Comprehensive Two-Dimensional Gas Chromatography/ Mass Spectrometric Analysis of Pepper Volatiles

RAPID COMMUNICATIONS IN MASS SPECTROMETRY
Rapid Commun. Mass Spectrom. 2006; 20: 28232836

Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rcm.2665
Comprehensive two-dimensional gas chromatography/

mass spectrometric analysis of pepper volatiles
Z. L. Cardeal1*, M. D. R. Gomes da Silva2 and P. J. Marriott3**
1
Chemistry Department, ICEx, Federal University of Minas Gerais, Avenida Antonio Carlos 6627, 31270-901 Belo Horizonte, MG, Brazil
REQUIMTE, Departamento de Qumica, Faculdade de Ciencias e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal
3
Australian Centre for Research on Separation Science, RMIT University, Department of Applied Sciences, GPO Box 2476V Melbourne 3001,
Australia
2
Received 13 April 2006; Revised 12 July 2006; Accepted 12 July 2006
The headspace compositions of 13 pepper and peppercorn samples of different species, colloquially
also referred to as pepper, were analyzed, and more than 300 compounds were tentatively characterized by means of comprehensive two-dimensional gas chromatography in tandem with flame
ionization detection, quadrupole mass spectrometric detection and time-of-flight mass spectrometric
detection (GC T GC-FID, GC T GC/qMS and GC T GC/TOFMS, respectively). The analysis of
volatile organic compounds (VOCs) was performed after solid-phase microextraction (SPME) using
a 75-mm PDMS/DVB fibre. Fingerprint comparison between the three techniques permitted peaks to
be assigned in the GC T GC-FID experiment based on the analogous MS analysis, taking into account
retention shifts arising from method variations.
When using GC T GC/TOFMS, about five times more peaks were identified than in GC T GC/qMS.
Retention indices for all peaks were calculated in the bi-dimensional column set comprising of a 5%
phenyl polysilphenylene-siloxane primary column and a polyethylene glycol second column. The
spectra obtained by both mass detection techniques (qMS and TOFMS) give very similar results
when spectral library searching was performed. The majority of the identified compounds eluted as
pure components as a result of high-resolution GC T GC separations, which significantly reduces coelution, and therefore increases the likelihood that pure spectra can be obtained. The differences
between TOFMS and qMS (in fast scanning mode) spectra were generally small. Whilst spectral
quality and relative ion ratios across a narrow peak (e.g. wb 100150 ms) do vary more for the fast
peaks obtained in GC T GC/qMS operation, than with TOFMS, in general adequate spectral
matching with the library can be achieved. Copyright # 2006 John Wiley & Sons, Ltd.
Pepper originated in the monsoon forests of the Malabar
Coast in southwest India, was first brought to Europe via
Arab traders to Greece, and then quickly brought to other
countries. The Romans used pepper to season food and
preserve meat. It has been used in trading as an exchange
medium in a similar manner to coinage and, in classical
times, tributes were paid in pepper; both Attila the Hun and
Alaric I the Visigoth demanded pepper as a substantial part
of Romes ransom. More than 2000 species of peppers are
used, with uses extending to traditional medicine as an
antibacterial, especially to heal wounds. The volatile organic
compound (VOC) fraction of the Piper species has previously
been isolated and more than 200 compounds were identified
after hydrodistillation,16 liquid-liquid extraction,7,8 headspace sampling of pepper/water mixtures by static headspace (HS),9,10 dynamic headspace sampling (purge and
*Correspondence to: Z. L. Cardeal, Chemistry Department, ICEx,
Federal University of Minas Gerais, Avenida Antonio Carlos
6627, 31270-901 Belo Horizonte, MG, Brazil.
E-mail: zenilda@ufmg.br
**Correspondence to: P. J. Marriott, Australian Centre for Research
on Separation Science, RMIT University, Department of Applied
Sciences, GPO Box 2476V Melbourne 3001, Australia.
E-mail: philip.marriott@rmit.edu.au
trap),1114 and solid phase microextraction (SPME)15 procedures. Nevertheless, none of the above methodologies
directly studied the emitted VOCs from the pepper matrices
in their natural state, without any pretreatment. It has been
reported that thermal treatment and the presence of water
promote many chemical or enzymatic reactions12,15 that
introduce major variations to the pepper aroma profile
composition. To our knowledge, the application of comprehensive two-dimensional gas chromatography hyphenated
with flame ionization detection (GC GC-FID), quadrupole
mass spectrometry (GC GC/qMS) and time-of-flight mass
spectrometry (GC GC/TOFMS) has not been applied to the
analysis and identification of aroma compounds of pepper.
In GC GC two columns are serially coupled through a
suitable interface. A chromatographic method is considered
comprehensive1618 when (a) the transfer from the first to
the second column is qualitatively and quantitatively
complete and (b) the orthogonality principle is respected,
i.e. the separation systems must be independent, and the
separation achieved on the first column is preserved on the
Copyright # 2006 John Wiley & Sons, Ltd.
2824 Z. L. Cardeal, M. D. R. Gomes da Silva and P. J. Marriott
second one. The sample is transferred from the first to the

second column without mass loss. In this work we used as an
interface a longitudinal modulation cryofocusing system
(LMCS)19,20 that has already proved reliable for diverse
analytical applications.2123 The interface, located between
the two columns, allows a further fractionation of the
primary separation obtained on the first column, while
preserving the previously achieved separation. The separation mechanisms in the two columns should be based on
different and independent physicochemical interactions (e.g.
boiling point vs. polarity), in order to maximize the
individual column capacities. Since the different fractions
of an eluate are cryofocused before the separation on the
second column, the resulting segments (peaks) of the
modulation are more intense in the second column due to
narrowing of the now focused chromatographic peak.
Moreover, the column bleed from the primary column can
be effectively resolved from components on the second
column, which is important when identification tasks are
performed. The modulated chromatogram is then, by means
of suitable computer software, transformed into a contour
plot comprising two dimensions, or as a 3D plot when peak
intensity is considered. These presentations are reconstructions of the separate linear modulated chromatograms
projected on the 2nd dimension in each modulation. Of
signal importance to the analyst is the far greater peak
separation achieved by GC GC, effectively within the same
time period as a conventional GC analysis, which arises due
to fast GC separation on the second column, i.e. within about
36 s. These very sharp peaks must be adequately detected,
and identified, and for reliable area measurement about ten
data point samples per peak are required. Clearly, identification of GC GC peaks based on MS detection will
demand a fast duty cycle for data acquisition.
The TOF mass analyzer can generate (after bunching) up to
500 spectra s1 because of its high ion extraction ratio. Since
all fragment ions are almost simultaneously detected, the
TOFMS technique allows for effective spectral deconvolution, reportedly even in multicomponent co-elution situations. Combining the GC GC and MS techniques, one is
able to attain high resolution and sensitivity, which
constitute important goals in trace analysis, especially in
aroma (odour) research.
In this work, after VOC HS-SPME extraction of 13 different
varieties of pepper and peppercorn samples (herein all
samples will referred to by the generic name of pepper) from
different geographical origins, GC GC/qMS allowed the
tentative identification of more than 300 compounds. The
Brazilian pink pepper sample was also analyzed by using
GC GC/TOFMS, which exemplifies the identification
capabilities of this detector relative to the more commonly
available, but slower scanning, qMS detector. Retention
indices were calculated for GC GC/qMS data and were
used to support the tentative assignment of individual peaks.
EXPERIMENTAL
Sample preparation and SPME procedure
The pepper corn or crushed samples were purchased from
local markets and spice purveyors. The 13 pepper varieties
were: A: Black Pepper (Piper nigrum, India); B: Cubeb Pepper

