Susanne Böhme and Hans-Ulrich Baier, Shimadzu Europa GmbH, Duisburg, Germany.
This article reports on multidimensional gas chromatography using a multi-Deans switch that allows easy heart-cutting. The
main difference in this switch device when compared with others is that the inlet pressure for the second column is not changed
even when switching is performed. This enables unchanged retention times of the peaks eluting after cut positions regardless
of how many cuts have been done.1,2 Therefore, time consuming correction of retention times by changing head pressures is unnecessary. Here, the MDGC-2010 system
was used by coupling a GC-2010 with FID detection to a GCMS-QP2010 Plus.
Figure 1
|
In flavour and fragrance analysis coelutions are often observed using one dimensional chromatography. To resolve those problematic
regions, they are transferred into the second column. For the first dimension, here a Carbowax column (30 m, 0.25 mm i.d.
and 0.25 µm film) was coupled to an RTX-5 (30 m, 0.25 mm, 0.25 µm) in the second dimension. In a first run, the sample is
analysed without cuts. A determination of the concentrations is usually done by performing an area normalization report using
FID detection. To identify the peaks in the first dimension, a FID–MS splitting was done. Parallel detection of FID and MS
in the first dimension was realized by a capillary split connection to feed the effluent into FID and MS simultaneously. Both
columns, the split connection and the second dimensional column were mounted into the MS detector using a special connector.3 The layout is shown in Figure 1. In standby mode (no cuts APC2 (P2) is operated close to natural mid-point pressure) the
injected sample gives a FID chromatogram and MS data simultaneously. This is shown in Figure 2.
Figure 2
|
A perfume sample was injected. The top chromatogram represents the FID data while the bottom is related to MS. The FID–MS
splitting ratio was 1:1/25. The column temperature programme was 60 °C, 1 min, then 20 °C/min to 200 °C, and 7.5 °C/min to
250 °C, 20 min. The initial linear velocity was set to 28 cm/s. No shift was observed between the retention time of the peaks
measured with the FID relative to the ones for MS. For the splitting a deactivated fused silica tube (1 m, 0.15 mm i.d.) was
used.
The advantage of this configuration is obvious: while the FID chromatogram can be used for an area normalization report the
MS data can be used to identify coelution regions. These can be transferred to the second column in a second run. This is
demonstrated in Figure 3 showing the FID data with 3 cut positions (10.82–11 min, 16.51–16.73 min, 21.98–22.14 min). The positions
of the cuts in the chromatogram can be defined easily by click and drag operations in the uncut chromatogram using the MDGCsolution
software. The switching times are then present in the method, which is stored and runs with cuts can then be started.
Figure 3
|
For comparison, Figure 3 shows the chromatogram without cuts (top) to demonstrate the precision of the system performance.
Clearly, no shifts of the uncut peaks are observed. The missing peaks in the bottom chromatogram are transferred into the
second dimension (via L2 , see Figure 1). To prevent interference from the MS–FID splitting of the first dimension with cut
peaks, line L1 has to be blocked for analytes in cut runs. This is achieved by a pressure P2', which is slightly higher than
P2 in standby mode. This reverses the flow in the line connecting the MS–FID splitting between the mid pressure point (M1)
and the split point (S1). All pressures and flows are set only once and stored in the corresponding methods. Therefore, chromatograms
without and with cuts can be run in batch analysis automatically.
Figure 4
|
Figure 4 shows the MS data for the cut run. The three major peaks are visible together with smaller peaks that coeluted in
the first dimension. As a purple line the MS data are shown when no cuts were done and L1 was operated in reverse flow mode
(no analytes should pass L1) to demonstrate that the blockage of the MS splitting line for the analytes is working properly
after the first dimension.
To conclude, this multidimensional GC/GC–MS configuration offers easy and reliable analysis of flavour and fragrance samples
where coelution can be easily identified by a parallel FID–MS detection in the first dimension.
References
1. Luigi Mondello et al., LC GC Europe, 21 (3), 130–137 (2008).
2. Shimadzu News 2/2008: http://www.shimadzu.eu/info/news/default.aspx?News=2/2008
3. Shimadzu Application note 74
Shimadzu Europa GmbH
Albert-Hahn-Str. 6-10, D-47269 Duisburg, Germany
tel. +49 203 76 87 0 fax +49 203 76 66 25
E-mail: shimadzu@shimadzu.eu
Website: http://www.shimadzu.eu/ 
Follow Us
RSS
Bookstore
