.01� 0.27 1.59� 0.42 2.12� 1.03 <0.05 A AB BMethyl eugenol 1.73� 0.63 1.54� 0.22 2.24� 0.46 0.83� 0.48 <0.05 AC B Cb-Caryophyllene 0.54� 0.19 0.57� 0.13 0.45� 0.12 0.70� 0.40 NS A A Aa-Guaiene 0.45� 0.27 0.38� 0.15 0.37� 0.12 0.86� 0.48 <0.05 A AB Ca-Humulene 0.34� 0.20 0.29� 0.09 0.27� 0.09 0.66� 0.37 <0.05 A AB CGermacrene D 1.15� 0.89 0.91� 0.21 0.78� 0.27 2.78� 1.45 <0.05 A AB CPentadecane 0.19� 0.17 0.21� 0.16 0.27� 0.13 1.45� 0.42 <0.05 A AB Cd-Guaiene 0.32� 0.20 0.30� 0.09 0.22� 0.12 0.62� 0.33 <0.05 AB A BHeptadecane 1.36� 0.97 1.16� 0.47 0.95� 0.17 2.96� 1.74 <0.05 A AB CFarnesol 0.90� 0.42 1.19� 0.44 0.65� 0.24 0.78� 0.33 <0.05 A B ABOctadecane 0.12� 0.09 0.10� 0.05 0.09� 0.02 0.28� 0.16 <0.05 A AB CNonadecene 2.21�1.66 2.00� 1.34 1.59� 0.39 4.43� 2.99 <0.05 AB A BNonadecane 7.46� 5.72 7.02� 5.33 5.31�1.43 14.35� 9.87 NS A A AEicosane 0.58� 0.50 0.58� 0.58 0.42� 0.10 1.04� 0.77 NS A A AHexanol 2.40� 2.29 2.49� 2.94 1.74� 0.65 3.92� 3.20 NS A A AaNS=not significant.bResults of the HSD test: the same letter in the same row indicates no significant differences (P <0.05).cSD <0.005.


59

belonging to the genus Cymbopogon (monocotyledon), such asC. martinii, one of the common adulterants of R. damascenaessential oil, C. nardus and C. citratus, from which naturalgeraniol and geranyl acetate can be extracted, assimilatecarbon dioxide according to the C4 or Hatch-Slack cycle,and are characterized by significantly higher d13C values, ran-ging from �14 to �10%.[40]

Since any potential addition of natural flavors from C4plants at the usual level of 5–8%[15] cannot be easily identifiedby measuring the d13C values of bulk samples, the d13Cvalues of some single constituents of R. damascena samplesusing GC/C/IRMS were determined. A DB-WAX capillarycolumn was selected for this analysis, since it allowed a gooddegree of separation of citronellol, geraniol and nerol underthe applied chromatographic conditions.With regard to the application of GC/C/IRMS to essential

oils, it has been reported that monoterpenes, includingcitronellol, from the essential oils of C3 plants, such as pepper-mint, mandarin, mixed camphor, cedar-wood, turpentine andpeony, have d13C values in the range of �34 to �26%, mostly


between �29 and �27%.[41] Linalool and linalyl acetate fromlavender, spike lavender, bois de rose, bergamot, geranium,clary sage, petit grain and coriander have shown d13C valuesfrom �32 to �25%.[42] The characteristic flavor componentsof mandarin essential oil have displayed d13C values from�34.5 to �25.5%.[23,24] d13C values between �33 and �24%have been described for bergamot essential oil (containinggeranyl acetate),[32] between �31 and �23% for neroli essen-tial oil (containing nerol, geranyl acetate, farnesol)[31] andfrom �31 to �25% for lemon essential oil (containing nerol,geraniol, geranyl acetate).[19] The d13C values of citronellol,nerol and geraniol of two samples of R. damascena essentialoil ranged between �28.9 and �24.7%.[33] The d13C values ofthe aldehydes neral, geranial (and of their combination) andcitral from lemon balm, Listea cubeba, lemon myrtle, lemonand lemon tea tree oil have been found to be from �29 to�24%,[20] and those from Lippia citriodora and some commer-cial lemon oils have shown d13C values between �23 and�21%,[20] at the limit of the variability field for C3 plantcomponents.