(Piper cubeba, Indonesia); C: Green Pepper (Piper nigrum,
Australia); D: Long pepper (Piper retrofractrun, Indonesia);
E: White Pepper (Piper nigrum, Malaysia). F: Australian
Pepperberrie (Tasmannia lanceolata, Australia); G: Brazilian
Pink Pepper (Schinus terebinthifolius, Brazil); H: Cayenne Pepper
(Capsicum frutescens, French Guiana); I: Jamaica Pepper
(Pimenta dioica, Jamaica); J: Mingnonette Pepper, a French
mixture of coriander (Coriandum sativum), Tellicherry Black
Pepper (Piper nigrum, India) and Muntok White Pepper
(Piper nigrum, Indonesia); K: Shichimi Pepper, Japan, a
mixture of hot chile (Capsicum frutescens), and Sichuan
Pepper (Zanthoxylum piperitum) with sea grass, sesame
(Sesamum indicum) seeds and orange (Citrus simensis) peel;
L: Chinese Pink Pepper (Zanthoxylum sansho piperitum,
China); M: Sichuan pepper (Zantoxylum simulans piperitum,
China). Each sample was ground with a glass mortar and
pestle. For the SPME procedure an aliquot of 1.0 g of ground
pepper was introduced into a 22 mL Pyrex vial, in which 3 mL
of a 10 mg/L hydrocarbon (C9-C22) stock solution in hexane
was added. The vial was then immediately sealed with a
Teflon-lined rubber septum/aluminium cap. The pepper vial
samples were equilibrated at 408C during the sorption step.
The manual SPME holder and the SPME fibres were
purchased from Supelco (Bellefonte, PA, USA). The SPME
fibres tested were 100 mm PDMS, 85 mm polyacrylate (PA),
65 mm StableFlexTM polydimethylsiloxane/divinylbenzene
(PDMS/DVB),
75 mm
carboxen/polydimethylsiloxane
(CAR/PDMS), 65 mm CarbowaxTM/divinylbenzene (CW/
DVB) and 50/30 mm divinylbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS). Prior to use, all fibres
were conditioned following the manufacturers recommendations. After initial screening (data not shown) a 65 mm
PDMS/DVB coated fibre was chosen to extract pepper
VOCs. The fibre was exposed to the sample headspace
during a suitable sorption period of 20 min at 408C
(conditions as determined by prior study, not shown), and
introduced into the GC injection port to allow thermal
desorption of the analytes at a temperature of 2208C for a
0.10 min period.
Equipment
The three GC GC systems were retrofitted with a longitudinally modulated cryogenic system (LMCS; Chromatography Concepts, Doncaster, Australia). The LMCS was
operated at a modulation period of 6 s at a temperature
between 20 and 308C. All GC systems consisted of an
Agilent Technologies 6890 model GC system (Agilent
Technologies, Burwood, Australia).
The GC GC/qMS instrument was equipped with a 5973
mass-selective detector, a model 6873 autosampler and
ChemStation software. The MS detector, with fast electronic
upgrade to allow faster scanning, was operated with a
transfer line temperature of 2808C, the detector voltage was
1.7 kV, and a mass scan range of m/z 45400 was used to give
a data acquisition rate of 19.77 Hz. Spectra were matched
with both the Adams and NIST 98 MS libraries using
ChemStation software.
The GC GC/TOFMS system consisted of a Pegasus III
time-of-flight mass spectrometer (LECO Corp, St. Joseph, MI,
DOI: 10.1002/rcm
Comprehensive 2D GC/MS analysis of pepper volatiles
USA) operated at a storage rate of 100 Hz, with a mass range

of m/z 45400 and a multichannel plate voltage of 1.7 kV. Data
were processed using LECO Corp ChromaTOFTM software.
In both the GC GC/TOFMS and GC GC/qMS systems,
a 0.50 m length and 0.1 mm i.d. deactivated fused-silica
interface capillary was used (SGE International, Ringwood,
Australia); for GC GC/qMS, 0.21 m of the capillary was
located inside the interface zone and 0.29 m inside the oven,
while, for GC GC/TOFMS, 0.17 m was in the interface zone
and 0.33 m in the oven.
The GC GC-FID system used was equipped with an
Optic 2 programmable injector and a multicapillary liner
(ATAS, Veldhoven, The Netherlands). The flame ionization
detector (FID) was operated at a data collection frequency of
100 Hz, at 3008C. Agilent ChemStation software was used for
data acquisition and instrument control.
The columns used for GC GC experiments (FID and
qMS) comprised a BPX5 (5% phenyl polysilphenylenesiloxane) primary column; 30 m 0.25 mm i.d. 0.25 mm
film thickness (df), directly coupled to a BP20 (polyethylene
glycol) second column with dimensions 1.5 m 0.1 mm
i.d. 0.1 mm df. For GC GC/TOFMS a BPX5 primary
column (dimensions as above) was directly coupled to a
BP20 second column (1.0 m 0.1 mm i.d. 0.1 mm df). The
columns were purchased from SGE International. The oven
temperature program for all GC GC analyses began at
358C, was held at that temperature for 5 min, raised to 2108C
at 38C/min, then to 2408C at 40 C/min and held for 4 min at
2408C. The carrier gas was hydrogen at a flow rate of 1.0 mL/
min for the GC GC-FID system. Helium was used as the
carrier gas at a flow rate of 1.3 mL/min with the GC GC/
TOFMS and GC GC/qMS systems.
RESULTS AND DISCUSSION

