AppNote 4/2005
A Selectable Single or Multidimensional GC System with Heart-Cut
Fraction Collection and Dual Detection
for Trace Analysis of Complex Samples
Edward A. Pfannkoch and Jacqueline A. Whitecavage
Gerstel, Inc., Caton Research Center, 1510 Caton Center Drive,
Suite H, Baltimore, MD 21227, USA
KEYWORDS
Multidimensional, GC-GC, fast GC, flavor, fragrance,
sulfur, PFPD
ABSTRACT
Identification of important trace components in complex
samples like fragrances, natural products, petroleum fractions or polymers can be challenging. Achieving the mass
on column and resolution necessary to locate and identify
trace components using a single chromatographic separation
can be difficult if not impossible. Parallel detection with a
selective detector and MSD can help locate the region of
interest within the complex chromatogram, but lack of sufficient resolution may still preclude reliable identifications.
Heartcutting and 2D GC using columns with different polarities can significantly improve the resolution of complex
regions and is the most powerful approach for trace component identification. This configuration requires a selective
detector after both the pre-column and analytical column to
track the components of interest through both separations,
adding complexity.
We used Stir Bar Sorptive Extraction as a solventless
means to introduce sufficient mass of sample onto the
pre-column. When additional mass is necessary, a cryotrap after the pre-column can function as a fraction
collector to accumulate fractions from many replicates
of the sample. We coupled low thermal mass (LTM)
GC column modules with dissimilar column phases
using a software-controlled column switching device to
perform heartcutting 2D-GC on two sample types. This
system was configured with dual detectors to track and
identify trace components of interest and a 6-port high
temperature valve to allow selecting single column
or multidimensional operation with dual detection in
either configuration.
Separation and identification of selected trace flavor components from alcoholic beverages and sulfur
compounds from food products demonstrated the effectiveness of this system. The main advantages of this
configuration were the simple selection of single or two
dimensional operation, the ability to collect multiple
fractions to maximize signal from trace components in
the second dimension, and parallel MSD and selective
detection to assure correct identification of components
of interest.
INTRODUCTION
Gas chromatography (GC) is widely used to analyze
samples with complex volatiles profiles which are difficult to resolve on a single chromatographic column.
When coupled to a mass selective detector (MSD)
the goal is often tentative identification of specific
compounds present in the sample. Issues arise when
complex samples are analyzed using GC/MS and coelution interferes with the mass spectral identification
of individual compounds.
In many cases, it is only necessary to further resolve groups of peaks in a region of interest. Regions
of interest may be known from prior studies, or may be
identified by using selective detection techniques such
as olfactometry for odors or element-specific detectors
like the Pulsed Flame Photometric Detector (PFPD) for
sulfur. Resolving these regions can be readily accomplished using standard GC equipment and data analysis
software by connecting two columns with different
phases and selectively transferring compounds from
one column into the other. This approach is commonly
known as classic heartcut GC-GC.
In this study, we coupled two GC columns with
dissimilar phases in independently programmable low
thermal mass (LTM) heating modules using a softwarecontrolled GC-GC heartcut system with a cryogenic
trap at the head of the second column. A manual 6-port
valve in the system allows the user to select whether
the main detectors (MSD, Olfactory or PFPD) are inline after the pre-column or main column simplifying
switching between single and multidimensional operation. The performance of the system was illustrated
by identifying aroma components in Amaretto, and
organosulfur species in sauerkraut and soy sauce.
EXPERIMENTAL
Instrumentation. Analyses were performed on a 6890
GC equipped with a 5973 mass selective detector, a
flame ionization detector (Agilent Technologies), a
5380 PFPD pulsed flame photometric detector (OI
Analytical), PTV inlet (CIS 4, GERSTEL) and thermal desorption unit with autosampler (TDS 2 & TDS
A, GERSTEL), a multi-dimensional column switching
system with cryotrap (MCS 2/CTS 2, GERSTEL), an
olfactory detection port (ODP, GERSTEL), a six-port
high temperature valve (Valco), and dual Modular Accelerated Column Heaters (MACH, GERSTEL).
AN/2005/04 - 2
Figure 1. Schematic diagram of GC-GC system with 6-port valve for selecting single or multidimensional
operation.
AN/2005/04 - 3
Analysis conditions.