9

F. Pellati et al.

600

The d13C values of neral, geranial and citral from C4 plants,such as lemongrass and citronella, have been described in therange from �14 to �7%.[20,21] For each sample analyzed inthese studies, these three aldehydes were characterized byvery similar d13C values. Within a closed system such as aplant, all the secondary metabolites are isotopically correlatedwith each other and with the primary metabolites (carbohy-drates), with the exception of moderate shifts due to isotopeeffects in the course of their biosynthesis, i.e. in the case ofmetabolic branching.[43] Intermolecular isotopic correlationshave been shown in several authentic essential oils, thusbecoming indicative of the authenticity of natural fla-vors.[19,22–24,31,32] This relationship is particularly true forstructurally related compounds, such as geraniol and geranylacetate, nerol and neryl acetate or linalool and linalyl acetate.It is evident that the three above-mentioned esters have d13Cvalues (around 1%) that are similar to or slightly lower thanthose of the corresponding alcohols in the essentialoil.[19,22,30–32,42]

Table 5 shows the d13C values of geranyl acetate, citronel-lol, nerol, geraniol, nonadecane, heneicosane and farnesol ofR. damascena essential oil samples. As explained in the

Table 5. d13C values (%) of specific compounds of commercial

Compoundb Geranyl acetate Citronellol Nero

Retention time (s) 1368 1392 1426BULG 01 �20.3 �27.8 �24.0BULG 02 �21.1 �28.2 �25.4BULG 03 �20.9 �28.4 �25.2BULG 04 �18.3 �27.5 �25.9BULG 05 �20.3 �28.0 �24.2BULG 06 �20.1 �27.3 �24.7BULG 07 �25.6 �28.1 �25.2BULG 08 �20.7 �27.0 �24.2Mean �20.9 �27.8 �24.9SD 2.1 0.5 0.7Min �25.6 �28.4 �25.9Max �18.3 �27.0 �24.0TURK 01 �20.7 �27.0 �25.1TURK 02 �20.4 �27.9 �25.2TURK 03 �19.4 �27.2 �24.4TURK 04 �19.7 �26.4 �23.6TURK 05 �20.3 �27.7 �24.7TURK 06 �20.5 �27.8 �25.5TURK 07 �21.6 �27.5 �23.6TURK 08 �21.3 �30.0 �22.0Mean �20.5 �27.7 �24.3SD 0.7 1.1 1.2Min �21.6 �30.0 �25.5Max �19.4 �26.4 �22.0IRAN 01 �19.9 �27.8 �23.0IRAN 02 �18.2 �27.9 �23.8IRAN 03 �19.6 �27.2 �23.2Mean �19.2 �27.6 �23.3SD 0.9 0.4 0.4Min �19.9 �27.9 �23.8Max �18.2 �27.2 �23.0aGC/C/IRMS conditions as in Experimental section.bCompounds are listed in order of elution time.*Data not shown for co-eluting peaks or % relative peak area <


Experimental section, we took into account the main com-pounds characterized by baseline separation, and also thesecondary compounds with good resolution and % relativepeak area higher than 0.5 in GC/MS or GC/FID analysis.