HS-SPME and HS-SPME-GC T GC method
development
The SPME parameters fibre type, absorption time, and
temperature and desorption time were studied and set at
acceptable values by using the GC GC-FID system with
Brazilian pepper as a test sample. Among the fibres
investigated, the PDMS/DVB fibre, containing micropores
average diameter, provided higher sorption capacity
of 17 A
and a broader range of sorbed VOC solutes from pepper than
the other fibres tested (PDMS, PA, CAR/PDMS, CW/DVB
and DVB/CAR/PDMS), over the target molecular mass and
polarity range of compounds present.
Extraction temperatures between 30 and 808C with a
10 min extraction period were investigated to determine fibre
coating sorption efficiency. The results showed that temperatures above 408C reduced the quantity extracted because
of the exothermic effects of the sorption process. Consequently, a temperature of 408C was selected as adequate
for the extraction procedure.
For the selection of the extraction time period, sorption
times of 1.0, 2.0, 10.0, 20.0, 30.0, 60.0 and 120.0 min were
tested at 408C. The results showed that the signal increased
up to an extraction time of 20 min, with little appreciable
changes thereafter. This duration of 20 min was then chosen
for sorption of VOCs in all subsequent studies of pepper
2825
samples. Note that different compounds reach maximum

sorbed amount at different times; 20 min may not give the
maximum response for some components, but 20 min was a
convenient compromise. Desorption time periods of 0.1, 0.5,
1.0 and 2.0 min were tested, at an injector temperature of
2008C. A time of 0.1 min was chosen since it is adequate for
complete desorption, and avoided carryover.
The LMCS period of modulation and the trap temperature
were optimized. The temperatures studied were from 208C to
408C in decrements of 108C. Lower temperatures generally
result in better resolution since they give reduced peak width
in the second dimension. However, when the trap temperature was too low the second-dimension peak widths tended
to broaden, probably due to slower re-heating of the cold
column zone. Thus, a temperature between 208C and
308C was then selected for all GC GC experiments.
Modulation periods of 4, 5, 6 and 7 s were studied, with 6 s
chosen since it presents an adequate time for completion of
the 2nd dimension (2D) analysis and an acceptable
modulation ratio (i.e. sampling events of the 1D peak),
without excessive wrap-around.
Reproducibility was investigated by performing triplicate
GC GC-FID analyses on Brazilian pepper using the
optimized sampling and analytical conditions above, and
this confirmed previous results for methoxypyrazine
sampling from wine headspace.24 The reproducibility was
deemed adequate to allow sample-to-sample comparison
both with respect to GC GC retention reliability and in the
confirmation of the presence of peaks in different samples to
support the identifications described below.
GC T GC-FID and GC T GC/qMS analysis

Comparing the contour plots obtained after GC GC FID
with GC GC/qMS (Figs. 1(A) and 1(B), respectively) for the
Brazilian pepper variety, one can observe that the plots are
similar, allowing easy pattern comparison and peak assignments in the FID chromatogram based on the GC GC/qMS
TIC response. For identification and quantitation (the latter is
not an aim of this work) this is of importance since direct
comparison between GC GC FID and GC GC/qMS is
thus possible, although GC GC/qMS provides only
semiquantitative capability due to its slower data acquisition
rate.25 These results show the possibility of fingerprint type
comparison, which becomes even more important if many
similar samples are investigated, since this facilitates sampleto-sample comparison and component identification. In this
study 13 pepper samples were extracted by HS-SPME and
submitted to GC GC-FID and GC GC/qMS analysis. For
the sake of clarity, only the hydrocarbon spots from the stock
solution introduced into the pepper sample are indicated,
along with some compounds which are presented in order to
facilitate visual comparisons between the contour plots;
compound numbering corresponds to those of Table 1. A
group of three compounds identified by the letter a
illustrates the wrap-around effect in the GC GC/qMS
system; their identification was not conclusive since the
spectral quality matches were not sufficiently good to be
considered for positive identification. Compounds 123 and
163 likewise elute within one modulation period in GC GCFID, but wrap-around in GC GC-qMS.
DOI: 10.1002/rcm
Figure 1. HS-SPME-GC GC-FID (A) and HS-SPME-GC GC/qMS (B) contour plots for the analysis of the
Brazilian pepper sample. Hydrocarbons from C9 to C18 from a stock solution spiked into the sample are indicated.
Major components are indicated by zones with white centres. Analytical conditions for each case are detailed in
the Experimental section. Peak numbers are according to Table 1. Peak group a comprises up to three components
(see text).
The tentative identification of components was made by

comparing spectra with Adams and NIST 98 mass spectral
database libraries, supported by retention index data. Table 1
shows all the identified peaks for the 13 pepper samples, by
using GC GC/qMS with the respective retention indices
calculated according to the van den Dool and Kratz
equation28 and Adams retention indices.29 The calculated
retention indices were obtained through the modulated
chromatogram, considering the most intense modulated
peak of the compound of interest, and those for both adjacent
nearest n-alkanes (in all cases identified through analysis of
the individual mass spectra of the compounds). This
modulated peak is a pulsed peak generated by the
modulation process, that produces several peak events
which have a distribution about the weighted mean retention
of the peak. The cryomodulation process adds a maximum of
one modulation period to the total retention time. The
maximal absolute index difference that one admits as
acceptable in this study, compared with the available
literature indexes, was 30. This variation is considered
reasonable, if one takes into account that the literature values
were determined in a one-dimensional system with a nonpolar column phase (5% phenyl polysilphenylene-siloxane).
In this work, the system is bi-dimensional, and, although
the first dimension column has a similar phase composition,
and consequently similar polarity, to the one used as the
reference column phase,29 the second dimension column was
a polyethylene glycol based polar phase column that not

