Anal Bioanal Chem (2011) 400:3109–3123

DOI 10.1007/s00216-011-4976-5

ORIGINAL PAPER

Application of high-performance liquid chromatography diode

array detection and mass spectrometry to the analysis
of characteristic compounds in various essential oils
Claudia Turek & Florian C. Stintzing

Received: 2 February 2011 / Revised: 28 March 2011 / Accepted: 31 March 2011 / Published online: 16 April 2011
# Springer-Verlag 2011

Abstract A high-performance liquid chromatography Introduction

(HPLC) method was established using an analytical
reversed-phase column and gradient elution to achieve Throughout centuries, essential oils have been known and
chromatographic separation of typical compounds in applied in folk medicine and for cosmetic purposes, the first
essential oils. For detection, a diode array detector record probably dating back to the Egyptians [1]. Biosynthe-
monitoring different wavelengths simultaneously as well sized by a considerable number of aromatic plants, essential
as a mass spectrometer (MS) were used. Atmospheric oils represent mixtures of a huge variety of volatile
pressure chemical ionization operating in the positive mode substances, characterized by a strong sensorial impact. Most
turned out to be a suitable tool to detect volatiles of of the more or less lipophilic components are structurally
different chemical classes and to identify them in essential related mono- and sesquiterpenoids of different chemical
oil matrices. Characteristic fingerprints of eucalyptus, classes like hydrocarbons (the so-called terpenes), alcohols,
lavender, may chang, pine, rosemary, thyme, and turpentine ketones, aldehydes, esters, or oxides [1]. Furthermore,
essential oils monitored at a representative wavelength aromatic compounds (phenylpropanoids) and paraffinic
(220 nm) demonstrated the suitability of HPLC in essential hydrocarbon derivatives can be found as plant volatiles as
oil analysis. Additional monitoring wavelengths (210, 250, well [2]. Essential oils are widely used in flavor and cosmetic
and 280 nm) provided useful information about the identity industry due to their strong and generally pleasant odor.
of the specific component and opened the possibility to Moreover, antimicrobial, antifungal, nematicidal, insecticid-
differentiate presumably coeluting compounds by means of al, antioxidant, and diverse pharmacological properties have
their distinct absorption behavior. Finally, peak assignment been reported [3–8]. The multitude of uses requires a
in seven essential oils was performed on the basis of thorough knowledge on the composition of essential oils.
characteristic retention times and UV and MS data of a Usually, gas chromatography (GC), often in combination
broad set of reference volatiles. with headspace, solid-phase microextraction, and/or mass
spectrometry has been the method of choice to characterize
Keywords Essential oils . Terpenoids . Volatiles . HPLC/ the complex mixture of essential oils [9–11]. Often neglected
DAD . Mass spectrometry . LC/APCI-MS however, structural alterations of thermally labile compounds
may occur during GC analyses due to high injector temper-
atures or catalytically active surfaces like columns or liners
Electronic supplementary material The online version of this article [12–17]. Moreover, GC analyses of water-based matrices
(doi:10.1007/s00216-011-4976-5) contains supplementary material, such as detergents and cosmetic products require time-
which is available to authorized users. consuming steps during sample work-up. Therefore, high-
C. Turek : F. C. Stintzing (*) performance liquid chromatography (HPLC) may represent a
Department of Research and Development, supplementary or even alternative method because of its
WALA Heilmittel GmbH,
versatility, sensitivity, and selectivity [12, 18]. Yet, only a
Dorfstrasse 1,
73087 Bad Boll/Eckwälden, Germany comparatively small number of reports on essential oil
e-mail: [email protected] analyses by HPLC can be found in the literature. To the
3110 C. Turek, F.C. Stintzing

most part, its use has been reserved for the detection of the sylvestris L., Pinaceae, Caesar & Loretz, Hilden, Germany),
non-volatile fraction in citrus oils [e.g., 9, 19–22], as a sample rosemary oil (Rosmarinus officinalis L., Lamiaceae, Sensient
cleaning or fractionating step prior to GC [e.g., 23–25] and in Essential Oils, Bremen, Germany), thyme oil (Thymus
rare cases for the detection of specific terpenoid volatiles vulgaris L., Lamiaceae, Brüder Unterweger, Thal-Assling,
[18, 26–28]. Still, reports on a more systematic approach to Austria), and turpentine oil (Pinus pinaster SOLAND.,
HPLC analyses of essential oils are scattered [12, 29–39]. Pinaceae, Brüder Unterweger, Thal-Assling, Austria). All
Therefore, the aim of the present paper was to establish a high- essential oils were obtained by steam distillation of the
performance diode array detection mass spectrometry respective plant species except turpentine essential oil, which
(HPLC/DAD/MS) method for the characterization of essential was the distillate of pine turpentine.
oils commonly applied in pharmaceutical and cosmetic
preparations. Retention, UV, and MS data of a broad range HPLC/DAD/MSMS analyses
of standard substances common in essential oils and belong-
ing to different chemical classes were systematically Chromatographic analyses were carried out on an Agilent
compiled. On this basis, compound assignment in essential 1200 HPLC system (Agilent Technologies Inc., Palo Alto,
oils from eucalyptus, lavender, may chang, pine, rosemary, CA), equipped with ChemStation Software, a micro-vacuum
thyme, and turpentine was possible. degasser G1322A, a binary pump G1312A, an autosampler
G1329A, a thermostatic column compartment G1316A, and a
diode array detector G1315B. An analytical Modulo-Cart MS
Materials and methods Uptisphere 3ODB reversed-phase (RP) column (150 mm×
2.1 mm ID) with a particle size of 3 μm from Interchim
Chemicals and reagents (Montluçon, France), combined with a Modulo-Cart QS RP
security guard column (2 μm, 10 mm×2.1 mm ID) was used
Analytical reference standards were obtained from Carl for chromatographic separation at 23 °C. Using purified water
Roth (Karlsruhe, Germany) except p-anisaldehyde, bornyl (mobile phase A) and acetonitrile (mobile phase B) at a flow
acetate, carvacrol, β-citronellol, geraniol, lavandulyl ace- rate of 0.21 mL/min, HPLC analyses started isocratically with
tate, α-phellandrene (Fluka, Buchs, Switzerland), eugenol 30% B for 2 min, followed by a linear gradient to 100% B in
(European Directorate for the Quality of Medicines & 40 min, then an isocratic step for 10 min before re-
HealthCare, Strasbourg, France) and α-bisabolol, camphor, equilibration to starting conditions within 10 min. The
carveol, caryophyllene oxide, citral, p-α-dimethylstyrene, injection volume of each sample was 10 μL. Simultaneous
farnesol, linalool, linalyl acetate, linalool oxide, nerol, monitoring was performed at 210, 220, 250, and 280 nm
ocimene, and 3-octanone (Sigma-Aldrich, Steinheim, (bandwidth and slit 4 nm), respectively. Moreover, UV–vis
Germany). Prior to analysis, standard compounds as well as spectra were recorded (190 to 720 nm) in order to obtain three-
essential oils were diluted in ethanol (gradient-grade for liquid dimensional (3D) spectral information of each reference
chromatography, Merck, Darmstadt, Germany). Purified water compound and essential oil sample.
(0.055 μS/cm) and acetonitrile of liquid chromatography/mass The HPLC system was connected in series with a Bruker
spectrometry (LC/MS)-grade (Sigma-Aldrich, St. Louis, MO) Daltonic (Bremen, Germany) HCT Ultra ion trap mass
were used for chromatography. spectrometer fitted with an atmospheric pressure chemical
ionization (APCI) ion source operating in the positive
Synthesis of thymol methyl ether mode. The instrument was controlled by an Esquire Control
Software (6.1).
Thymol methyl ether was prepared by methylation of Direct infusion of standard solutions was used for optimi-
thymol according to Beckert et al. [40] and subsequently zation of MS parameters. Due to the broad structural diversity
dissolved in GC-grade methanol (Merck, Darmstadt, of the components considered, the following settings were
Germany) for analysis. chosen to best suit all compounds: capillary voltage, 3,950 V;
corona current, 2,500 nA; skimmer voltage, 40 V; capillary
Essential oils exit voltage, 116.7 V; and trap drive, 22.4. Nitrogen was used
as drying gas at a flow rate of 5 L/min at 300 °C. The nebulizer
The following commercially available essential oils were used was set at a pressure of 40 psi, and the vaporizer temperature
in this study: eucalyptus oil (Eucalyptus sp., Myrtaceae, was 350 °C. Mass spectra of the HPLC eluates were recorded
Sensient Essential Oils, Bremen, Germany), lavender oil during the chromatographic separation from m/z 15–650. To
(Lavandula angustifolia MILL., Lamiaceae, Vigalex, Sofia, gain further structural information, collision-induced frag-
Bulgary), may chang oil (Litsea cubeba PERS., Lauraceae, mentation spectra were acquired in the auto MS–MS mode
Sensient Essential Oils, Bremen, Germany), pine oil (Pinus with a compound stability of 100%.
HPLC analysis of essential oils 3111