It is evident that for the analyzed essential oils the d13Cvalues for most of the compounds are in the usual range fornatural aromatic substances from C3 plants, except for gera-nyl acetate with higher values (up to �18%). No significantdifferences (P <0.05) are evident between the three geo-graphic origins. The comparison of the d13C values of geranylacetate with those of geraniol indicated that the ester hasvalues more than 4% higher than its parent compound,except for the sample BULG 07. This is in disagreement withwhat has previously been described, where it has been shownthat an ester has a similar or slightly lower d13C value thanthe corresponding alcohol.[19,22,30–32,42]

As a confirmation of what has been reported for otheressential oils, the GC/C/IRMS analysis of an authentic R.damascena essential oil, directly obtained in the laboratoryby hydro-distillation from fresh petals, was carried out. Inthe authentic sample, the d13C values of geraniol (�27.5%)and geranyl acetate (�27%) were found to be similar, as

essential oils of R. damascenaa

l Geraniol Nonadecane Heneicosane Farnesol

1485 1567 1828 2223�26.2 * �31 �27�26.0 * �30 �31�27.1 * �29 �32�26.0 * �29 �29�24.0 * �30 �30�26.0 * �29 �31�25.8 * �30 �32�25.0 * �29 �32�25.8 �30 �31

0.9 1 2�27.1 �31 �32�24.0 �29 �27�25.1 * �30 �29�24.5 * �31 �30�26.0 * �29 �30�24.9 * �29 �27�24.0 * �30 �31�25.7 * �30 �29�25.1 * �28 �31�24.0 * �28 �28�24.9 �29 �29

0.7 1 1�26.0 �31 �31�24.0 �28 �27�25.1 �31 �28 �30�25.9 �30 �29 �30�24.1 �30 �29 *�25.0 �29 �30

0.9 1 0�25.9 �29 �30�24.1 �28 �30

0.5 in GC/MS or GC/FID analysis.



expected for a natural product. In the light of this, the highd13C values of geranyl acetate in the commercial rose essentialoils can be explained by the addition of flavors or essentialoils from C4 plants, such as C. martinii (palmarosa). Todeeper investigate this hypothesis, two commercially avail-able samples of C. martinii essential oil were fully character-ized by GC/MS, GC/FID and GC/C/IRMS. In agreementwith literature results,[13,14] the main volatile compounds ofthese samples were found to be geraniol (79.8–82.1%)and geranyl acetate (8.6–10.4%). The d13C values of geranioland geranyl acetate in the two C. martinii essential oilswere found to be �10.2 and �11.5% and �10.8 and �9.9%,respectively, which are typical for a C4 plant. Therefore, itcan be hypothesized that the commercial rose essential oilsare made up of a mixture containing 95% R. damascenaessential oil, e.g. with 20% geraniol (d13C: –27.5%) and0.5% geranyl acetate (d13C: –27%), and 5% C. martinii, e.g.with 80% geraniol (d13C: –10%) and 10% geranyl acetate(d13C: –11%), resulting in 23% geraniol (d13C: –24.5%) and1% geranyl acetate (d13C: –19%). Among the R. damascenaessential oils, only BULG 07 showed an isotopic profile thatcould indicate the absence of the addition of an essential oilfrom a C4 plant.The addition of natural and/or synthetic components to

essential oils is a fairly common practice in the essential oilindustry. Enantioselective capillary GC and GC/C/IRMSare important tools in the authenticity control of flavors andessential oils.[16,44] The adulteration of a commerciallyavailable rose oil sample by the addition of the unnaturalenantiomers, (+)-cis-rose oxide, (+)-trans-rose oxide and (R)-(+)-citronellol, has been previously described.[45] In relationto GC/C/IRMS, this study demonstrated for the first timethat geranyl acetate can be considered as a suitable endogen-ous reference compound for the authenticity assessment ofrose essential oil.While reliable criteria for authenticity are reflected by

the enantiomeric composition of major chiral constituentsin some essential oils,[46] enantioselective GC is less conclu-sive for those oils with varying enantiomeric composition ofthe chiral constituents.[46] GC/C/IRMS did not show theselimitations and it is therefore a highly recommended techni-que for detecting possible adulteration of essential oils.