unexpectedly shifted the retention indices to larger values
than the literature values (hence the D (RILit) (RICalc) values
in Table 1 are mostly negative values). This shift depends on
the polarity of each molecule, since in polyethylene glycol
phases the interaction mechanism between solute and
stationary phase is dominated by the polarity properties of
the compound, whereas in a high methyl content polysiloxane phase the compound separation is dominated by
evaporation enthalpy effects, or in other words, according
to boiling points. Nevertheless, all the identified compounds
were tentatively assigned provided that the mass quality
correlation was better than 85%. In those cases where the
absolute retention index differences were around 30, the
identification was only considered positive if the particular
compound had a mass spectral match better than 90%.
Comparison with some co-injected standards gave the same
(large) retention index deviation for some (mainly polar)
compounds, confirming that retention index deviations from
the database values were not unexpected. In this work,
retention indices based on a bi-dimensional system are
reported. This could be useful for future essential oil
characterization research when GC GC is to be performed,
since there is no published work with this type of data (i.e.
index data on a two-column system). Naturally, the retention
indices will vary depending on the column configurations
selected, especially if polyethylene based columns are used,
DOI: 10.1002/rcm
2827
Table 1. Peak identification after HS-SPME-GC GC/qMS and HS-SPME-GC GC/TOFMS analysis, with calculated and
literature retention data.26 See text for analytical conditions. Each pepper species is assigned with a letter. A: Black Pepper
(Piper nigrum, India); B: Cubeb Pepper (Piper cubeba, Indonesia); C: Green Pepper (Piper nigrum, Australia); D: Long Pepper
(Piper retrofractrun, Indonesia); E: White Pepper (Piper nigrum, Malaysia). F: Australian Pepperberrie (Tasmannia lanceolata,
Australia); G: Brazilian Pink Pepper (Schinus terebinthifolius, Brazil); H: Cayenne Pepper (Capsicum frutescens, French Guiana);
I: Jamaica Pepper (Pimenta dioica, Jamaica); J: Mingnonette Pepper, a French mixture of coriander (Coriandum sativum),
Tellicherry Black Pepper (Piper nigrum, India) and Muntok White Pepper (Piper nigrum, Indonesia); K: Shichimi Pepper, Japan, a
mixture of hot chile (Capsicum frutescens), and Sichuan Pepper (Zanthoxylum piperitum) with sea grass, sesame (Sesamum
indicum) seeds and orange (Citrus simensis) peel; L: Chinese Pink Pepper (Zanthoxylum sansho piperitum, China); M: Sichuan
pepper (zantoxylum simulans piperitum, China). All the peaks have a match quality spectrum better than 85%
Pepper samples
Peak
Nr.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Compound
a
Hexanal
Heptan-2-one
a-Thujenea,b,c,d
Methyl hexanoate
Tricyclene
a-Pinenea,b,c,d
a-Fenchene
Camphenea,b,c,d
3-Methyl-6-(1-methylethylidene)-cyclohexene
Isobutyl butanoate
3,7,7-trimethyl-3,5-cycloheptatriene
Benzaldehyde
Sabinenea,b,c,d
b-Pinenea,b,c,d
3-Ethyl-2,5-dimethyl-hexane
Myrcenea,c,d
6-Methyl-hept-5-en-2-one
7-Methyl-non-1-yne
b-Thujene
Octan-2-one
2-Pentylfuran
Cyclofenchene
Linalool oxideb
Pseudolimonene
d-2-Carene
a-Phellandrenea,b,c,d
Hexyl acetate
Z-Dehydroxy linalool oxide
a-Terpinenea,c,d
p-Cymenea,b,c,d
2-Bornene
3-Isopropenyl-5,5-dimethyl-cyclopentene
Sylvestrene
o-Cymene
Z-b-ocimenea,c
Limonenea,b,c,d
b-Phellandrenea,c,d
1,8-Cineoleb,c
1,5-Dimethyl-cycloocta-1,5-diene
2,2,6-Trimethyl-cyclohexanone
E-b-ocimenea,b,c
Thujol
d-3-Carenea,b,c,d
Pentyl-isobutanoate
2,3,6-Trimethyl-1,5-hepta-1,5-diene
g-Terpinenea,b,c,d
E-Linalool oxideb
Z-Sabinene hydratea
Nonan-2-one
Acetophenone
Z-Linalool oxide
m-Cymene
Terpinolenea,b,c,d
RI Lit.
802
892
930
927
927
939
953
950
960
975
980
980
995
986
991
993
1004
1002
1003
1003
1008
1017
1026
1031
1026
1037
1029
1030
1031
1050
1031
1060
1073
1070
1090
1065
1087
1085
1088
RI Calc.
G
x
x
x
930
932
934
937
949
956
957
961
980
981
983
985
987
988
995
996
996
997
997
1000
1003
1012
1014
1016
1019
1019
1025
1025
1027
1028
1032
1034
1035
1037
1041
1042
1047
1047
1054
1055
1061
1062
1063
1066
1080
1082
1083
1088
1090
1090
1093
0
5
7
2
4
6
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
21
8
5
7
7
9
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
6
1
8
2
8
11
11
x
x
x
x
x
x
x
x
x
x
x
x
x
x
10
8
12
13
16
11
8
1
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
4
x
30
x
x
6
7
12
7
23
3
5
5
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Continues

DOI: 10.1002/rcm
Table 1. (Continued)
Pepper samples
Peak
Nr.
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
Compound
2-Ethyl-p-xylene
Z-2,5-Ethenyltetrahydro-a,a,5-trimethyl-furanmethanol
4-Carene
6,7-Epoxymyrcene
Ethenyl-dimethyl-benzene (isomer)
p-Cymenene
Linaloola,b,c,d
Fenchonec
Isopentyl isovalerate
o-Cresol acetate
2-(2-Methyl-1-propenyl)-bicyclo[2.2.1]heptane
Methyl benzoate
E-Sabinene hydratea,c,d
3,7-Dimethyl-octa-1,5,7-trien-3-ol,
1,6-Dimethyl-hepta-3,5-dien-2-one
Z-Thujone
3-Methyl-but-3-enyl-isovalerate
p-Mentha-1,3,8-triene
4-Ethyl-1,2-dimethyl-benzene
E-Thujone
3,4-Dimethylcyclohexanol
Z-Myroxide
allo-Ocimeno
a-Pyronene
Menth-2-en-1-ol
E-p-Mentha-2,8-dien-1-ol
a-Campholenal
E-Non-3-en-1-ol
Z-Limonene oxide
terc-Butylbenzene
4-Acetyl-1-methylcyclohexene
1,2,3,4,5,8-Hexahydronaphthalene
E-Limonene oxide
2-Methyl-6-methylen-octa-1,7-dien-3-one
5-Methylundecane
E-Myroxide
Epoxyterpinolene
E-Sabinolb
4-Methylundecane
Z-Pinocarveol
2-Methylundecane
Camphora,b,c
Menthone
4-Butyl-cyclohexene
1,3-Dimethyl-5-(1-methylethyl)-benzene
2,6-Dimethyl-bicyclo[3.2.1]octane
Sabina ketone
6,6-Dimethyl-2-methylene-bicyclo[2.2.1]heptan-3-one
Isoborneola
Pinocarvone
Isomenthone
p-Mentha-1,5-dien-8-ol
d-Terpineol
Thuj-3-en-2-one
a,a,4-Trimethyl-cyclohex-3-ene-1-methanamine
Isopulegone
4,5-Dihydro-4-(2-methyl-3methylenebut-4-yl)-2(3H)-furanone
Terpinen-4-ola,b,c,d
Hexyl-2-methylbutanoate
terc-Butyltoluene
RI Lit.
1093
1091
1097
1087
1103
1139
1091
1098
1102
1115
1110
1114
1135
1132
1122
1123
1126
1137
1142
1183
1145
1142
1184
1146
1153
1159
1162
1165
1163
1170
1166
1171
1177
1236
RI Calc.
1093
1094
1097
1099
1103
1103
1105
1106
1108
1109
1111
1112
1114
1115
1118
1121
1122
1125
1125
1126
1126
1127
1132
1135
1137
1138
1141
1143
1145
1146
1147
1147
1149
1152
1155
1156
1156
1158
1159
1160
1163
1164
1166
1166
1172
1177
1178
1179
1179
1180
1181
1181
1182
1185
1187
1187
1187
1192
1195
1197
1204
1207
G
x
\
x
x
x
x
x
x
6
12
8
19
5
30
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
21
16
x
x
x
x
x
x
19
7
15
x
x
x
x
12
x
x
8
0
15
15
15
8
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
7
x
x
28
11
x
x
x
16
x
x
24
20
19
x
x
x
x
x
x
x
x
x
20
x
x
19
16
19
15
21
16
x
x
x
x
x
x
x
20
32
x
x
x
x
x
Continues