Table 1 LC/DAD and APCI(+)-mass spectral data for volatile compounds commonly found in essential oils

Compounda tR [min] lmax, UV [nm]b Formula Mr [g/mol] APCI(+)-MS m/zmaxc MS2 (m/zmax) m/z

Alcohols
α-Bisabolol 35.0 <200 C15H26O 222.37 205d 149, 121
Borneol 19.9 <200 C10H18O 154.25 137d 95, 81
cis-Carveol 18.1 <200 C10H16O 152.23 135d 107, 93
trans-Carveol 18.5 <200 C10H16O 152.23 135d 107, 93
β-Citronellol 24.1 <200 C10H20O 156.27 157e 137, 83
cis-Farnesol 33.7 <200 C15H26O 222.37 205d 149, 121
trans-Farnseol 34.2 <200 C15H26O 222.37 205d 149, 121
Geraniol 21.6 <200 C10H18O 154.25 137d 95, 81
Lavandulol 20.9 <200 C10H18O 154.25 155e 137, 111, 81
Linalool 21.9 <200 C10H18O 154.25 137d 95, 81
Nerol 21.6 <200 C10H18O 154.25 137d 95, 81
Terpinene-4-ol 20.9 <200 C10H18O 154.25 137d 107, 95, 81
α-Terpineol 19.2 <200 C10H18O 154.25 137d 107, 95, 81
Aldehydes
p-Anisaldehyde 11.1 221, 283 C8H8O2 136.15 137e 109
Geranial 24.1 241 C10H16O 152.23 135d 107, 93
Neral 23.2 241 C10H16O 152.23 135d 107, 93
Ketones
e
Camphor 20.7 <200, 288 C10H16O 152.23 153 135, 109, 95
e
Carvone 20.3 238 C10H14O 150.22 151 123, 109
e
3-Octanone 23.5 <200, 278 C8H16O 128.21 129 111, 70
e
Verbenone 14.0 258 C10H14O 150.22 151 123, 109
Terpenic hydrocarbons
Camphene 38.6 <200 C10H16 136.23 137e 107, 95, 81
δ-3-Carene 39.7 <200 C10H16 136.23 166f 147, 137, 123
β-Caryophyllene 47.8 <200 C15H24 204.35 234f; 203g 217, 216, 161; 147
α-Cedrene 51.2 <200 C15H24 204.35 204h 161, 147, 119
p-Cymene 33.7 212sh, 266, 274sh C10H14 134.22 135e 107, 93
p-α-Dimethylstyrene 31.7 206, 248 C10H12 132.20 133e 117, 105
Limonene 38.6 <200 C10H16 136.23 153i; 135g 135, 107; 105
β-Myrcene 37.1 224 C10H16 136.23 155j 137, 109
Ocimene 36.8 236, 276sh C10H16 136.23 153i 135, 109
α-Phellandrene 38.7 264 C10H16 136.23 194k; 153i; 135g 150; 109; 107
α-Pinene 40.8 206 C10H16 136.23 166f; 137e 107; 107
β-Pinene 39.0 200 C10H16 136.23 137e 107, 93, 81
Sabinene 36.8 206 C10H16 136.23 137e 107, 93, 81
α-Terpinene 38.2 268 C10H16 136.23 153i 135, 109
γ-Terpinene 38.3 <200 C10H16 136.23 153i 135
Terpinolene 38.3 <200 C10H16 136.23 208l; 153i; 135g 133; 133, 109; 107
Esters
Bornyl acetate 32.0 210sh C12H20O2 196.29 137m 95, 81
Lavandulyl acetate 31.3 <200 C12H20O2 196.29 137m 95, 81
Linalyl acetate 31.9 <200 C12H20O2 196.29 137m 107, 95, 81
Terpenoid oxides
Caryophyllene oxide 35.2 202 C15H24O 220.35 203d 175, 161, 147, 95
1,8-Cineol 22.1 <200 C10H18O 154.25 137d 95, 81
Limonene oxide 24.0 <200 C10H16O 152.23 135d 107, 93
3112 C. Turek, F.C. Stintzing

Table 1 (continued)

Compounda tR [min] lmax, UV [nm]b Formula Mr [g/mol] APCI(+)-MS m/zmaxc MS2 (m/zmax) m/z

cis-Linalool oxide 10.1 <200 C10H18O2 170.25 153d 135

trans-Linalool oxide 11.2 <200 C10H18O2 170.25 153d 135
Phenols
Carvacrol 22.7 216sh, 278 C10H14O 150.22 149g 135, 121, 109
Eugenol 17.7 202, 226sh, 282 C10H12O2 164.20 163g 131
Thymol 23.8 216sh, 278 C10H14O 150.22 150h 135, 121, 109
Ethers
Anethole 27.8 206, 258, 288sh C10H12O 148.20 149e 121
Thymol methyl ether 33.7 220sh, 274 C11H16O 164.24 165e 149, 123
a
Alphabetical order within particular compound class
b
UV spectra were background-corrected by the software. sh = shoulder
c
For the sake of clarity, only the prominent base peak mass-to-charge ratios are shown. Further details are given in the text section and the corresponding
Online Resources
d
Equivalent to [M−17]+
e
Equivalent to [M+1]+
f
Equivalent to [M+30]+
g
Equivalent to [M−1]+
h
Equivalent to [M]+
i
Equivalent to [M+17]+
j
Equivalent to [M+19]+
k
Equivalent to [M+58]+
l
Equivalent to [M+72]+
m
Equivalent to [M−59]+