CONCLUSIONS

The described results demonstrated that stable isotope ratioanalysis, using GC/C/IRMS, represents a reliable technique, incombination with GC/MS and GC/FID, for the completephytochemical characterization of rose essential oil. The addi-tion of a natural cheaper oil from a C4 plant (C. martinii) wasobserved in most of the samples analyzed. Since the reportedadulteration was undetected by GC/FID and GC/MS analysisof rose essential oils, the d13C isotopic fingerprint representsan effective tool for the authenticity assessment of theseeconomically important natural products.

60

AcknowledgementThe authors are grateful to Dr Carlo Dappino of Agronatura(Spigno Monferrato, Alessandria, Italy) for providing anauthentic sample of R. damascena essential oil.


REFERENCES

[1] M. Jalali-Heravi, H. Parastar, H. Sereshti. Development ofa method for analysis of Iranian damask rose oil: combina-tion of gas chromatography–mass spectrometry with che-mometric techniques. Anal. Chim. Acta 2008, 623, 11.

[2] A. Almasirad, Y. Amanzadeh, A. Taheri, M. Iranshahi.Composition of a historical rose oil sampke (Rosa damascenaMill., Rosaceae). J. Essent. Oil Res. 2007, 19, 110.

[3] F. Ayci, M. Aydinli, Ö. A Bozdemir, M. Tutaş. Gas chromato-graphic investigation of rose concrete, absolute and solidresidue. Flavour Fragr. J. 2005, 20, 481.

[4] E. Basim, H. Basim. Antibacterial activity of Rosa damascenaessential oil. Fitoterapia 2003, 74, 394.

[5] N. Kumar, P. Bhandari, B. Singh, S. S. Bari. Antioxidant activityand ultra-performance LC-electrospray ionization-quadrupoletime-of-flight mass spectrometry of phenolics-based finger-printing of rose species: Rosa damascena, Rosa bourboniana andRosa brunonii. Food Chem. Toxicol. 2009, 47, 361.

[6] V. Hajhashemi, A. Ghannadi, M. Hajiloo. Analgesic andanti-inflammatory effects of Rosa damascena hydroalcoholicextract and its essential oil in animal models. Iran. J. Pharm.Res. 2010, 9, 163.

[7] M. H. Boskabady, S. Mani, H. Rakhshandah. Relaxanteffects of Rosa damascena on guinea pig tracheal chains andits possible mechanism(s). J. Ethnopharmacol. 2006, 106, 377.

[8] M. H. Boskabady, M. N. Shafei, Z. Saberi, A. Somayeh. Phar-macological effects of Rosa Damascena. Iran. J. Basic Med. Sci.2011, 14, 295.

[9] W. N. Setzer. Essential oils and anxiolytic aromatherapy.Nat. Prod. Commun. 2009, 4, 1305.

[10] G. Bianchi, M. Nuzzi, A. A. Leva, A. Rizzolo. Developmentof a headspace-solid phase micro extraction method tomonitor changes in volatile profile of rose (Rosa hybrida, cvDavid Austin) petals during processing. J. Chromatogr. A2007, 1150, 190.

[11] G. Matheis, in Flavourings, Production, Composition, Applica-tions, Regulations, (Ed: H. Ziegler). Wiley-VCH, Weinheim,Germany, 2007, pp. 137–389.

[12] R. S. Singhal, P. R. Kulkarni, D. V. Rege, in Handbook of Herbsand Spices, (Ed: K. V. Peter). Woodhead Publishing Ltd.,Cambridge, UK, 1997, pp. 22–29.

[13] G. R. Mallavarapu, B. R. R. Rao, P. N. Kaul, S. Ramesh,A. K. Bhattacharya. Volatile constituents of the essential oilsof the seeds and the herb of palmarosa (Cymbopogon martinii(Roxb.) Wats. var.motia Burk.). Flavour Fragr. J. 1998, 13, 167.

[14] V. S. Dubey, G. R. Mallavarapu, R. Luthra. Changes in theessential oil content and its composition during palmarosa(Cymbopogon martinii (Roxb.) Wats. var. motia) inflorescencedevelopment. Flavour Fragr. J. 2000, 15, 309.