DOI: 10.1002/rcm
2829
Pepper samples
Peak
Nr.
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
Compound
RI Lit.
RI Calc.
Cryptone
Hexyl isovalerate
2,6-Dimethyundecane
a-Terpineola,b,c,d
Myrtenad
Methyl chavicol
Octyl acetate
p-Cymen-8-ol a
Z-Anethole
Safranal
E-Dihydrocarvone
Z-Dihydrocarvone
2-Butyl-1,1,3-trimethyl-cyclohexane
a-Phellandrene epoxyde
2,6-Dimethyl-octa-3,7-diene-2,6-diol
Eucarvone
6,6-Dimethyl-2-(2-methylpropyl)bicyclo[3.1.1]heptan-3-one (isomer)
4-Methylene isophorone
Z-Sabinene hydrate acetate
E-Carveola,b
E-Chrysanthenyl acetate
Hexyl-cyclohexane
6-Methyl-dodecane
5-Methyl-dodecane
2,4-Dimethylundecane
Linalyl acetate
4-Methyl-dodecane
Pulegone
E-sabinene hydrate acetate
Methyl veratrole
1-Methyl-3-(1-methylethenyl)-cyclohexene
2,3,4-Trimethyl-decane
Z-Carveol
Methyl citronellate
Cumin aldehydea
Carvonea
3-Carvomenthenone
3-Methyl-dodecane
5-Isopropenyl-2-methylcyclopent-1-enecarboxaldehyde
2-Isopropyl-5-methylcyclohex-3-en-1-one
Carvenone
Piperitonea
a-Terpinen-7-al
5-Tridecene
Methyl 3-phenylpropanoate
Bornyl acetateb
Undecan-2-one
2-Isopropyl-2,5-dimethyl-cyclohexanone
E-Anethole
Terpinenyl acetate
2,3,4-Trimethyl-benzaldehyde
Safrolea,b,c
6-Methyl-tridecane
2,3,5,8-Tetramethyl-decane
Lyratyl acetate
E-Piperetone epoxyde
Methyl geranate
Isoascaridole
2,5,6-Trimethyl-hepta-1,3,6-triene
Myrtenyl acetate
Piperitenoneb
d-Elemenea,b,c
1186
1244
1210
1211
1213
1213
1216
1216
1216
1216
1216
1218
1218
1224
1228
1229
1229
1233
1237
24
33
1189
1196
1196
1214
1183
1253
1197
1201
1218
1221
1217
1238
1257
1237
1256
1229
1261
1242
1243
1253
1258
1253
1285
1289
1294
1285
1300
1315
1287
1325
1303
1327
1343
1338
1242
1243
1246
1246
1247
1251
1253
1253
1256
1257
1258
1258
1259
1260
1260
1261
1264
1265
1265
1269
1270
1271
1275
1277
1277
1282
1289
1293
1297
1303
1306
1307
1307
1312
1314
1316
1318
1321
1330
1332
1332
1334
1336
1339
1344
G
M
x
x
x
24
20
20
2
33
37
21
17
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
24
22
29
8
x
x
x
x
x
x
x
x
1
x
x
21
2
x
x
x
x
x
32
3
23
22
16
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
19
24
3
x
x
8
9
22
7
3
27
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
7
29
x
x
x
x
x
x
x
x
9
4
6
x
x
x
x
x
x
x
Continues

DOI: 10.1002/rcm
Pepper samples
Peak
Nr.
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
Compound
Exo-2-hydroxycineole acetate
Cytronellyl acetate
a-Cubebenea,b,c,d
a-Terpinyl acetateb
Linalyl butanoate
8,9-Dehydro-neoisolongifolene
Aciphyllene
2,6-dimethyl-3-octylacetate
a-Ylangene
Neryl acetate
Isoledene
Cyclosativene
2,7-Dimethyl-2,6-octa-2,6-dien-1-ol
Geranyl acetateb
a-Copaenea,b,c,d
Eugenola,d
E-Tetradec-5-ene
b-Cubebenea,b,c,d
E-Tetradec-7-ene
b-Bourboneneb,c
b-Elemenea,b,c,d
b-Maaliene
terc-Butyl-1-tetralone
Isoeugenol
Sativene
7-epi-Sesquithujene
Cyperene
Italicene
a-Gurjunenea,c
Z-a-Bergamotene
Methyl eugenola,b
Z-caryophyllenea,b,c
a-Santalene
b-Ylangene
epi-b-Santalene
E-caryophyllenea,b,c,d
g-Elemenea
E-a-Bergamotenea,b
b-Gurjunenec
a-Guaienea,d
b-Cedreneb
b-Copaeneb
Cyclohex-1-ene-1-methanol4-(1-methylethenyl)-acetate
Thujopsene
Aromadendrenec
Z-b-Farnesene
E-Muurola-3,5-diene
E-b-Farnesenea,b,c,d
a-Clovene
a-Humulenea,b,c,d
Dehydro aromadendrane
9-epi-E-Caryophyllene
Z-Muurola-4(14),5-diene
allo-Aromandereneb,c
Dodecanol
g-Muurolenea,b,c,d
Ethyl-deca-(2E,4Z)-dienoate
E-Cadina-1(6),4-diene
b-Chamigrene
g-Gurjunene
g-Curcumene
a-Amorphenec
RI Lit.
1353
1351
1349
1375
1362
1376
1371
1381
1380
1359
1388
1388
1391
1382
1407
1392
1391
1399
1406
1410
1413
1404
1409
1418
1421
1447
1419
1437
1435
1434
1440
1421
1432
1431
1441
1443
1450
1457
1454
1455
1463
1466
1467
1460
1471
1480
1469
1477
1478
1477
1483
1485
RI Calc.
1352
1354
1356
1359
1363
1365
1367
1372
1377
1379
1380
1382
1383
1386
1387
1390
1391
1394
1396
1397
1397
1398
1400
1403
1405
1408
1420
1420
1421
1425
1428
1431
1431
1432
1432
1440
1441
1442
1446
1446
1447
1450
1452
1456
1456
1459
1467
1467
1468
1469
1471
1471
1478
1479
1486
1487
1488
1488
1488
1490
1493
1493
1
5
10
x
x
G
x
x
x
x
x
x
x
x
x
M
x
x
x
x
x
x
2
17
4
11
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
5
7
31
x
x
6
9
6
16
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
4
13
17
21
14
11
12
24
22
13
11
15
21
4
7
12
6
26
18
25
15
16
17
10
14
14
8
5
11
19
15
7
19
11
10
13
10
8
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Continues