Identification of essential oil compounds [15] as well as Hinshaw [16] point out unusual peak forms,
appearance of earlier eluting peaks, or baseline rising due to
Prior to essential oil analyses, each standard compound thermal or catalytic decompositions. Moreover, LC/MS has
(concentrations ranging from 0.1 to 10 mg/mL) was analyzed been applied, among others, for investigation of non-volatiles
to obtain chromatographic and mass spectrometric data. As in citrus oils [e.g., 20, 21] and for separation and identification
the purity of most reference substances was at least 95%, of terpene oxidation products in atmospheric secondary
differentiation between main components and eventual minor organic aerosols [e.g., 41–46]. But, as far as we are aware,
impurities was promptly possible by examining their partic- no coupling of HPLC/DAD to MS in order to characterize
ular 3D UV–vis spectra. On this basis, individual essential oil volatile compounds in essential oils has been published
compounds in real samples were assigned by comparison with hitherto. Therefore, testing of a broad range of analytical
retention times (tR) and UV and mass spectra. settings was required with respect to eluent composition,
gradient design, detection wavelengths, along with different
ionization techniques to optimize both separation and detec-
Results and discussion tion of volatile compounds with a great structural diversity.
Due to their lipophilic character, complete solvation of
Method development essential oils prior to analysis and for primary adsorption on
the stationary phase is mandatory. After extensive testing
Up to the present, HPLC analyses on essential oil volatiles have including different essential oils, reference compounds, and
been scarcely reported [12, 18, 26–39], and comprehensive LC mixtures therefrom, best separation was achieved using water
methods applicable to a greater variety of essential oils are still and acetonitrile as eluents. Methanol instead of acetonitrile led
rare. This is even more remarkable considering that essential to peak broadening not affordable for analysis of complex
oil components have been shown to undergo chemical essential oils. Likewise, acid addition to the aqueous eluent
alterations during GC analysis. Bruhn, for instance, illustrated did not only turn out to be unnecessary for improved
significant changes in GC chromatograms by increasing ionization but even decreased separation capacity (data not
injector temperature from 200 °C to 350 °C [13], and Klee shown). In order to figure out optimum monitoring wave-
HPLC analysis of essential oils 3113

Fig. 1 HPLC fingerprint of 15

Intens. a
essential oil from lavender at [mAU] Intens..
a 220 nm and b 210 nm (dilution [[mAU]] b 7 11 15
800
1:500; for peak assignment, see 1.200
600 20
Table 2) 1.100
22
400
1.000
200
900
0
800 10 20 30 40 Time [min]

700
600 20
500
400
300
7 11 22
200
16
12 14 17 21 23
100 3 45 6 8 9 10 13 18 19
0
10 20 30 40 50 Time [min]

lengths, 3D DAD spectra were recorded for all analytes as its capability was compared with electrospray ionization.
suggested by Villa et al. [27]. Lacking conjugated double Optimization of ionization parameters such as lens tunings,
bonds, absorption maxima were below 200 nm for several voltage adjustments, temperature, and nebulizer regulations
terpenoids. Therefore, wavelengths of 195 [34] or 210 nm by direct syringe injection of standard compounds in both
[27] have been previously applied. In the present study, positive (+) and negative (−) modes manifested that APCI
however, 220 nm turned out to be a suitable monitoring (+)-MS afforded the best sensitivity and detection capacity
wavelength considering both maximum absorption and lowest (data not shown). Though APCI(−)-MS resulted in a high
signal-to-noise ratios of the great variety of compounds in total ion current (TIC) for phenols with intense signals for
essential oils. In addition, 210, 250, and 280 nm were the [M−1]- ion, it did not yield distinct mass spectra for all
simultaneously chosen to cover the specific absorption other compounds analyzed.
properties of other accompanying components. Thus, separa-
tion of the target compounds was possible, but peak identity HPLC/DAD/APCI-MS analysis of common volatile
and purity could also be determined. Since spectroscopic data compounds
were not very specific for some structurally related compo-
nents, additional information was obtained by tandem mass A number of common essential oil components, representing a
spectrometry to compensste for this deficiency. broad structural variety, were subjected to HPLC/DAD/APCI
Though APCI has been indicated as the ionization method (+)-MS measurements. Retention times (tR), UV maxima
of choice for small, non-, or medium-polar molecules [20, 47], (lmax, UV), and the prominent base peak mass-to-charge ratios

Fig. 2 a HPLC fingerprint of

Intens. a
essential oil from thyme at [[mAU] 7
220 nm (dilution 1:4,000; for 600
peak assignment, see Table 3).
550
b Detailed section from 37.5 to Intens. 14
38.8 min; solid line 220 nm, 500 [mAU]
b
scattered line 250 nm, dotted 14 13
450
line 280 nm 12
400 10
350 8
10 6
300
4
250 2
0
200
150
100 6 12
17
50 16
2 5 8 9 15
0
10 20 30 40 50 Time [min]
3114 C. Turek, F.C. Stintzing

Fig. 3 a HPLC fingerprint of Intens. aa

essential oil from rosemary at [mAU] 20
220 nm (dilution 1:500; for peak
assignment, see Table 4). Intens. b 14
350 [mAU] 19
b Detailed section from 36.0 to 200
39.5 min; solid line 220 nm,
300 160
scattered line 250 nm (intensity
×2 magnified), dotted line 120
280 nm (intensity ×2 magnified) 250 80 15 16
13 17 18
40 12
200
0 22

150 10

100

50 21 23
2 3 5 6 8 9 11 24 25
7
0
10 20 30 40 50 Time [min]

(m/zmax) are presented in Table 1. All compounds analyzed noids. But, no general elution order according to chemical
were detectable in TIC mode of APCI(+)-MS and at 220 nm, classes [12, 29] was found (Table 1). Although no universal
except 3-octanone (280 nm), linalool oxide, and borneol. The rules could be drawn between chemical structures and their
two latter displayed little absorption at 210 nm, but reasonable respective fragmentation patterns in MS, similarities were
detection was possible at 190 nm and by MS. Since it appears found for most compounds analyzed. Fragment ions at m/z 81
that a compilation of retention, UV, and MS data is still (C6H9+), 93 (C7H9+), 95 (C7H11+), 107 (C8H11+), 109
missing in the literature, such data have been carefully (C8H13+), and 121 (C9H13+, assigned according to Maleknia
monitored for a comprehensive set of essential oil volatiles. et al. [49]) were very frequent among all classes of
Some reference standards were obtained as mixtures of terpenoids. MS2 mostly produced neutral loss of water and
isomers, e.g., citral, consisting of the cis- and trans-form fragmentation of the hydrocarbon backbone, but significant
neral and geranial, respectively. For peak assignment, the differences in fragment intensities of diastereomers like cis-
cis- were presumed to elute prior to the trans-isomers in trans-isomers could not be observed (Table 1).
RP-HPLC in reference to previously published articles [42,
48] and analogous to analysis of the separately available Alcohols Alcohols typically show strong signals of
isomeric alcohols nerol and geraniol (Table 1). [M+H−H2O]+ [41, 47], corroborating the results for
With the exception of aromatic p-cymene and p-α- terpenoid alcohols of the present study. Only citronellol as
dimethylstyrene, oxygen-containing molecules eluted prior well as lavandulol exhibited [M+H]+ base peak ions. Trace
to hydrocarbons and monoterpenoids preceded sesquiterpe- ions of [2M+H−2H 2O] + and of [M−H] + previously

Fig. 4 a HPLC fingerprint of

Intens. a 19
essential oil from pine at [mAU]
220 nm (dilution 1:500; for
peak assignment, see Table 5). 1.000
b Detailed section from 36.0 to
39.5 min; solid line 220 nm, 900 Intens. b 10
scattered line 250 nm [mAU] 17
800
(intensity ×2 magnified),
400
dotted line 280 nm 700
(intensity ×6 magnified) 300 13 14
600
200 11 12
500 15
9
100 8 16
400
0
300 22

200
18
6
100 1 4 5 20 21 23 24
25
0
10 20 30 40 50 Time [min]
HPLC analysis of essential oils 3115

Fig. 5 HPLC fingerprint of Intens. 10

essential oil from turpentine at [mAU]
220 nm (dilution 1:1,000; for
1.200
peak assignment, see Table 6)
1.100
1.000
900
800
700
600
500
400
300
9
200
6
100 3 45 8 11 12
7
0
10 20 30 40 50 Time [min]

assigned to [M+H−H2]+ by Glasius et al. [41] were also fragment of m/z 109 were registered for verbenone and
detected (see Electronic Supplementary Material Fig. S1 for carvone (Electronic Supplementary Material Fig. S3 for
α-terpineol). verbenone).