TDS 2
splitless,
20°C, 60°C/min, 250°C (5 min)
PTV
0.2 min solvent vent (50 mL/min),
splitless
-120°C, 12°C/s, 280°C (3 min)
Pre-Column 15 m RTX-5 (Restek), LTM
di= 0.25 mm, df= 0.25 μm
Main Column 30 m INNOWax (Agilent), LTM
di= 0.25 mm, df= 0.25 μm
Initial flows: 1.6 mL/min for pre-column
1.5 mL/min for main-column
MACH:
50°C; 20°C/min; 280°C (5 min)
(Amaretto)
for pre-column
40°C (10 min); 10°C/min;
220°C (5 min) for main-column
MACH:
40°C (1 min); 25°C/min;
(Soy sauce) 280°C (5 min) for pre-column
60°C (8 min); 60°C/min;
120°C (2 min); 7°C/min;
220°C (5 min) for main-column
MACH:
40°C (1 min); 15°C/min;
(Sauerkraut- 280°C (5 min) for pre-column
juice)
40°C (12 min); 10°C/min;
220°C (5 min) for main-column
Sample preparation. 10 mLs of sample diluted 1:10 in
bottled H2O (Amaretto) or prepared neat (soy sauce &
sauerkraut juice) were pipetted into a 10 mL headspace
vial. Stir Bar Sorptive Extraction was performed by
adding a Twister to the sample, capping the vial, and
extracting for 1 hr at room temperature with 1200 rpm
stirring. The stir bars were removed, rinsed with bottled
H2O, blotted dry and placed into conditioned thermal
desorption tubes for analysis.
RESULTS AND DISCUSSION
Identification of aromas in Amaretto liquor. We configured the selectable single/multidimensional system
with dual detection (Olfactory Detection Port/MSD) to
identify aroma compounds in Amaretto liquor. A Twister stir bar used to extract the 1:10 diluted Amaretto
was first analyzed with the single dimension system
to identify regions in the chromatogram containing
characteristic aroma compounds (Figure 2). The group
of peaks eluting near 7 minutes contained a spicy citrus
aroma.
spicy
citrus
Abundance
9000000
7000000
6000000
5000000
sweet almond
8000000
4000000
vanilla
3000000
2000000
1000000
Time-->
5.00
6.00
7.00
8.00
9.00
10.00
11.00
12.00
Figure 2. Single dimension separation of Amaretto flavor extract on a 30m x 0.25mm x 0.25μm DB-5MS
column with main fragrance regions identified.
AN/2005/04 - 4
The 6-port valve was then turned to the multidimensional configuration and a second Twister stir bar was
analyzed with a heart cut in the 7.9-8.5 minute region
to transfer the group of peaks to the cold trap at the
head of the second (main) column. Heating the cold
trap to start the second dimension separation (Figure
3) gave a good separation of peaks, allowing identification of decanal as the compound responsible for
the spicy citrus aroma. In addition, seven more trace
aroma compounds could be detected in this same heart
cut fraction (Table 1).
Abundance
Heart Cut
7.9-8.5 min
500000
A
400000
300000
200000
100000
800000
600000
400000
14.00
16.00
Ethyl benzoate
1000000
12.00
Menthol
10.00
Decanal
8.00
"spicy citrus"
6.00
Ethyl octanoate
Menthone
Time-->
4.00
Abundance
B
200000
Time-->
16.00
18.00
2 3
20.00
1
4
5
22.00
6 7
24.00
Figure 3. (A) Pre-column separation of Amaretto extract on a 30m x 0.25mm x 0.25μm DB-5MS column with
heartcut region (7.9-8.5 minutes) highlighted. (B) Main column separation of heartcut region with major “spicy
citrus” aroma identified as decanal. Descriptors for additional aroma regions are listed in Table 1.
Table 1. Amaretto – 1 cut (3 stir bars).
Peak No.
Retention Time
Descriptor
Compound I.D.
Match Quality
1
19.04
Floral
Ethyloctanoate
94
2
19.98
Spicy citrus
Decanal
91
3
20.27
Green
Unidentified
4
20.45
Peppery
Unidentified
5
21.03
Rubbery
Unidentified
6
22.44
Oily
Unidentified
7
22.72
Musty
Unidentified
AN/2005/04 - 5
Identification of sulfur compounds in sauerkraut juice.
We then configured the selectable single/multidimensional system with dual detection (PFPD/MSD) to
identify sulfur compounds in foods. Figure 4 shows
the first dimension separation of a Twister extract of
sauerkraut juice. Over 65 sulfur compounds were detected in this extract.
Abundance
6000000
The 6-port valve was then turned to the multidimensional configuration and a second Twister stir bar was
analyzed with a heart cut in the 8.3-9.3 minute region
to transfer a group of peaks containing sulfur species to
the cold trap at the head of the second (main) column.
Heating the cold trap to start the second dimension
separation (Figure 5) gave a good separation of peaks,
allowing identification of trithiolane (Figure 6) and
several other compounds in this fraction.