DOI: 10.1002/rcm
2831
Pepper samples
Peak
Nr.
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
Compound
b,c
a-Curcumene
Germacrene Da,b,c,d
Z-b-Guaienea,c
epi-Bicyclosesquiphellandrene
g-Himachalene
d-Selinene
1R,3Z,9S-4,11,11-Trimethyl-8methylenebicyclo[7.2.0]undec-3-ene
a-Zingiberenea,c,d
E-Muurola-4(14)-5-diene
b-Selineneb,d
Valencenec
g-Amorphene
Viridiflorene
2-Isopropyl-5-methyl-9-methylenebicyclo[4.4.0]dec-1-ene
Elixene
a-Bisabolenec
a-Cuprenene
a-Selineneb,c,d
a-Muurulone
Bicyclogermacrenea,c
E-b-Guaienea
1-Ethyl-3-vinyl-adamantane
Cuparene
b-Bisabolenea,b,d
g-Cadinenec
2a-Isopropenyl-6a-isopropyl-3a-methyl-3avinylcyclohexanone
g-Bisabolene
b-Curcumenenec
E-Calamenenea,b
7-epi-a-Selinene
Zonarene
d-Armophene
d-Cadinenea,b,d
b-Sesquiphellandrene
E-Cadina-1,4-dienea,b,c
Selina-3,7(11)-diene
b-Vetivenene
Z-Calamenenea,b
Octahydro-4,4,8,8-tetramethyl-4a,7-methano-4a
H-naphth[1,8a-b]oxirene
1,2,3,4,6,8a-Hexahydro-1-isopropyl4,7-dimethyl-naphthalene,
1,1,4,8-Tetramethyl-cycloundeca-4Z,7Z,10Z-triene
Italicene ether
a-Cadinene
a-Calacorene
E-Isocroweacin
Elemola,b
E-Nerolidola,b,c,d
g-Gurjunenepoxide
Germacrene Ba,c
b-Calacorene
Hexadecene
Ledold
Palustrol
Caryophyllene oxidea,b,d
Spathulenola,c
Guaiola,b
E-Tridecen-6-yne
Selina-6-en-4-ol
RI Lit.
RI Calc.
1481
1485
1493
1495
1496
1497
1498
1500
1501
1505
14
11
4
x
x
1483
1493
1494
1494
1490
1496
1496
1497
1490
1507
1506
1498
1500
1500
1503
1505
1506
1514
1515
1516
1529
1522
1530
1512
1523
1523
1535
1547
1533
1540
11
12
17
11
12
12
1511
1511
1512
1513
1513
1513
1514
1514
1518
1520
1520
1522
21
4
6
15
13
13
11
1523
1523
1523
1530
1530
1533
1533
1536
1538
1539
1540
1541
1544
8
7
6
8
0
21
10
13
3
8
7
1
G
x
x
x
x
x
x
x
x
x
x
x
x
17
8
1505
1506
1507
1507
1508
1509
1510
1561
1566
1590
1569
1583
1578
1601
1552
1554
1554
1560
1564
1570
1575
1575
1578
1585
1590
1593
1593
1605
1606
1617
1619
1624
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
13
14
6
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
1551
1538
1539
1546
1555
1550
1563
x
x
16
15
14
9
20
12
x
x
x
x
x
x
x
x
x
x
x
x
x
17
19
0
24
22
28
16
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Continues

DOI: 10.1002/rcm
Pepper samples
Peak
Nr.
298
299
300
301
302
303
304
305
306
307
308
309
Compound
Ledene oxide
Humulene epoxide
a-Muurolola,b,d
3,4-Dimethyl-cyclohex-3-en-1-carboxaldehyde
2,4,5,6,7,7a-Hexahydro-3-(1-methylethyl)7a-methyl-1H-2-indenone
Cubenola
Heptadec-8-ene
Tetradecanol
b-Eudesmola,b,d
Cadalene
Cyclocorolonone
3,5,6,7,8,8a-Hexahydro-4,8a-dimethyl-6(1-methylethenyl)-2(1H)-naphthalenone
RI Lit.
1608
1642
1647
1673
1661
1677
1761
RI Calc.
1631
1635
1638
1638
1663
1674
1679
1687
1688
1705
1788
1790
G
x
27
4
x
x
x
x
x
27
14
27
28
27
x
x
x
x
x
x
x
x
x
x
D: Difference between the literature and the calculated retention indices; (RILit) (RICalc).

Value not determined due to the absence of the hydrocarbon C8 (n-octane).

Data confirmed by TOFMS.
a
Described in Piper sp. in Ref. 5.
b
c
d
since they are very polar and their stability may be variable.
Nevertheless, the database composed of the data here
constitutes a developmental step in essential oil analysis
using a GC GC system.
GC T GC/TOFMS analysis
Figure 2 shows a GC GC/TOFMS analysis of the same
Brazilian pepper variety as that shown in Fig. 1, but as a 3D
relief plot, and again it is clearly demonstrated that the
contour plot presents a high correlation (match) with the
Figure 2. HS-SPME-GC GC/TOFMS 3D relief plot for the

analysis of the Brazilian pepper sample. Analytical conditions
are detailed in the Experimental section.
previous contour plots obtained, perhaps better matching

with the GC GC-FID result of Fig. 1(A). Here the same
compounds or group of compounds are also indicated in
order to allow easy comparison with the previous 2D contour
plots. For the purposes of this study, the maximum number
of processed peaks was limited to 1000 (peaks with signal-tonoise (S/N) > 500). The resulting table gave 1000 identified
compounds, but with some incorrect assignments (i.e.
multiple peak entries for compounds eluting at different
retention positions, but reported as the same compound;
analytes modulated into two second-dimension peaks or
broad second dimension peaks recognized as two or more
individual peaks by the data processing), as already
described.22 In the case of the Brazilian pepper sample,
where GC GC/TOFMS was performed as an example to
compare the identification capabilities of this system with the
more commonly available system based on qMS detection,
only the compounds that were identified as being the same
by both GC GC/TOFMS and GC GC/qMS analyses were
reported in the identification peak listing in Table 1 (all other
compounds in Table 1 not assigned as being contained in the
Brazilian pepper sample were only identified by means of
GC GC/qMS). More than 700 compounds were identified
in the Brazilian pepper by GC GC/TOFMS, the great
majority in trace amounts, of which only 139 could be
detected by GC GC/qMS. The GC GC/TOFMS system
allowed a clear chromatographic separation of the peaks in
both dimensions, for example as shown in the segment of the
contour plot in Fig. 3, which is also achieved by GC GC/
qMS since the basic analytical separation system is the same.
In this case, however, easier data processing is possible with
the LECO ChromaTOFTM system, than the manual processing steps needed for GC GC/qMS data to generate a
contour plot. Algorithmic calculations with GC GC/
DOI: 10.1002/rcm
2833
Figure 3. HS-SPME-GC GC/TOFMS expanded segment 3D relief plot of the Brazilian pepper sample. Compounds are numbered according to Table 1, except those not co-identified by GC GC/qMS, which are simply
named on the plot. Analytical conditions are detailed in the Experimental section.
TOFMS data recognize minimal variation in retentions of coeluting peaks such that superimposed spectra may be
spectrally deconvoluted to allow detection of the individual
components, even after the extra separation power of
GC GC. In both cases, not only is better identification of
the compounds in the contour plot possible, but the
production of pure MS spectra of each constituent is
achieved. This fact is significant and offers considerable
advantages over one-dimensional (1D)-GC, since it is now
possible to obtain pure mass spectra of almost all the
compounds, and thus an opportunity for building mass
spectral libraries with more reliable spectral data, ideally
supported by first- and/or second-dimension retention data.
By way of example, as seen in Fig. 3, Z-carveol, 4,4dimethyl-cyclohex-2-en-1-ol, linalyl acetate and an unidentified peak (n.i.) should provide high-purity mass spectra,
whereas they would have overlapped in the first-dimension
column. The latter component offers an interesting possibility; if its spectrum is apparently pure, but it still cannot be
assigned to any library entries with good reliability, it
suggests that the GC GC technique may provide a route
towards the development of more complete libraries.
Authentic compound structures would still be the absolute
arbiter for proper compound identification and library entry
confirmation. Furthermore, E-sabinene hydrate acetate and
sabinyl acetate are now also completely separated, and also
resolved from column bleed arising from siloxane groups,
which is a typical interference in GC. Reduction in siloxane
interferences again provides improved spectral quality. The
same argument can be drawn for almost all of the
compounds methyl citronellate, carvotanacetone, carvone,
5-isopropenyl-2-methycyclopent-1-enecarboxaldehyde and
cumin aldehyde, which have similar first-dimension
retention times but are resolved on the 2D column, and
are subsequently identified with spectral qualities better then
90%. Using GC GC/TOFMS, it was possible to identify a
great variety of compounds, amongst them monoterpenes,
sesquiterpenes, oxygenated terpenoids and other volatile
compounds, such as aliphatic and aromatic hydrocarbons,