Aldehydes The aliphatic aldehydes neral and geranial Terpenic hydrocarbons The ionization behavior of mole-
(Electronic Supplementary Material Fig. S2) showed strong cules depends on their chemical structure as well as on
[M+H−H2O]+ signals accompanied by weaker [M±H]+ the prevailing conditions in the ion source. As solvent
signals, in which the [M+H]+ ion was more pronounced for composition changes in gradient HPLC, ionization
the trans- compared with the cis-isomer (data not shown). conditions and the responses of the eluting compounds
The protonated molecule ion proved to be diagnostic for p- alter during a chromatographic LC/MS run [42]. In the
anisaldehyde, producing an intense [M+H−CO]+ signal in present study, terpenic hydrocarbons eluting under quite
the MS2 mode. nonpolar conditions proved to be the least easy to pin
down as they manifested good ionization in aqueous
Ketones Ketones were characterized by strong [M+H]+ conditions, but changed ionization behavior when ex-
signals. MS2 experiments with camphor and 3-octanone led posed to less polar eluent composition (data not shown).
to neutral loss of water and fragmentation of the hydrocar- While ionization patterns of most of the compounds
bon backbone, while neutral loss of CO and a strong investigated were explicit and coherent, mass spectra of

Fig. 6 a HPLC fingerprint of

essential oil from eucalyptus at
Intens. a
[mAU]
220 nm (dilution 1:100; for peak
assignment, see Table 7). 900
b Detailed section from 36.5 to
800 Intens. b 19
39.5 min; solid line 220 nm, [mAU] 15
scattered line 250 nm, dotted 26
700 700
line 280 nm
600 500 22
17
23
500 300
21 24
18 25
100 20
400
0
300

200

100 5 7 9 16
4 6 8 1011 12 1314 27 28
0
10 20 30 40 50 Time [min]
3116 C. Turek, F.C. Stintzing

Fig. 7 HPLC fingerprint of es-

Intens. 7
sential oil from may chang at [mAU]
220 nm (dilution 1:2,000; for
peak assignment, see Table 8)
700
5

600

500

400

300

200
6 14 -18
100 3 13
8 9 10 11 19 20 21
1 2 4 12
0
10 20 30 40 50 Time [min]

some terpenes turned out to be ambiguous and therefore however, that although several monoterpene isomers
cannot be explained so far (Table 1). It is noteworthy, differ only in the position of one double bond, e.g., α-

Table 2 Compound assignment in essential oil from lavender by HPLC/DAD/APCI(+)-MS analysis

Numbera, b
tR [min] lmax, UV [nm]c APCI(+)MS m/z MS2 m/z Assignment

1 10.0 – 153; 135d 135 cis-Linalool oxide

2 11.2 – 153; 135d 135 trans-Linalool oxide
3 17.1 230 139 121 n.i..
4 19.2 <200 137 107, 95, 81 α-Terpineol
5 20.8 <200 137 107, 95, 81 Terpinene-4-ol
155 137 Lavandulol
6 21.4 <200 137 95, 81 Nerol
7 21.9 <200 137 107, 95, 81 Linalool
8 23.4 <200, 280 129 – 3-Octanone
9 23.9 204, 258 149 – n.i..
10 31.3 <200 137; 197d 95, 81 Lavandulyl acetate
11 32.1 <200 137; 273d 107, 95, 81 Linalyl acetate
12 33.8 212sh, 266, 274sh – – p-Cymene
13 35.3 202 203; 221d 175, 161, 147, 95 Caryophyllene oxide
14 35.8 216 – – n.i..
15 36.8 234, 276sh 153; 135d 135, 109 Ocimene
16 37.1 225 155 137, 109 β-Myrcene
17 38.2 268 153 – α-Terpinene
18 39.0 201 – – β-Pinene
19 40.9 204 – – α-Pinene
20 46.3 222sh 205; 203 161, 159, 147, 145 n.i..
21 47.4 210, 260 205; 203 161, 159, 147, 145 n.i..
22 47.9 <200 234; 203 217; 147, 145 β-Caryophyllene
23 49.4 <200 234 – n.i..

n.i. not identified, – no signal available, sh shoulder

a
Dilution 1:500 in ethanol prior to analysis
b
Peak assignment refers to Fig. 1
c
UV spectra were background-corrected by the software
d
Auxiliary diagnostic m/z
HPLC analysis of essential oils 3117

pinene/β-pinene or terpinolene/α-phellandrene/α-terpinene/ γ-terpinene and a strong signal in limonene, α-phellandrene

γ-terpinene, their particular UV data together with mass and terpinolene, while β-myrcene showed [M+H+H2O]+
spectral pattern revealed distinct differences. [M+H]+ ions base peak ions as described by Aznar et al. [50] (Table 1). In
were present for all hydrocarbons analyzed and for each case, MS2 experiments led to neutral loss of water.
camphene, sabinene, β-pinene, as well as the aromatic Signals derived from [M+H−H2]+ and [M]+ ions were often
compounds p-cymene and p-α-dimethylstyrene, a character- present in ion clusters of ±1 u in altering ratios (data not
istic [M+H]+ base peak was detected. α-Pinene provided shown). As depicted in Table 1, ionization of some
strong signals of the protonated molecule at m/z 137 but also monoterpenes under the conditions applied led to several
exhibited intense ions of m/z 166, both signals alternating as signals of low and ambiguous intensities. In order to improve
base peaks. The same phenomenon was true for β- ionization, methanol, water, and 5% (v/v) aqueous formic
caryophyllene with m/z 234 and m/z 203 (Table 1). Accom- acid were added to the LC flow via a T-connection outside of
panying signals of ions 30 u higher than M+ were also found the ion source at a velocity of up to 53 μL/min. No better
for limonene and β-myrcene (data not shown) and yielded ionization, however, was achieved. In further experiments,
the base peak ion in δ-3-carene (Electronic Supplementary corona current was reduced to 500 nA which led to a more
Material Fig. S4). This mass shift of 30 u, previously distinct [M+H]+ signal for δ-3-carene, α-phellandrene, and
assigned to the formation of NO+ adducts [50, 51] has been geranial. But, at the same time, signal intensity and peak
earlier reported in APCI-MS analysis of limonene [50]. form quality decreased in most cases (data not shown).
Terpinolene as well as γ-terpinene yielded signals at m/z Although for some terpenes indistinct and ambiguous mass
167, most probably the analogous protonated adduct ion spectra need to be accepted, several isobaric compounds can
[MHNO]+. Further mass shifts starting from M+ of 58 u be distinguished by their particular mass spectra. Moreover,
(resulting in m/z 194 as was the case for camphene, lack in proper MS differentiation may be compensated by
δ-3-carene, limonene, β-myrcene, α-phellandrene, β- retention as well as UV data.
pinene, sabinene, and α- and γ-terpinene) and 72 u (resulting
in m/z 208 for β-myrcene, α-phellandrene, terpinolene, and Esters In agreement with the ejection of an acetic acid
β- and α-pinene) were characteristic for several monoter- molecule after protonation (Table 1) as suggested previous-
penes (data not shown), but no explanation can be given ly [41, 51], acetyl esters are marked by strong [M−O(CO)
about their formation so far. MS2 experiments on these CH3]+ base peak signals. Lavandulyl acetate also exhibited
adducts only produced ambiguous fragments. Furthermore, a distinct signal of the quasimolecular ion [M+H]+
addition of 17 u, probably due to [M+H+H2O−H2]+ ions, (Electronic Supplementary Material Fig. S5). Linalyl
was frequently encountered in mass spectra of volatile acetate showed trace signals of m/z 273 (tentatively
hydrocarbons and formed the base peak in ocimene, α- and assigned as [2(M−HOOCCH3)+H]+) which were also found