4
A
5000000
4000000
3000000
1
2
2000000
7
3
1000000
Time-->
Abundance
8
5
2.00
4.00
6
6.00
8.00
1 2 3 4 56 7 8
9
10.00
12.00
14.00
1
2
3
4
5
6
9
1000000
800000
600000
7
8
9
400000
16.00
18.00
20.00
Carbon disulfide
Dimethyldisulfide
Allylisothiocyanate
Dimethyltrisulfide
4-Methylthiobutane nitrile
Methyl (methylthio)methyl
disulfide
Dimethyltetrasulfide
1-Isothiocyanato propane
2-Phenylisothiocyanate
B
200000
Time-->
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00
Figure 4. Single dimension separation of commercial sauerkraut juice extract with TIC (A) and PFPD (B)
response. Peaks corresponding to main organosulfur compounds are numbered.
AN/2005/04 - 6
Abundance
150000
Heart Cut
8.3-9.3 min
A
100000
50000
Time-->
Abundance
2.00
4.00
6.00
10.00
8.00
12.00
14.00
16.00
18.00
3
B
7
200000
6
4
100000
5
8
9
1
2
22.00
Time-->
Abundance
24.00
26.00
28.00
7e+07
1
2
3
4
5
6
7
8
9
6e+07
5e+07
4e+07
3e+07
2e+07
30.00
32.00
Trithiolane
4-Methylthiobutane nitrile
1-Isothiocyanato propane
Eugenol isomer
1-Decanoic acid
Geranic acid
Indole
3-Methyl-1H-indole
Benzenepropanoic acid
C
1e+07
Time-->
22.00
24.00
26.00
28.00
30.00
32.00
Figure 5. Pre-column separation of sauerkraut juice extract with heartcut region (8.3-9.3 minutes) highlighted.
Main column separation TIC (B) and PFPD response (C) of heartcut region with three sulfur and six additional
compounds identified.
AN/2005/04 - 7
Abundance
8000
4044
124
78
6000
4000
60
2000
m/z-->
Abundance
40
60
81
100
80
120
78
124
S
8000
6000
45
4000
S
60
S
2000
m/z-->
40
60
80
100
120
Figure 6. Wiley138 library match for trithiolane
detected in sauerkraut juice.
Fast multidimensional GC with fraction collection in
soy sauce. We used the same selectable single/multidimensional system with dual detection (PFPD/MSD) to
identify sulfur compounds in soy sauce. Figure 7 shows
the first dimension separation of a Twister extract of
soy sauce run with fast heating (50C/min). Over 30
sulfur compounds were detected in this extract.
The 6-port valve was then turned to the multidimensional configuration and a second Twister stir bar
was analyzed with a heart cut in the 5.51-5.92 minute
region containing several sulfur compounds. This region was chosen because of the major interference from
a large benzoic acid peak. Heating the cold trap at the
head of the main column to start the second dimension
separation (Figure 8) gave a good separation of peaks,
all well resolved from the benzoic acid interference.
Several compounds were identified in this fraction,
but despite obtaining better mass spectra of the sulfur
species no convincing mass spectral library matches
were obtained.
Abundance
A
1.4e+07
1.2e+07
1e+07
8000000
6000000
4000000
2000000
4.00
6.00
4e+08
3e+08
2e+08
1e+08
Time-->
8.00
10.00
B
2,2-Diethyl-1,3-benzoxathiole (64)
Time--> 2.00
Abundance
2.00
4.00
6.00
8.00
10.00
Figure 7. Single dimension separation of soy sauce extract with TIC (A) and PFPD (B) response. Only one
organosulfur compound could be tentatively identified.
AN/2005/04 - 8
Abundance
Heart Cut
5.51-5.92 min
600000
A
500000
400000
300000
200000
100000
Time-->
Abundance
2.00
4.00
6.00
5 6,7
600000
8.00
9
8
B
500000
2
400000
1
300000
34
200000
100000
Time--> 10.00
Abundance
12.00
1
14.00
2
16.00
20.00
22.00
4
7e+07
1
2
3
4
5
6
7
8
9
6e+07
5e+07
3
4e+07
18.00
3e+07
2e+07
C
Sulfur 1
Sulfur 2
Sulfur 3 (Fig. 9)
Sulfur 4
Ethyl benzoate (87)
Phenethyl acetate (90)
Ethyl-3-pyridine carboxylate (95)
3-Phenyl furan (91)
Benzoic acid (91)
1e+07
Time--> 10.00
12.00
14.00
16.00
18.00
20.00
22.00
Figure 8. Fast pre-column separation of soy sauce extract with heartcut region (5.51-5.92 minutes) highlighted.
Main column separation TIC (B) and PFPD response (C) of heartcut region with three sulfur and six additional
compounds identified.
AN/2005/04 - 9
Abundance
95 110
67
8000
6000
41
81
55
4000
- CH3SH
2000
m/z-->
Abundance
158
143
20
40
80
60
100
120
140
160
120
140
160
95
67
81
8000
6000
55
27
2000
m/z-->
110
41
4000
20
40
60
80
100
Figure 9. NIST98 library match for trace sulfur compound detected in soy sauce. Several sulfur compounds
exhibited loss of 48 AMU, perhaps consistent with loss of CH3SH fragment as shown.
