Identification by GC-MS of cymene isomers and 3,7,7-trimethylcyclohepta-1, 3,5-triene in essential oils

Article abstract of DOI:10.1007/s10600-006-0256-6

Retention indices of cymene isomers published in popular GC-MS atlases were found to be erroneous by analyzing synthetic samples. The following retention indices (RI) were found using a nonpolar phase (diphenyl: dimethylpolysiloxane, 5:95) for four essential-oil components with indistinguishable mass spectra: 3,7,7-trimethylcyclohepta-1,3,5-triene (RI = 970), m-cymene (RI = 1022), p-cymene (RI = 1024), and o-cymene (RI = 1039). The relative distributions of these compounds were evaluated based on the analysis of about 1000 essential oils. Simple methods were given for preparing standard mixtures of isomeric compounds for identification by GC-MS.

Full text of DOI:10.1007/s10600-006-0256-6

Chemistry of Natural Compounds, Vol. 42, No. 6, 2006
IDENTIFICATION BY GC—MS OF CYMENE ISOMERS
AND 3,7,7-TRIMETHYLCYCLOHEPTA-1,3,5-TRIENE
IN ESSENTIAL OILS
1
2
E. P. Romanenko and A. V. Tkachev
UDC 547.913+543.544.43+543.51
Retention indices of cymene isomers published in popular GC—MS atlases were found to be erroneous by
analyzing synthetic samples. The following retention indices (RI) were found using a nonpolar phase
(diphenyl:dimethylpolysiloxane, 5:95)forfour essential-oil components withindistinguishablemassspectra:
3
,7,7-trimethylcyclohepta-1,3,5-triene (RI = 970), m-cymene (RI = 1022), p-cymene (RI = 1024), and
o-cymene (RI = 1039). The relative distributions of these compounds were evaluated based on the analysis
of about 1000 essential oils. Simple methods were given for preparing standard mixtures of isomeric
compounds for identification by GC—MS.
Key words: essential oils, 3,7,7-trimethylcyclohepta-1,3,5-triene, m-cymene, p-cymene, o-cymene, retention indices,
GC—MS.
Monoterpenes of the p-menthane series are widelydistributed in the plant world and are the main components of many
essential oils and plant extracts. One of the most common compounds of this type is p-cymene (1), which is the starting material
for synthesizing products for various purposes, mainly oxidized derivatives [1], exhibits biological activity [2], is a component
of common spices.
Identification of 1 in plant extracts is often difficult because the isomeric m- (2) and o-isomers (3) have mass spectra
that are practicallyindistinguishable from that of p-cymene. The situation is further complicated bythe fact that GC—MS data
in commonlyused atlases published in 1989 [3], 1995 [4], and 2001 [5] give erroneous retention indices (RI) and elution orders
for o- and p-cymene. As a result, many published scientific studies of essential oil compositions incorrectly identify cymene
isomers. An example of this is a technique for precise analysis of lavander oil [6] in which peaks for cymene isomers are
incorrectly identified based on the citation of an atlas [4].
We synthesized samples of cymene isomers 1-3, confirmed their structures by PMR and 13C NMR spectra, and
determined their GC—MS properties. As it turned out, there is another component, 3,7,7-trimethylcyclohepta-1,3,5-triene (4),
with similar properties (see Experimental).
2
3
4
1
1
) N. N. Vorozhtsov Novosibirsk Institute of Organic Chemistry, pr. Akad. Lavrent′eva, 9, Novosibirsk, 630090,
Russian Federation; 2) Novosibirsk State University, ul. Pirogova, 2, Novosibirsk, 630090, fax (383) 330 97 52, e-mail:
atkachev@nioch.nsc.ru. Translated from Khimiya Prirodnykh Soedinenii, No. 6, pp. 569-571, November-December, 2006.
Original article submitted June 2, 2006.
©
0
009-3130/06/4206-0699 2006 Springer Science+Business Media, Inc.
699