Table 3 Compound assignment

in essential oil from thyme by Numbera, b
tR [min] lmax, UV [nm]c APCI(+)MS m/z MS2 m/z Assignment
HPLC/DAD/APCI(+)-MS
analysis 1 19.1 – 137 95, 81 α-Terpineol
2 20.3 236 151; 109d 123, 109 Carvone
3 20.8 – 153 135, 109, 95 Camphor
4 20.8 – 137 107, 95, 81 Terpinene-4-ol
5 21.9 <200 137 107, 95, 81 Linalool, 1,8-cineol
6 22.8 216sh, 278 149 135, 121 Carvacrol
7 23.6 216sh, 278 150; 135d 135, 121, 109 Thymol
8 28.1 206, 240sh – – n.i..
9 33.1 218sh, 275 – – n.i..
10 33.7 212sh, 266, 274sh 135 – p-Cymene
n.i. not identified, – no signal 11 35.2 – 203; 221d 175, 161, 147 Caryophyllene oxide
available, sh shoulder 12 36.9 224 155 – β-Myrcene
a
Dilution 1:4,000 in ethanol prior 13 38.1 268 153 – α-Terpinene
to analysis 14 38.3 <200 153 – γ-Terpinene
b
Peak assignment refers to Fig. 2 15 38.9 200 – – β-Pinene
c
UV spectra were background- 16 40.7 206 – – α-Pinene
corrected by the software
d
17 47.8 <200 234; 203 217, 161; 147 β-Caryophyllene
Auxiliary diagnostic m/z
3118 C. Turek, F.C. Stintzing

for the corresponding alcohol linalool (data not shown). An MS2 experiments led to an ion signal at m/z 135 for both
m/z 273 can therefore be used to distinguish linalyl acetate thymol and carvacrol which was tentatively assigned to
from the coeluting bornyl acetate. [M+3H−H2O]+. Eugenol provided a distinct MS2 fragment
at m/z 131, probably due to ejection of the protonated
Terpenoid oxides Oxides all yielded strong base peak signals methoxy group. Rearrangement loss of CO as described by
due to [M+H−H2O]+ (Table 1), accompanied by smaller Holčapek et al. [47] could not be detected.
amounts of [M+H]+ ions for caryophyllene oxide and
limonene oxide (data not shown) and even [M+H−2H2O]+ Ethers The base peaks of the methyl ethers analyzed
in the case of linalool oxide, a molecule exhibiting two were characterized by their quasimolecular [M+H]+ ions
oxygen atoms (Electronic Supplementary Material Fig. S6). (Table 1). Thymol methyl ether showed further strong
MS2 produced fragmentation of the hydrocarbon backbone [M]+ signals (m/z 164), m/z 149 (tentatively assigned as
and neutral loss of water in the case of linalool oxide. [M+3H−H2O]+ or [M−CH3]+•), and m/z 123, a fragment
that cannot be explained so far. Anethole provided weaker
Phenols All phenols analyzed are characterized by ion signals of m/z 190 (Electronic Supplementary Material
clusters of ±1 u around M+ with the highest intensity of Fig. S8), possibly the acetonitrile adduct of the protonated
[M]+ for thymol (Electronic Supplementary Material Fig. S7) molecule [M+H+CH3CN]+, as described earlier [47, 52].
and [M−H]+ in the case of carvacrol and eugenol (Table 1). MS2 from the base peak of anethole produced a strong

Table 4 Compound assignment in essential oil from rosemary by HPLC/DAD/APCI(+)-MS analysis

Numbera, b
tR [min] lmax, UV [nm]c APCI(+)MS m/z MS2 m/z Assignment

1 14.2 – 151; 109d 123, 109 Verbenone

2 17.8 200, 226sh, 282 163; 131d 131 Eugenol
3 19.2 <200 137 107, 95, 81 α-Terpineol
4 20.0 – 137 95, 81 Borneol
5 20.8 <200, 288 153 135, 109, 95 Camphor
6 21.9 <200 137 95, 81 Linalool
7 22.2 <200 137 95, 81 1,8-Cineol
8 31.7 206, 249 – – p-α-Dimethylstyrene
9 32.2 208sh 137 95, 81 Bornyl acetate
10 33.8 212sh, 266 135 – p-Cymene
11 35.3 202 203; 221d 147 Caryophyllene oxide
12 36.3 242 – – n.i..
13 36.8 234, 276sh 153; 135d – Ocimene
14 37.0 224 155 137, 109 β-Myrcene
15 38.2 268 153 135, 109, 95 α-Terpinene
16 38.4 232 – – n.i..
17 38.7 <200 153; 137; 135; 194d – Limonene, camphene
18 38.7 164 194; 153; 135; 166d – α-Phellandrene
19 39.0 200 137 107, 93, 81 β-Pinene
20 40.9 204 166; 137 – α-Pinene
21 46.9 <200 221; 205; 203 – n.i..
22 48.0 <200 234; 203 217, 216, 161; 147 β-Caryophyllene
23 49.0 <200 203; 221d – n.i.
24 50.1 210 – – n.i.
25 51.9 206 204 – n.i.

n.i. not identified, – no signal available, sh shoulder

a
Dilution 1:500 in ethanol prior to analysis
b
Peak assignment refers to Fig. 3
c
UV spectra were background-corrected by the software
d
Auxiliary diagnostic m/z
HPLC analysis of essential oils 3119

Table 5 Compound assignment in essential oil from pine by HPLC/DAD/APCI(+)-MS analysis

Numbera, b
tR [min] lmax, UV [nm]c APCI(+)MS m/z MS2 m/z Assignment

1 14.1 258 151; 109d 123, 109 Verbenone

2 19.3 – 137 107, 95, 81 α-Terpineol
3 20.0 – 137 81 Borneol
4 31.7 204, 248 – – p-α-Dimethylstyrene
5 32.1 210sh 137 95, 81 Bornyl acetate
6 33.8 212sh, 266, 274sh 135 – p-Cymene
7 35.2 – 203; 221d 147 Caryophyllene oxide
8 36.3 240 – – n.i.
9 36.8 211, 278, 268sh, 288sh 137 – n.i.
10 37.0 224 155 137, 109 β-Myrcene
11 37.3 208, 268, 260sh, 278sh – – n.i.
12 38.2 268 153 109 α-Terpinene
13 38.4 <200 208; 153; 135; 167d 133, 109; 107 Terpinolene
14 38.5 232 208; 153; 137; 135 n.i.
15 38.7 <200 153; 137; 135; 166d – Limonene, camphene
16 38.7 263 194; 153; 135; 166d – α-Phellandrene
17 39.0 200 137 107, 81 β-Pinene
18 39.8 <200 166; 137d 123 δ-3-Carene
19 40.9 206 166; 137 – α-Pinene
20 46.3 222sh – – n.i.
21 46.8 <200 221; 205; 203 – n.i.
22 47.9 <200 234; 203 217, 161; 147, 145 β-Caryophyllene
23 50.1 201 234; 203d 217 n.i.
24 51.0 203 – – n.i.
25 51.8 205 204 – n.i.

n.i. not identified, – no signal available, sh shoulder

a
Dilution 1:500 in ethanol prior to analysis
b
Peak assignment refers to Fig. 4
c
UV spectra were background-corrected by the software
d
Auxiliary diagnostic m/z

signal at m/z 121. According to Holčapek et al. [47], for methyl ethers, this fragment can be tentatively
reporting neutral loss of CH3O• as possible fragmentation assigned to [M+4H−CH3O]+•.

