Page 100 - Multidimensional Chromatography

92 Multidimensional Chromatography

pulled apart in the 2D space, but even minor or trace components that would never
have been exposed in the single column are often clearly placed as fully separated
peaks.
The fact that we have peaks within a 2D space implies that where no peak is
found represents a true detector baseline or electronic noise level. In a
conventional petroleum sample, a complex unresolved mixture response causes an
apparent detector baseline rise and fall throughout the GC trace. It is probably a fact
that in this case the true electronic baseline is never obtained. We have instead a
chemical baseline comprising small response to many overlapping components. This
immediately suggests that we should have more confidence in peak area measure-
ments in the GC GC experiment.

4.3.4 CONSIDERATIONS OF RETENTION ON THE SECOND
COLUMN–WRAPAROUND
For unresolved peaks eluting from the first column into the modulator, each pulsed
solute packet delivers a group of components to column 2. It does this every modu-
lation event, e.g. every 4 s. There is no guarantee that each solute will have a
retention less than the modulation time, and so some peaks may not have reached
the detector by the time the next packet of solute is launched into the short fast sec-
ond column. The procedure for generation of the matrix format is illustrated in
Figure 4.10. The column data are taken in blocks of data points equal to the number
of points corresponding to the modulation time, and put into a column. This proce-
dure is repeated until the full data are transformed. If a peak from one modulation
event happens to be retained longer than the sampling period, then it will not be
included in that period’s transformed data block. Let us say the solute has a 2D
retention time of 5 s. Since the data are transformed into matrix form for 2D presen-
tation, using the 4 s modulation time as one matrix edge, then the 5 s retained solute
will have an apparent transformed time of 1 ( 5 4) s, and its peak will appear to
be centred on a time of 1 s in the 2D plot. Thus, while we normally use the labels
‘first dimension retention’ (which is correct) and ‘second dimension retention’ for
the 2D space, the above solute does not have an absolute retention of 1 s on the sec-
ond column. (It may be more appropriate to say ‘apparent retention’ or use a similar
term to reflect that the time does not necessarily have an absolute meaning on the
2D axis.) Here, the solute has ‘wrapped around’ one set of matrix transformations
of data and so appears in the subsequent matrix data line. It is possible to have
solutes wraparound more than once. A useful way to decide if wraparound occurs is
to study peak widths in the second dimension. Those peaks retained more will have
increasingly wide peaks. In a recent study, for instance, of derivatized sterols (24),
the high elution temperature meant that the sterols eluted during the isothermal hold
period (at 280°C). Thus, the later peaks not only became increasingly broad in the
first dimension, but because they were less volatile they were more retained on the
2
second column. A modulation of 4 s was used, and the last sterol had a t R time of