[15] M. Valussi, in Il grande manuale dell’aromaterapia, Fondamentidi scienza degli oli essenziali, (Ed: M. Valussi). TecnicheNuove, Milano, Italy, 2005, p. 507.

[16] A. Mosandl. Enantioselective capillary gas-chromatographyand stable-isotope ratio mass-spectrometry in the authenti-city control of flavors and essential oils. Food Rev. Int. 1995,11, 597.

[17] W. J. Meier-Augenstein. Applied gas chromatographycoupled to isotope ratio mass spectrometry. J. Chromatogr.A 1999, 842, 351.

[18] R. Braunsdorf, U. Hener, G. Przibilla, S. Piecha, A. Mosandl.The influence of analytical and technological procedures onthe 13C/12C ratio of orange oil compounds. Z. Lebensm.Unters Forsch. 1993, 197, 24.

[19] R. Braunsdorf, U. Hener, S. Stein, A. Mosandl. Comprehen-sive CGC-IRMS in the authenticity control of flavors andessential oils.1.lemon oil. Z. Lebensm. Unters Forsch. 1993,197, 137.


1

F. Pellati et al.

602

[20] T.-T. Nhu-Trang, H. Casabianca, M.-F. Grenier-Loustalot.Authenticity control of essential oils containing citronellaland citral by chiral and stable-isotope gas-chromatographicanalysis. Anal. Bioanal. Chem. 2006, 386, 2141.

[21] S. Wagner, P. Vreca, A. Leis, H. Boechzelt. Carbon isotoperatio analysis of authentic and commercial essential oils oflemon balm. Nat. Prod. Commun. 2008, 3, 1165.

[22] C. Frank, A. Dietrich, U. Kremer, A. Mosandl. GC-IRMS inthe authenticity control of the essential oil of Coriandrumsativum L. J. Agric. Food Chem. 1995, 43, 1634.

[23] S. Faulhaber, U. Hener, A. Mosandl. GC/IRMS analysis ofmandarin essential oils. 2. d13C(PDB) values of characteristicflavor components. J. Agric. Food Chem. 1997, 45, 4719.

[24] L. Schipilliti, P. Q. Tranchida, D. Sciarrone, M. Russo,P. Dugo, G. Dugo, L. Mondello. Genuineness assessmentof mandarin essential oils employing gas chromatography-combustion-isotope ratio MS (GC-C-IRMS). J. Sep. Sci.2010, 33, 617.

[25] B. Faber, K. Bangert, A. Mosandl. GC–IRMS and enantiose-lective analysis in biochemical studies in dill (Anethumgraveolens L.). Flavour Fragr. J. 1997, 12, 305.

[26] S. Bilke, A. Mosandl. 2H/1H and 13C/12C isotope ratios oftrans-anethole using gas chromatography-isotope ratiomass spectrometry. J. Agric. Food Chem. 2002, 50, 3935.

[27] C. Ruff, K. Hör, B. Weckerle, T. König, P. Schreier. Authenti-city assessment of estragole and methyl eugenol by on-linegas chromatography-isotope ratio mass spectrometry.J. Agric. Food Chem. 2002, 50, 1028.

[28] S. Sewenig, U. Hener, A. Mosandl. Online determination of2H/1H and 13C/12C isotope ratios of cinnamaldehyde fromdifferent sources using gas chromatography isotope ratiomass spectrometry. Eur. Food Res. Technol. 2003, 217, 444.

[29] M. Greule, C. Hänsel, U. Bauermann, A. Mosandl. Feedadditives: authenticity assessment using multicompo-nent-/multielement-isotope ratio mass spectrometry. Eur.Food Res. Technol. 2008, 227, 767.

[30] J. Jung, S. Sewenig, U. Hener, A. Mosandl. Comprehensiveauthenticity assessment of lavender oils using multielement/multicomponent isotope ratio mass spectrometry analysis andenantioselective multidimensional gas chromatography-massspectrometry. Eur. Food Res. Technol. 2005, 220, 232.