Table 6 Compound assignment

in essential oil from turpentine Numbera, b
tR [min] lmax, UV [nm]c APCI(+)MS m/z MS2 m/z Assignment
by HPLC/DAD/APCI(+)-MS
analysis 1 14.1 – 151; 109d 123, 109 Verbenone
2 19.3 – 137 107, 95, 81 α-Terpineol
3 33.8 212sh, 266 – – p-Cymene
4 35.3 202 203; 221d 147 Caryophyllene oxide
5 36.3 244 – – n.i.
n.i. not identified, – no signal 6 37.0 224 155 – β-Myrcene
available, sh shoulder 7 38.4 233 – – n.i.
a
Dilution 1:1,000 in ethanol prior 8 38.7 <200 153; 194 d
– Limonene, camphene
to analysis 9 39.0 200 137 107, 81 β-Pinene
b
Peak assignment refers to Fig. 5 10 40.9 205 166; 137 – α-Pinene
c
UV spectra were background- 11 46.8 <200 – – n.i.
corrected by the software
d
12 47.8 <200 234; 203 – β-Caryophyllene
Auxiliary diagnostic m/z
3120 C. Turek, F.C. Stintzing

Characterization of essential oils by HPLC/DAD/APCI-MS with the help of unlike lmax, UV (e.g., α-phellandrene and
analyses limonene, Tables 4, 5, and 7) or by means of characteristic
base peak ions, e.g., camphor and terpinene-4-ol in thyme oil
The separation capacity of the HPLC/DAD/APCI(+)-MS (Table 3). Still, few substances are chemically too much alike
method established was applied to the analysis of seven to be distinguishable (e.g., 1,8-cineol and linalool in thyme
commercially available essential oils originating from different oil, Table 3). Characteristic fingerprints of the essential oils
plant genera (Figs. 1–7). By comparison with retention, MS analyzed are presented in Figs. 1–7 with identical time
and UV data of reference standards, all typical components of frames to facilitate direct profile comparison. Monitoring at
the particular essential oil could be identified (Tables 2–8). 220 nm turned out to be best, although simultaneous
Given that essential oils are composed of a multitude of detection at further wavelengths (210, 250, and 280 nm)
chemically related substances, particular attention was paid to provided additional information about peak identity and
potential coelution events. As a matter of fact, several possible coelution events, as for α- and γ-terpinene in thyme
simultaneously eluting volatiles can be differentiated either oil (Fig. 2b).

Table 7 Compound assignment in essential oil from eucalyptus by HPLC/DAD/APCI(+)-MS analysis

Numbera, b
tR [min] lmax, UV [nm]c APCI(+)MS m/z MS2 m/z Assignment

1 10.1 – 153; 135d 135 cis-Linalool oxide

2 11.2 – 153; 135d 135 trans-Linalool oxide
3 14.0 – 151; 109d 123, 109 Verbenone
4 18.1 <200 135 107, 93 cis-Carveol
5 19.2 <200 137 107, 95, 81 α-Terpineol
6 19.9 233 153 – n.i.
7 20.3 238 151; 109d – Carvone
8 20.8 <200 137 107, 95, 81 Terpinene-4-ol
9 21.5 <200 137 95, 81 Nerol
10 22.1 <200 137 95, 81 1,8-Cineol
11 23.5 <200, 241 135 107, 93 Neral
12 24.4 <200, 241 135; 153d 107, 93 Geranial
13 31.1 231 – – n.i.
14 31.7 205, 249 133 – p-α-Dimethylstyrene
15 33.8 212sh, 266, 274sh 135 107, 93 p-Cymene
16 35.1 232, 268sh – – n.i.
17 36.8 206 137 – Sabinene
18 36.9 234, 276sh 153; 135d – Ocimene
19 37.0 224 155 137, 109 β-Myrcene
20 38.2 267 153 135, 109 α-Terpinene
21 38.4 232 208; 194; 153; 135 – n.i.
22 38.4 <200 208; 153; 135; 194d; 167d – γ-Terpinene, terpinolene
23 38.6 <200 153; 137; 135; 194d; 166d – Camphene, limonene
24 38.7 264 194; 153; 135; 166d; 137d – α-Phellandrene
25 39.0 200 137 107, 93, 81 β-Pinene
26 40.8 205 166; 137 – α-Pinene
27 46.2 222sh 205; 203 – n.i.
28 47.8 <200 – – β-Caryophyllene

n.i. not identified, – no signal available, sh shoulder

a
Dilution 1:100 in ethanol prior to analysis
b
Peak assignment refers to Fig. 6
c
UV spectra were background-corrected by the software
d
Auxiliary diagnostic m/z
HPLC analysis of essential oils 3121

Table 8 Compound assignment

in essential oil from may chang Numbera, b
tR [min] lmax, UV [nm]c APCI(+)MS m/z MS2 m/z Assignment
by HPLC/DAD/APCI(+)-MS
analysis 1 18.5 <200 109; 127d 81, 68 n.i.
2 21.4 <200 137 95, 81 Nerol
3 21.5 <200 137 95, 81 Geraniol
4 21.9 <200 137 – Linalool
5 23.4 241 135 107, 93 Neral
6 24.0 <200 157 – β-Citronellol
7 24.3 241 135; 153 d 107, 93 Geranial
8 33.8 212sh, 266 – – p-Cymene
9 36.8 206 137 – Sabinene
10 37.1 224 155 – β-Myrcene
11 38.7 <200 153; 135 135, 107; 105 Limonene
12 39.0 200 – – β-Pinene
13 40.9 206 – – α-Pinene
14 43.7 296 269 227, 213 n.i.
n.i. not identified, – no signal 15 43.9 296 287 269, 109 n.i.
available, sh shoulder 16 44.3 216sh, 300 269 227, 213 n.i.
a
Dilution 1:2,000 in ethanol prior 17 44.6 216sh, 300 287 269, 205, 109 n.i.
to analysis 18 45.0 <200 287 269 n.i.
b
Peak assignment refers to Fig. 7 19 46.4 220sh, 292 289 271, 163 n.i.
c
UV spectra were background- 20 46.7 214sh, 258sh 289 271, 231 n.i.
corrected by the software
d
21 47.9 <200 234; 203 – β-Caryophyllene
Auxiliary diagnostic m/z

Lavender oil Because compounds diverge in their absorp- camphor and terpinene-4-ol by APCI(+)-MS analysis. The
tion behavior, response factors at specific wavelength may identity of 5 (Fig. 2, Table 3) could not be clearly assigned
differ considerably. This can be clearly seen in the because linalool and 1,8-cineol showed neither differences in
fingerprint of lavender oil (Fig. 1). The absorption UV nor MS data. Nevertheless, ratio determination of peak
coefficients of the two main components linalool (7) and intensity in UV (220 nm) compared with TIC signal intensity
linalyl acetate (11) at 220 nm are lower than those of in APCI(+)-MS chromatograms, remarkably higher for
ocimene (15) coeluting with β-myrcene (16), whose linalool than for 1,8-cineol, militated in favor of linalool.
lmax, UV are close to 220 nm. Therefore, the profile at
210 nm is presented as well (Fig. 1b), as it provides a better Rosemary oil The HPLC fingerprint of rosemary oil at
understanding of lavender oil composition. Table 2 presents 220 nm (Fig. 3) showed peak splitting at tR of linalool and
HPLC/DAD data and characteristic ions from compounds 1,8-cineol, indicating that both components were present.
detected in essential oil of lavender. Although some peak Furthermore, all characteristic compounds usually found in
identities yet remain unknown, all typical components essential oil of rosemary were identified by HPLC/DAD/
could be assigned by comparison with standards. Mass APCI(+)-MS analysis (Table 4). No comparative reference
spectra and tR of 20–21 and 23 were suggestive of compound could be found for six peaks, but four of them
sesquiterpenes and 14 of a sesquiterpenoid alcohol, whereas (21, 23–25, Table 4) could be tentatively assigned as
no assumption was possible in the case of 3 and 9 (Table 2). sesquiterpenes because of their high tR and characteristic
In HPLC/DAD analysis, separation of terpinene-4-ol and ions. Monitoring of different wavelengths enabled detection
lavandulol could not be achieved, however, differentiation of 12 and α-terpinene (Fig. 3, Table 4) at 250 and 280 nm,
was possible due to distinct ions in APCI(+)-MS analysis. respectively, and revealed coelution of α-phellandrene with
Furthermore, β-myrcene was not completely separated limonene and/or camphene. APCI(+)-MS analysis of the
from ocimene, but unlike lmax, UV as well as base peak peak at tR =38.7 min provided ions characteristic for a couple
ions revealed their degree of coelution. of monoterpenes but did not yield compound-specific data.