To illustrate the fraction collection capability on the
cold trap, three Twister stir bars were desorbed simultaneously in a single desorption tube and the heart cut
fraction was trapped in the cold trap. Two more tubes
containing three Twisters each were subsequently
desorbed with the same heart cut region. Figure 10
shows the additional signal available in the second
dimension separation after accumulating the fractions
from nine Twisters.
AN/2005/04 - 10
Abundance
A
1000000
800000
*
600000
400000
200000
Time-->
Abundance
12.00
11.00
14.00
13.00
15.00
16.00
105
4000
61
40
3000
74
1000
88 101
109 122 133
69
55
51
95
81
50
m/z-->
Abundance
B
162
117
44
2000
17.00
75
129
138
100
125
150
166
150
175
105
30000
C
162
61
117
74
20000
41
45
10000
88
69
81
65
50
122
109
5155
35
m/z-->
101
75
95
100
133 150
129
166
138
125
150
175
Figure 10. Overlay of heartcut region (5.51-5.92) TIC from single Twister and 3x3 Twisters (with cold fraction
collection) illustrating the additional MS response possible with fraction collection. Mass spectra from single
Twister (B) and nine Twisters (C) also show the increased abundance across the spectrum.
AN/2005/04 - 11
CONCLUSIONS
The major steps to trace compound identification by
heartcut GC-GC are:
- Introduce sufficient mass of the compounds of interest into the column to obtain adequate response
at the selective detector (ODP or PFPD) and the
MSD.
- Identify the regions of the chromatogram in which
target compounds elute using a selective detector.
- Reconfigure the GC for heartcut GC-GC.
- Heart cut regions containing target compounds and
interfering matrix components onto a second, orthogonal GC column to separate matrix interferences
from the compounds of interest.
- Identify the resolved components using selective
detection and MSD in parallel.
Identification of odors in complex samples can be most
effectively accomplished by multidimensional GC-GC/
MS if an olfactometry detector using the human nose
is used to help pinpoint the compounds responsible for
the odor on both the pre-column and main column.
Using this approach, we identified decanal and ethyl
octanoate as contributors to the characteristic odor
associated with Amaretto, in addition to the expected
benzaldehyde and vanillin.
Identification of organosulfur compounds in foods
can be most effectively accomplished by multidimensional GC-GC/MS if a PFPD configured for sulfur response is used to help pinpoint the compounds on both
the pre-column and main column. Using this approach,
over 65 sulfur species were detected in sauerkraut juice
extract, and eleven could be tentatively identified by
MS library matching. Over 30 sulfur species were detected in soy sauce extract with only one tentatively
identified based on MS library matching. Mass spectra
were obtained for most other sulfur compounds, and
structure elucidation is ongoing.
AN/2005/04 - 12
AN/2005/04 - 13
GERSTEL GmbH & Co. KG
Eberhard-Gerstel-Platz 1
45473 Mülheim an der Ruhr
Germany
+49 (0) 208 - 7 65 03-0
+49 (0) 208 - 7 65 03 33
[email protected]
www.gerstel.com
GERSTEL Worldwide
GERSTEL LLP
Level 25, North Tower
One Raffles Quay
Singapore 048583
+65 6622 5486
+65 6622 5999
[email protected]
www.gerstel.com
GERSTEL Brasil
Av. Pascoal da Rocha Falcão, 367
04785-000 São Paulo - SP Brasil
+55 (11)5665-8931
+55 (11)5666-9084
[email protected]
www.gerstel.com.br
GERSTEL K.K.
1-3-1 Nakane, Meguro-ku
Tokyo 152-0031
SMBC Toritsudai Ekimae Bldg 4F
Japan
+81 3 5731 5321
+81 3 5731 5322
[email protected]
www.gerstel.co.jp
D QM-SY
ST
IFIE
RT
CE
GERSTEL AG
Wassergrabe 27
CH-6210 Sursee
Switzerland
+41 (41) 9 21 97 23
+41 (41) 9 21 97 25
[email protected]
www.gerstel.ch
EM
GERSTEL, Inc.
701 Digital Drive, Suite J
Linthicum, MD 21090
USA
+1 (410) 247 5885
+1 (410) 247 5887
[email protected]
www.gerstelus.com
R
Information, descriptions and specifications in this
Publication are subject to change without notice.
GERSTEL, GRAPHPACK and TWISTER are registered
trademarks of GERSTEL GmbH & Co. KG.
O
9001:20
0
FI
C AT N O . 0
E
IS
CE
TI
08
86
D
ISO 9001
© Copyright by GERSTEL GmbH & Co. KG
Awarded for the
active pursuit of
environmental sustainability