aldehydes, esters, ethers and alcohols, although, as reported
above, the compounds sabinyl acetate, carvotanacetone and
4,4-dimethyl-cyclohex-2-en-1-ol are not included in Table 1
as belonging to Brazilian pepper, since they were not coidentified by GC GC/qMS (presumably the identification
procedure through the modulated chromatogram and/or the
sensitivity of the quadrupole did not allow successful
identification). The increased analytical separation achieved
with GC GC hyphenated with TOFMS, assembled in one
instrument, with consequent maximization of separation and
improved spectral resolution, clearly has a very important
role in the characterization of the kind of matrices
represented by essential oils. By extension, the GC GC/
qMS system, although less sensitive, can still be very helpful
when identification tasks are to be performed for sample
screening and characterization.
Comparison between qMS and

TOFMS mass fragmentation patterns
Figure 4 presents a plot of the abundances of the ions at m/z
54 [82 CO], 82 [M C5H10], 95 [M C4H9], 110 [M
C3H6], 124 [M C2H4], 137 [M CH3] and 152 [M] for
piperitone obtained using GC GC/qMS. This compound
was only identified in the Brazilian pepper and Chinese pink
pepper. The data shown in Fig. 4 are for successive scans over
the most intense peak among the three modulated pulses
obtained for this compound, as shown in Fig. 5. By means of
the fast acquisition capability afforded by the electronic
upgrade for fast scanning in qMS, it was possible to acquire
eight spectra over the peak, with reasonable mass spectral
pattern information. The first and the last spectra correspond
to the beginning and end of the peak elution, that here took
about 200 ms. For comparison, all spectra obtained were
background-subtracted, using the same background spectrum. The first and the last spectra (corresponding to the
rather low abundance extremities of the peak), when
compared with the library database, present a relatively
DOI: 10.1002/rcm
Relative Abundance
Quality match
data 38641
72%*
data 38642
94%
data 38643
91%
data 38644
95%
data 38645
95%
data 38646
94%
data 38647
91%
data 38648
83%
average spectrum 91%
54
82
95
1 1 0 12 4
137 1 52
m /z
Figure 4. Plot of the mass abundances for the most important piperitone fragment ions and
molecular ion (m/z 152) for successive scans obtained by fast qMS. The scan mass range was
m/z 45400. Quality matches are presented for each scan. For data scan # 38641, mass
spectra library search matched piperitone as second choice.
method for the m/z 82/110 ion abundance ratio, and about
four times more variation for the m/z 110/152 abundance
ratio. TOFMS apparently is not quite absolute in its ion
abundance ratio over a given peak, and in spite of the
variations experienced in qMS, the quadrupole is still fast
enough to give adequate spectral search quality, except for
the peak extremity spectral data. In all plots, the average
spectrum (given as a bold line) is positioned in the middle of
the individual data points. The differences between the qMS
and TOFMS plots show the better quality spectra of TOFMS;
however, although having a lower acquisition rate, qMS
affords comparable results with TOFMS (for major components) when identification tasks are required. In the case of
the TOFMS plots, one can observe that the library reference
spectrum (lighter bold line) presents important shifts at m/z
95 and 110, compared with the individual data points, which
poor quality match (72% and 83%, respectively). In the case

of the first data point (scan # 38641), the library assigned
piperitone as second choice. All other spectral scan data
present a match quality of over 90%, and a first choice match
for piperitone. On comparison with the average spectrum
obtained over the total peak (also with the same background
subtraction) one can observe that the resulting spectrum
effectively represents the average of all the mass abundances
of the individual data points, as its average position in the
scattered data of Fig. 5 demonstrates.
Figure 6 presents the normalized ion abundances (based
on the abundance of m/z 82) and m/z ratios obtained for qMS
and TOFMS analysis for piperitone. The ion abundance for
m/z 110, for example, varies ca. 50% using qMS compared
with ca. 25% by TOFMS. Consequently, the fast scanning
qMS displays about twice the variation of the TOFMS
3.8x104
# 38648
# 38646
# 38647
# 38644
# 38645
# 38643
# 38641
1.8x104
# 38642
Abundance
2.8x104
0.8x104
35.58
35.60
35.62
35.64
35.66
35.68
35.70
35.72
35.74
35.76
35.78
Time (min)
Figure 5. Section of the GC GC/qMS TIC chromatogram for piperitone, where the arrows
indicate the three modulated peak pulses. The expanded region of the highest peak is presented,
with scan data number shown.