[31] I. Bonaccorsi, D. Sciarrone, L. Schipilliti, A. Trozzi, H.A. Fakhry,G. Dugo. Composition of Egyptian neroli oil. Nat. Prod.Commun. 2011, 6, 1009.

[32] L. Schipilliti, P. Dugo, L. Santi, G. Dugo, L. Mondello. Authen-tication of bergamot essential oil by gas chromatography-combustion-isotope ratio mass spectrometer (GC-C-IRMS).J. Essent. Oil Res. 2011, 23, 60.


[33] J. C. Bayle, H. Casabianca. Isotopic 13C/12C and chiralanalyses of constituents of essential oils of rose, concreteand absolute. Rivista Italiana EPPOS 1996, 7, 262.

[34] F. Camin, L. Bontempo, L. Ziller, C. Piangiolino, G. Morchio.Stable isotope ratios of carbon and hydrogen to distinguisholive oil from shark squalene-squalane. Rapid Commun.Mass Spectrom. 2010, 24, 1810.

[35] V. Gochev, L. Jirovetz, K. Wlcek, G. Buchbauer, E. Schmidt,A. Stoyanova, A. Dobreva. Chemical composition andantimicrobial activity of historical rose oil from Bulgaria. J.Essent. Oil Bear. Plants 2009, 12, 1.

[36] A. Bayrak, A. Akgül. Volatile oil composition of Turkishrose (Rosa damascena). J. Sci. Food Agric. 1994, 64, 441.

[37] J. C. Chalchat, M. M. Özcan. A comparative investigation ofthe composition of rose (Rosa damascena Mill.) oil producedby using two different methods. J. Essent. Oil Bear. Plants2009, 12, 447.

[38] J. Safaei-Ghomi, S. Akhoondi, H. Batooli, M. Dackhili.Chemical variability of essential oil components of two Rosax damascena genotypes growing in Iran. Chem. Nat. Compd.2009, 45, 262.

[39] A. Mostafavi, D. Afzali. Chemical composition of the essen-tial oils from Rosa damascena from two different locations inIran. Chem. Nat. Compd. 2009, 45, 110.

[40] B. N. Smith, S. Epstein. Two categories of 13C/12C ratios forhigher plants. Plant Physiol. 1971, 47, 380.

[41] D. J. Zhang, P. R Wang, Z. C. Mao, G. J. Xu. Stable carbonisotope composition of monoterpanes in essential oils andcrude oils. Sci. China Earth Sci. 2010, 53, 522.

[42] S. Hanneguelle, J. N. Thibault, N. Naulet, G. J. Martin.Authentication of essential oils containing linalool andlinalyl acetate by isotopic methods. J. Agric. Food Chem.1992, 40, 81.

[43] H. L. Schmidt, R. A. Werner, A. Rossmann, A. Mosandl,P. Schreier, in Flavourings, Production, Composition, Applications,Regulations, (Ed: H. Ziegler). Wiley-VCH,Weinheim, Germany,2007, pp. 602–663.

[44] A. Mosandl. Authenticity assessment: a permanent challengein food flavor and essential oil analysis. J. Chromatogr. Sci.2004, 42, 440.

[45] P. Kreis, A. Mosandl. Chiral compounds of essential oils. PartXII. Authenticity control of rose oils, using enantioselectivemultidimensional gas chromatography. Flavour Fragr. J.1992, 7, 199.

[46] W. A. König, C. Fricke, Y. Saritas, B. Momeni, G. Hohenfeld.Adulteration or natural variability? Enantioselective gaschromatography in purity control of essential oils. J. HighResol. Chromatogr. 1997, 20, 55.


essential oils of rosa

Documents

d13c values of bulk

rosa damascena

unusual d13c values

natural cheaperoil

natural aromatic substances

natural product

c3 plants

higher values