We used GC—MS to analyze about 1000 samples of commercially available essential oils and oils obtained under
laboratory conditions from plants of the families Apiaceae, Asteraceae, Lamiaceae, Pinaceae, Cupressaceae, Salicaceae,
Valerianaceae, Ericaceae, Araceae, Betulaceae, Grossulariaceae, Rosaceae, andRutaceae. Manyessential oils andplant extracts
contained 1, most often asa minor component in a mixture generallydominated byp-menthane-type derivatives. Small amounts
of 2, 3, and 4, mainly in products from conifers, were observed. m-Cymene (2) was found in some turpentines from pine (Pinus
spp.), larch (Larix spp.), and juniper (Juniperus spp.), always as an insignificant (content <0.05%) impurity and always with
p-cymene (p/m ratio 15:1 at a minimum). o-Cymene (3) was found in trace (up to 0.05%) amounts in monoterpene fractions
of resins and in extracts and essential oils from various parts ofconiferous trees (P. pityusa Stev., P. mugo Turra., and P. sibirica
R. Mayr.). The content of 3 was maximum (0.8% of the monoterpene fraction) in turpentine from Japanese larch resin
[
L. leptolepis (Sieb. et Zucc.) Gord.]. An insignificant amount of 4 was found in practically all samples including industrial
products from reprocessing of coniferous tree resins and in pine and fir oils, in the monoterpene fraction of pine (Pinus spp.),
larch (Larix spp.), fir (Abies spp.), spruce (Picea spp.), and microbiota (Microbiota decussata Kom.).
The origin of 1-4 is not yet clear. Apparently 1 in many instances is an artifact and is formed by oxidation of various
p-menthane-type derivatives. Compound 4 may be a dehydration product of 3-carene oxidation products. Heating 4 above
3
00 C leads in turn to its destruction and formation of a mixture of products including p- and m-cymene [7].
Because the difference in the RI of p- and m-cymene is only two units, these isomers can be reliably identified only by
using standards since noticeable deviations from the table values of the RI can be observed on going from some nonpolar phases
to others (and even on going to a column with the same type of phase but from different manufacturers). Several methods can
berecommendedfor preparingstandardmixturescontainingtheseisomers: 1)dehydrogenation oflimoneneproducesp-cymene;
2
) dehydrogenation of 3-carene produces a mixture of p- and m-cymene; 3) Friedel-Krafts alkylation of toluene by
isopropylbromide gives a mixture of o-, m-, and p-cymene; 4) thermal degradation of 3,4-caranediol diacetate produces
,7,7-trimethylcyclohepta-1,3,5-triene.
3
EXPERIMENTAL
General Comments. GC—MS was carried out in a Hewlett—Packard 5890/II gas chromatograph with an HP 5972A
mass-selective detector, 30-m HP-5 MS quartz capillary column [30 m × 0.25 mm, 0.25 µm stationary phase
diphenyl:dimethylsiloxane copolymer, 5:95), He carrier gas (1 mL/min), vaporizer temperature 280°C, column 50°C (2 min)-
0°C/min-280°C, ion source 173°C, interface between GC and MS detector 280°C, ionizing-electron energy 70 eV, data
(
1
collection 1.2 scans/s for mass range 30-650 amu].
Preparative GC was performed byrefurbishingfor preparativeworka Chrom-5chromatograph with a flame-ionization
detector and N carrier gas. Products were separated using a steel column [4500 × 6 mm, 15% Apiezon L on Chromaton N-AW
2
(
0.250-0.315 mm), column temperature 120°C, vaporizer and detector 250°C].
PMR and 13C NMR spectra of CDCl solutions were recorded on a Bruker DRX-500 spectrometer (500.13 MHz for
3
1H; 125.75 MHz, 13C) at 25°C using solvent signals (δ 76.90 ppm, δ 7.24) as internal standards.
C
H
Limonene Dehydrogenation. Limonene (170 mg, 1.2 mmol) was mixed with conc. H SO (0.1 mL) and stirred at
2
4
room temperature for 15 min. The isomerization products were extracted with hexane (5 × 1 mL). The combined extracts were
washed with water (1 mL) and dried over anhydrous MgSO . Solvent was removed to afford p-cymene (1, 40 mg) without
4
impurities of other isomers.
3
-Carene Dehydrogenation. 3-Carene(330mg, 2.4 mmol) was treated with conc. H SO (0.3mL)andstirredat room
2 4
temperature for 30 min. The reaction products were extracted with hexane (5 × 1 mL). The combined extracts were washed
with water (1 mL) and dried over anhydrous MgSO . Solvent was removed to afford a mixture of hydrocarbons (83 mg)
4
containing ~60% of p- (1) and m-cymene (2) in a 4:1 ratio.
Toluene Alkylation. A mixture of toluene (4.0 mL, 37.8 mmol) and AlCl (0.18 g, 1.3 mmol) was cooled to 0°C,
3
stirred vigorously, treated dropwise over 60 min with i-PrBr (1.0 mL, 10.6 mmol), held at 0°C for another 30 min after the
addition was finished, and poured onto ice. The organic layer was separated, washed with NaHCO solution (1 M, 2 × 1 mL)
3
and water (1 mL), and dried over MgSO . The excess of toluene was removed to afford a mixture (0.98 g) of 1, 2, and 3 in a
4
2
:5:1 ratio.
7
00