Thyme oil A chromatogram (Fig. 2) and compound table of Pine oil Comparable to rosemary oil, essential oil of pine
thyme oil (Table 3) is presented. Strong dilution was required (Fig. 4, Table 5) contains α-terpinene, α-phellandrene,
due to the intense thymol signal in HPLC/DAD. Although limonene, and/or camphene. Furthermore, detection at
compounds 3 and 4 co-eluted, they could be identified as 280 nm revealed two yet unidentified substances 9 and 11
3122 C. Turek, F.C. Stintzing

(Table 5) with similar UV spectra. Terpinolene could be To evaluate the applicability of the method to real
identified as shoulder peak of 14 (Table 5), a compound samples, several essential oils deriving from diverse plant
with lmax, UV =232 nm also present in rosemary essential genera were analyzed. Although a few substances were
oil (16, Table 4). Compounds characterizing pine oil could structurally too similar to be chromatographically different,
all be detected and assigned by HPLC/DAD/APCI(+)-MS an overall high level of separation was achieved. In
analysis as presented in Table 5. In addition, a couple of addition to the characteristic components of the particular
further substances, probably sesquiterpenes in consideration oil, minor compounds could also be determined, among
of their tR and mass spectra were noticed, some of them them even yet unidentified substances.
also occurring in lavender or rosemary oil. It is therefore suggested that HPLC should be more often
considered as a versatile method for essential oil analyses
Turpentine oil Compound 21 (Table 5) at tR =46.8 min is complementary to GC in the future. Providing a suitable tool
also present in essential oil from turpentine (11 in Table 6). especially for the analysis of water-containing preparations like
The HPLC fingerprint of turpentine oil at 220 nm, which is cosmetics, time-consuming processing essential for GC may
marked by a strong signal of α-pinene (10, Table 6), is thus be skipped. Moreover, possible rearrangements of labile
presented in Fig. 5. substances in GC due to high temperatures or catalytically
active surfaces can be avoided. Continuative investigations are
Eucalyptus oil Essential oil from eucalyptus (Fig. 6, Table 7) currently under way to expand the data basis to other essential
showed a strong MS signal of 1,8-cineol (10, Table 7), while oils, essential oil mixtures, as well as pharmaceutical prepara-
UV absorption was very low above 190 nm. Again, virtually tions and cosmetics including compound quantification.
all major characteristic components as well as minor
components could be identified. Moreover, five compounds Acknowledgments The authors wish to thank Simone Ritter,
Nadine Kirschmann, and Julia Bertrams for assistance in sample
remained unidentified, among them a tentative sesquiterpene preparation and trouble shooting.
(27, Table 7) already detected in lavender (20, Table 2) and
pine (20, Table 5) oil.
References
May chang oil The fingerprint of may chang oil (Fig. 7)
was dominated by huge signals of neral and geranial (5 and 1. Scott RPW (2005) In: Worsfold P, Townshend A, Poole C (eds)
7, Table 8). Apart from one early eluting substance (1, Encyc Anal Sci, 2nd edn. Elsevier, Amsterdam
Table 8) and a couple of unknown components at tR =43.7– 2. Grassmann J, Elstner EF (2003) In: Caballero B, Trugo L, Finglas P
46.7 min (14–20, Table 8), compounds could be clearly (eds) Encyc Food Sci Nutr, 2nd edn. Elsevier, Amsterdam
3. Burt S (2004) Essential oils: their antibacterial properties and
identified. Base peak signals of m/z 269, 287, and 289, lmax, potential applications in foods—a review. Int J Food Microbiol
UV from 258 to 300 nm (except 18) and the elution between 94:223–253
α-pinene and β-caryophyllene are suggestive of polyunsat- 4. Bakkali F, Averbeck S, Averbeck D, Idaomar M (2008) Biological
urated, oxygen-containing diterpenes. effects of essential oil—a review. Food Chem Toxicol 46:446–475
5. Cavanagh HMA (2007) Antifungal activity of the volatile phase
of essential oils: a brief review. Nat Prod Commun 2:1297–1302
6. Ntalli NG, Ferrari F, Giannakou I, Menkissoglu-Spiroudi U
Conclusions (2010) Phytochemistry and nematicidal activity of the essential
oils from 8 greek lamiaceae aromatic plants and 13 terpene
components. J Agric Food Chem 58:7856–7863
An HPLC/DAD/APCI-MS method has been developed 7. Podsędek A, Sikora M, Kalemba D (2009) The determination of
suitable to detect essential oil components of diverse antioxidant activity of essential oils. SOFW J 135:2–5
chemical classes. After extensive optimization of LC and 8. Zygadlo JA, Juliani HR Jr (2000) Bioactivity of essential oil
MS parameters, UV and MS data of a broad range of components. Curr Top Phytochem 3:203–214
9. Lockwood GB (2001) Review. Techniques for gas chromatogra-
representative reference compounds were compiled. phy of volatile terpenoids from a range of matrices. J Chromatogr
Simultaneous monitoring at different wavelengths pro- A 936:23–31
vided information of peak identities as well as of 10. Bicchi C, Joulain D (1990) Headspace-gas chromatographic
possible coelution events. Although many terpenoids analysis of medicinal and aromatic plants and flowers. Flavour
Fragr J 5:131–145
possess isobaric structures, differentiation was often 11. Franz CM (2010) Essential oil research: past, present and future.
achieved by careful inspection of dissimilar retention, Flavour Fragr J 25:112–113
UV absorption, or ionization behavior. APCI(+)-MS 12. Kovar K-A, Friess D (1980) Determination of essential oils in
analyses showed distinct base peak ions and characteris- mixtures—HPLC in comparison to GC. Arch Pharm 313:416–428
13. Bruhn W (1974) Zur gaschromatographischen Analyse ätherischer
tic fragments for most volatiles, however, formation of
Öle: Naturstoffe oder Artefakte? Dragoco Rep 21:115–121
some ions, mainly from nonpolar terpenes cannot be 14. Kallen J, Heilbronner E (1960) Das Gas-Chromatogramm einer
explained so far. labilen Verbindung (System: A→B). Helv Chim Acta 43:489–500
HPLC analysis of essential oils 3123