DOI: 10.1002/rcm
2835
Figure 6. Plots of the normalized mass abundances and m/z ratios obtained for qMS
and TOFMS analysis for piperitone. In all plots, the average spectra (bold line) and
library reference spectra (light bold line) are indicated.
are not observed in the qMS representation. The reason for

this is not clear; however, the fact that the library spectra are
quadrupole-generated spectra can apparently contribute to
this behaviour. Nevertheless, the match achieved for the
average piperitone TOFMS spectrum is 94%. A TOFMS
dedicated mass spectral library may be required to achieve
more accurate identifications in library searching.
CONCLUSIONS
The resemblances across all the contour plots through
GC GC-FID, GC GC/qMS and GC GC/TOFMS
analysis, although under slightly different analytical conditions, show that fingerprint recognition based upon
GC GC methods can be very useful in essential oil research,
for a better understanding of the different oil compositions.
The analyses by qMS with confirmation by TOFMS (which
was performed only on Brazilian pepper) allowed the
identification of more than 300 compounds, in the headspace
of pepper samples. This work demonstrates that even with a
low (or perhaps less than desirable) acquisition rate, it is
still possible to perform GC GC/qMS analysis with very
good peak identification assignments. The use of GC GC
suggests the possibility of obtaining, in an easy and effective
way, very high spectral quality for many compounds,
providing an important tool for essential oil research and this
may allow the development of a new database of spectra,
from either qMS or TOFMS instruments.
A table constructed for GC GC/qMS data, which should
be a readily accessible technique for the research community,
which includes retention indices in the first dimension but
measured in a bi-dimensional system, offers an important
step in matrix characterization. Deviations in experimental
retention indices of up to 30 compared with reference values
are considered acceptable; these differences were confirmed

by analysis of some authentic standards. Higher deviation
retention index values can still be acceptable, depending on
the reliability of the spectra and the matches obtained,
especially if standards are co-injected. More work should be
performed to determine the magnitude of the deviations, the
degree of acceptance of those values, and the long-term
reproducibility of a GC GC system with respect to
retention index values. The qMS system can be successfully
used for GC GC analysis and identification purposes with
good accuracy, if a fast scanning (10000 u/s) system is
available, with minimum dwell times, and a restricted mass
scan range. The TOFMS technique allowed the identification
of about five times the number of components found by
using qMS due to the higher acquisition rate, sensitivity and
deconvolution capability of TOFMS, which constitute
important attributes for hyphenation with GC GC.
Acknowledgements
The authors would like to acknowledge and thank Mr Paul
Morrison for his technical assistance, SGE International for
provision of capillary columns, and support from Agilent
Technologies for gas chromatography and quadrupole mass
spectrometry facilities. LECO Corp. is thanked for maintenance of the TOFMS system. Some pepper samples were
obtained from the spice traders: Herbies Spices (Rozelle,
NSW, Australia) and Peter Watson (Brunswick, Victoria,
Australia), who are thanked for their generosity. M. D. R.
Gomes da Silva acknowledges the scholarship from Fundacao Calouste de Gulbenkian, Portugal, and Z. L. Cardeal
was supported by CAPES (Ministerio da Educacao Brazil),
both to spend time as visiting scholars in the ACROSS
Laboratory at RMIT University.
DOI: 10.1002/rcm
REFERENCES
1. Chitwood RL, Pagborn RM, Jennings W. Food Chem. 1983; 11:
201.
2. Ekundayo O, Laakso I, Adegbola R-M, Oguntimein B,
Soforowa A, Hiltunen R. J. Agric. Food Chem. 1988; 36:
880.
3. Martins AP, Salgueiro L, Vila R, Tomi F, Canigueral S,
Casanova J, Proenca da Cunha A, Adzeit T. Phytochemistry
1998; 7: 2019.
4. Zollo PHA, Biyiti L, Tchoumbougnang F, Menut C, Lamaty
G, Bouchet Ph. Flavour Fragr. J. 1998; 13: 107.
5. Jirovetz L, Buchbauer G, Ngassoum MB, Geissler M.
J. Chromatogr. A 2002; 976: 265.
6. Pino J, Sauri-Duch E, Marbot R. Food Chem. 2006; 94: 394.
DOI: 10.1016/j.foodchem.2004.11.040.
7. Jagella T, Grosch W. Eur. Food Res. Technol. 1999; 209: 16.
8. Steinhaus M, Schieberle P. J. Agric. Food Chem. 2005; 53: 6049.
DOI: 10.1021/jf0506030.
11. van Ruth SM, Roozen JP. Food Chem. 1994; 51: 165.
12. Jiang J, Kubota K. J. Agric. Food Chem. 2001; 49: 1353. DOI:
10.1021/jf001166m.
13. van Ruth S, Boscaini E, Mayr D, Pugh J, Posthumus M. Int. J.
Mass Spectrom. 2003; 223/224: 55.
14. Luning PA, de Rijk T, Wichers HJ, Roozen JP. J. Agric. Food
Chem. 1994; 42: 977.
15. Plessi M, Bertelli D, Miglieta F. Lebensm.-Wiss. U.-Technol.

2002; 35: 260. DOI: 10.1006/fstl.2001.0853.
16. Bertsch W. J. High Resol. Chromatogr. 2000; 23: 167.
17. Dalluge J, Beens J, Brinkman UATh. J. Chromatogr. A 2003;
1000: 69. DOI: 10.1016 /S0021-9673(03)00242-5.
18. Gorecki T, Harynuk J, Panic O. J. Sep. Sci. 2004; 27: 359. DOI
10.1002/jssc 200301650.
19. Marriott PJ, Kinghorn RM. Anal. Chem. 1997; 69: 2582.
20. Marriott PJ, Shellie R. Trends Anal. Chem. 2002; 21: 573.
21. Shellie R, Mondello L, Marriott PJ, Dugo G. J. Chromatogr. A
2002; 970: 225.
22. Ryan D, Shellie R, Tranchida P, Casilli A, Mondello L,
Marriott PJ. J. Chromatogr. A 2004; 1054: 57. DOI: 10.1016/
j.chroma.2004.08.057.
23. Williams A, Ryan D, Guasca AO, Marriott PJ, Pang E.
J. Chromatogr. B 2005; 817: 97. DOI: 10.1016/j.chromb.2004. 05.021.
24. Ryan D, Watkins P, Smith J, Allen M, Marriott P. J. Sep. Sci.
2005; 28: 1075.
25. Guilhaus M, Mlynski V, Selby D. Rapid Commun. Mass
Spectrom. 1997; 11: 951.
26. Sumathykutty MA, Rao JM, Padmakumari KP, Narayanan
CS. Flavour Fragr. J. 1999; 14: 279.
27. Orav A, Stulova I, Kailas T, Muurisepp M. J. Agric. Food
Chem. 2004; 52: 2582. DOI: 10.1021/jf030635s.
28. van den Dool H, Kratz PD. J. Chromatogr. 1963; 11, 463.
29. Adams RP. Identification of Essential Oil Components by Gas
Chromatography/Quadrupole Mass Spectrometry, Allured Publishing Corp.: Carol Stream, Illinois, USA, 2001.

DOI: 10.1002/rcm

Comprehensive Two-Dimensional Gas Chromatography/ Mass Spectrometric Analysis of Pepper Volatiles

Cargado por

Información del documento

Título original

Derechos de autor

Formatos disponibles

Compartir este documento

Compartir o incrustar documentos

Opciones para compartir

¿Le pareció útil este documento?

¿Este contenido es inapropiado?

Copyright:

Formatos disponibles

Comprehensive Two-Dimensional Gas Chromatography/ Mass Spectrometric Analysis of Pepper Volatiles

Cargado por

Copyright:

Formatos disponibles

RAPID COMMUNICATIONS IN MASS SPECTROMETRY

Rapid Commun. Mass Spectrom. 2006; 20: 28232836