3
,7,7-Trimethyl-1,3,5-cycloheptatriene(4). 3,4-Caranedioldiacetate(750mg, 2.9mmol)thatwaspreparedasbefore
8] was heated in a sealed ampul at 230°C for 2 h, cooled to room temperature, diluted with hexane (10 mL), washed with
saturated aqueous NaHCO (2 × 1 mL) and water (1 mL), and dried over MgSO . Solvent was removed. The solid was
[
3
4
chromatographed over SiO (hexane eluent) to isolate a hydrocarbon fraction (120 mg) containing according to GC 4 (65%)
2
that was then purified by preparative GC.
PMR spectrum (δ, ppm, J/Hz): 0.98 (6H, s, Me -7), 2.01 (3H, s, Me-3), 5.10 (1H, d, J = 10.2, H-6), 5.16 (1H, d,
2
J = 10.3, H-1), 5.90 (1H, d, J = 10.3, H-2), 5.98 (1H, dd, J = 1.02, 6.6, H-5), 6.25 (1H, d, J = 6.6, H-4). 13C NMR spectrum
(CDCl ): 24.33 (q, CH –C-3), 26.07 [2q, (CH ) –C-7], 34.78 (s, C-7), 123.94 (d, C-5), 127.03 (d, C-4), 127.28 (d, C-2), 132.28
3 3 3 2
(
d, C-6), 133.59 (d, C-1), 138.46 (s, C-3).
Retention Indices (HP 5 MS Column, Diphenyl:dimethylsiloxane copolymer, 5:95) and Electron-Impact
Mass Spectra of 1-4.
p-Cymene (1). RI = 1024. Mass spectrum (m/z): 134 (27), 120 (10), 119 (100), 117 (12), 115 (5), 105 (2), 104 (2),
03 (3), 93 (2), 91 (17), 77 (5), 65 (4), 41 (2), 39 (2).
m-Cymene (2). RI = 1022. Mass spectrum (m/z): 134 (32), 120 (10), 119 (100), 117 (13), 115 (7), 105 (3), 104 (3),
03 (4), 93 (2), 92 (2), 91 (20), 79 (2), 78 (2), 77 (6), 65 (4), 63 (2), 51 (2), 41 (2), 39 (2).
o-Cymene (3). RI = 1039. Mass spectrum (m/z): 134 (31), 120 (10), 119 (100), 117 (13), 115 (7), 105 (2), 104 (3),
03 (4), 93 (2), 92 (2), 91 (20), 79 (2), 78 (2), 77 (5), 65 (4), 63 (2), 58 (2), 51 (2), 41 (2), 39 (2).
1
1
1
1
3
,7,7-Trimethylcyclohepta-1,3,5-triene (4). RI = 970. Mass spectrum (m/z): 134 (11), 120 (9), 119 (100), 117 (10),
15 (5), 105 (4), 104 (3), 103 (5), 93 (3), 92 (4), 91 (36), 79 (5), 78 (3), 77 (10), 65 (5), 63 (2), 51 (4), 41 (7), 39 (5).
REFERENCES
1
.
a) F. Vaudano and P. Tissot, Electrochim. Acta, 46, 875 (2001); b) G. Harms, R. Rabus, and F. Widdel, Arch.
Microbiol., 172, 303 (1999); c) R. R. L. Martins, M. G. P. M. S. Neves, A. J. D. Silvestre, A. M. S. Silva, and
J. A. S. Cavaleiro, J. Mol. Catal. A: Chem., 137, 41 (1999).
2.
3.
4.
B. Delgado, A. Palop, P. S. Fernandez, and P. M. Periago, Eur. Food Res.Technol., 218, 188 (2004).
R. P. Adams, Identification of Essential Oils by Ion Trap Mass Spectroscopy, Academic Press, San Diego (1989).
R. P. Adams, Identification of Essential Oil Components by Gas Chromatography/Mass Spectrometry, Allured
Publ. Corp., Carol Stream, IL, USA (1995).
5
.
R. P. Adams, Identification of Essential Oil Components by Gas Chromatography/Quadrupole Mass Spectroscopy,
Allured Publ. Corp., Illinois, USA (2001).
6.
7.
8.
R. Shellie, L. Mondello, P. Marriott, and G. Dugo, J. Chromatogr. A, 970, 225 (2002).
J. A. Berson and M. R. Willcott III, J. Am. Chem. Soc., 88, No. 11, 2494 (1966).
P. J. Kropp, J. Am. Chem. Soc., 88, No. 21, 4926 (1966).
701