15. Klee M (2010) Hot split injections, part 3—decomposition. Sep mixture of blackcurrant buds (Ribes nigrum L.). J Chromatogr
Sci GC Solut, Nantwich, Issue 7: 2–6. http://e-learn.sepscience. A 665:105–109
com/gcsolutions/ Accessed 13 Jan 2011 36. Ross MSF (1976) Analysis of cinnamon oils by high-pressure
16. Hinshaw JV (2002) Nonlinearity in a chromatographic system. liquid chromatography. J Chromatogr 118:273–275
LCGC North Am 20:120–126 37. Scott RPW, Kucera P (1979) Mode of operation and performance
17. Wylie PL, Klein KJ, Thompson MQ, Hermann BW (1992) Using characteristics of microbore columns for use in liquid chromatog-
electronic pressure programming to reduce the decomposition of raphy. J Chromatogr 169:51–72
labile compounds during splitless injection. J High Res Chroma- 38. Strack D, Proksch P, Gülz P-G (1980) Reversed phase high
togr 15:763–768 performance liquid chromatography of essential oils. Z Naturforsch
18. Da S, Rauber C, Guterres SS, Schapoval EES (2005) LC C/J Biosci 35:675–681
determination of citral in Cymbopogon citratus volatile oil. J 39. Yang D-Q, Yao C-F, Zeng H-P, Lai X-P, Yang T, Chen D-Y
Pharm Biomed Anal 37:597–601 (2002) Determination of the contents of Saussurea lappa essential
19. Frérot E, Decorzant E (2004) Quantification of total furocoumar- oil in Chinese traditional medicine Po Chai pills by HPLC.
ins in citrus oils by HPLC coupled with UV, fluorescence, and Huanan Shifan Daxue Xuebao, Ziran Kexueban 22–24
mass detection. J Agric Food Chem 52:6879–6886 40. Beckert R, Fanghänel E, Habicher WD, Knölker H-J, Metz P,
20. Dugo P, Mondello L, Dugo L, Stancanelli R, Dugo G (2000) LC- Schwetlick K (2009) Organikum, 23rd edn. Wiley, Weinheim
MS for the identification of oxygen heterocyclic compounds in 41. Glasius M, Duane M, Larsen BR (1999) Determination of polar
citrus essential oils. J Pharm Biomed Anal 24:147–154 terpene oxidation products in aerosols by liquid chromatography-
21. Dugo G, Tranchida PQ, Cotroneo A, Dugo P, Bonaccorsi I, ion trap mass spectrometry. J Chromatogr A 833:121–135
Marriott P, Shellie R, Mondello L (2005) Advanced and 42. Warnke J, Bandur R, Hoffmann T (2006) Capillary-HPLC-ESI-
innovative chromatographic techniques for the study of citrus MS/MS method for the determination of acidic products from the
essential oils. Flavour Fragr J 20:249–264 oxidation of monoterpenes in atmospheric aerosol samples. Anal
22. Buiarelli F, Cartoni GP, Coccioli F, Leone T (1996) Analysis of Bioanal Chem 385:34–45
bitter essential oils from orange and grapefruit by high- 43. Müller L, Reinnig M-C, Hayen H, Hoffmann T (2009) Charac-
performance liquid chromatography with microbore columns. J terization of oligomeric compounds in secondary organic aerosol
Chromatogr A 730:9–16 using liquid chromatography coupled to electrospray ionization
23. Nilsson U, Bergh M, Shao LP, Karlberg A-T (1996) Analysis of Fourier transform ion cyclotron resonance mass spectrometry.
contact allergenic compounds in oxidized d-limonene. Chroma- Rapid Commun Mass Spectrom 23:971–979
tographia 42:199–205 44. Larsen BR, Di Bella D, Glasius M, Winterhalter R, Jensen NR,
24. Schieberle P, Grosch W (1988) Identification of potent flavor Hjorth J (2001) Gas-phase OH oxidation of monoterpenes:
compounds formed in an aqueous lemon oil/citric acid emulsion. J gaseous and particulate products. J Atmos Chem 38:231–276
Agric Food Chem 36:797–800 45. Claeys M, Iinuma Y, Szmigielski R, Surratt JD, Blockhuys F, Van
25. Mondello L, Dugo P, Dugo G, Lewis AC, Bartle KD (1999) High- Alsenoy C, Böge O, Sierau B, Gómez-González Y, Vermeylen R,
performance liquid chromatography coupled on-line with high Van der Veken P, Shahgholi M, Chan AWH, Herrmann H,
resolution gas chromatography. J Chromatogr A 842:373–390 Seinfeld JH, Maenhaut W (2009) Terpenylic acid and related
26. Solinas V, Gessa C (1981) High-performance liquid chromato- compounds from the oxidation of α-pinene: Implications for new
graphic analysis of carvacrol and thymol in the essential oil of particle formation und growth above forests. Environ Sci Technol
Thymus capitatus. J Chromatogr 219:332–337 43:6976–6982
27. Villa C, Gambaro R, Mariani E, Dorato S (2007) High- 46. Winterhalter R, Herrmann F, Kanawati B, Nguyen TL, Peeters J,
performance liquid chromatographic method for the simultaneous Vereecken L, Moortgat GK (2009) The gas-phase ozonolysis of
determination of 24 fragrance allergens to study scented products. β-caryophyllene (C15H24). Part I: an experimental study. Phys
J Pharm Biomed Anal 44:755–762 Chem Chem Phys 11:4152–4172
28. Wu S, Lu G, Zhao L, Chen S, Zhou H, Li Q (2010) Determination 47. Holčapek M, Jirásko R, Lísa M (2010) Basic rules for the
of β-elemene in volatile oil of fresh Sarcandra glabra (Thunb.) interpretation of atmospheric pressure ionization mass spectra of
Nakai by HPLC. Zhongguo Xiandai Yingyong Yaoxue 27:59–61 small molecules. J Chromatogr A 1217:3908–3921
29. Benincasa M, Buiarelli F, Cartoni GP, Coccioli F (1990) Analysis 48. Sadilova E, Stintzing FC, Carle R (2006) Anthocyanins, colour
of lemon and bergamot essential oils by HPLC with microbore and antioxidant properties of eggplant (Solanum melongena L.)
columns. Chromatographia 30:271–276 and violet pepper (Capsicum annuum L.) peel extracts. Z
30. Hudaib M, Bellardi MG, Rubies-Autonell C, Fiori J, Cavrini V Naturforsch C/J Biosci 61:527–535
(2001) Chromatographic (GC-MS, HPLC) and virological evalu- 49. Maleknia SD, Bell TL, Adams MA (2007) PTR-MS analysis of
ations of Salvia sclarea infected by BBWV-I. Il Farm 56:219–227 reference and plant-emitted volatile organic compounds. Int J
31. Komae H, Hayashi N (1975) Separation of essential oils by liquid Mass Spectrom 262:203–210
chromatography. J Chromatogr 114:258–260 50. Aznar M, Tsachaki M, Linforth RST, Ferreira V, Taylor AJ
32. Kovar K-A, Bock E (1983) Quantitative determination of a (2004) Headspace analysis of volatile organic compounds from
mixture of extracts of drugs from volatile oils by high- ethanolic systems by direct APCI-MS. Int J Mass Spectrom
performance liquid chromatography. J Chromatogr 262:285–291 239:17–25
33. McKone HT (1979) High performance liquid chromatography of 51. Amelynck C, Schoon N, Kuppens T, Bultinck P, Arijs E (2005) A
essential oils. J Chem Educ 56:698 selected ion flow tube study of the reactions of H3O+, NO+ and O2•+
34. Ogawa CA, Diagone CA, Lanças FM (2002) Separation of with some oxygenated biogenic volatile organic compounds. Int J
monoterpenes in orange essential oil by capillary liquid chroma- Mass Spectrom 247:1–9
tography and micellar electrokinetic chromatography. J Liq 52. Zhou S, Hamburger M (1996) Application of liquid chromatography-
Chromatogr Relat Technol 25:1651–1659 atmospheric pressure ionization mass spectrometry in natural product
35. Píry J, Príbela A (1994) Application of high-performance analysis. Evaluation and optimization of electrospray and heated
liquid chromatography to the analysis of the complex volatile nebulizer interfaces. J Chromatogr A 755:189–204