Page 111 - Multidimensional Chromatography

Orthogonal GC–GC 103

There remains one key concern for the routine use of GC GC for quantitative
analysis, and that is the need for a real-time, 2D data presentation package with the
ability to present a comprehensive data report providing a list of all peaks found
within the 2D space, along with their heights and areas. Ideally, further interpretive
data might be provided, such as peak symmetry, resolution of neighbouring peaks,
etc. Again, we await the general availability of such software, and perhaps this is the
single most important issue to resolve before the methods outlined here become
more widely accepted. However, the large data file sizes obtained with fast data
acquisition should not be a problem.

4.5.1 COMPARISON OF FIGURES OF MERIT

Chromatographically, any procedure will only be of value if the analytical data or
figures of merit are acceptable to the analyst. It is necessary to demonstrate the relia-
bility of data when using the modulation process. The above points state in very gen-
eral terms why modulation gives advantages, but this still requires demonstration. In
a recent comprehensive study of analytical figures of merit for the analysis of deriva-
tized sterols, we compared the three modes of operation of the LMCS, namely nor-
mal capillary GC, comprehensive GC GC and targeted MDGC (24). This work
revealed that improved precision for raw peak areas is achieved with the selective
LMCS mode (RSD of 2 %; n 5), over that of normal GC and comprehensive gas
chromatography analysis (both with RSD ca. 4 %; n 5). Peaks were generally
25–40 times taller for the selective mode when compared with normal GC peaks,
based on peak height data, with detection limits of approximately 0.004 mg/l com-
pared with 0.1 mg/l for normal capillary gas chromatography and 0.02 mg/l for the
comprehensive mode. Note that peak areas are not increased when using the cryo-
genic system, although we can measure peak areas down to lower levels. The linear-
ity of calibration was best for targeted MDGC (e.g. for the results as obtained by the
procedure shown in the lower trace in Figure 4.7), presumably because there is less
uncertainty in the peak measurement. Calibration curve linearity for comprehensive
GC was similar to normal GC. The better height and area precision for the five repli-
cate analyses for selective MDGC analysis may be due to the better confidence of
measuring peak response when the peak is much higher than the level of detector
noise. The same injection procedure (splitless) was used for all analyses, i.e. with an
autoinjector, and hence the reproducibility of injection should have been similar for
all of the operational modes.
The reason for the differences in comparative figures for the two cryogenic modes
may be related to the fact that comprehensive GC gives a series of pulses for each
peak, and these must be summed to get the total response, while targeted MDGC
gives a single peak for the total component. Normal GC is improved upon by both
the other modes. These results indicate that there is every reason to believe that the
modulated peak response methods provide reliable and accurate data, at least as
good as normal GC, for the analysis of chromatographic peaks.