Page 95 - Multidimensional Chromatography

Orthogonal GC–GC 87

downstream column) as a rapid pulse. The pulse width has been estimated to be as
narrow as 10 ms or less, based on an experiment where a very short piece of column
located between the cryotrap and detector gave peaks about 50 ms wide at their
bases. Since the peak (or slice of a peak–see later) entering the cryotrap has a signif-
icantly broader bandwidth, the focusing process has the effect of concentrating the
packet of solute – increasing its mass flow rate, which is translated into a peak height
increase in the detection step. For instance, if a migrating peak is 40 cm wide, and is
then focused to say 4 mm wide at the cryotrapping zone, we have a 100-fold increase
in concentration. If the subsequent peak at the detector is 2 cm wide after it travels
through a short (fast) column, then the peak will be about 20 times taller than that on
a conventional capillary column operated normally. This argument can also be based
on peak widths in units of s. If the conventional capillary peak is 5 s wide at its base,
and the pulsed peak after the short second column is 200 ms wide, then a peak height
increase of 25-fold is achieved. These improvements are easily attained in the LMCS
experiment. Perhaps the more important effect of the focusing, however, is the
potential for separation or resolution of solutes which are co-trapped in the cryotrap
and collectively pulsed to column 2.
Accepting that the cryofocussing/remobilization process is both effective in the
collection of discrete sections of the effluent from column 1, and very rapid in rein-
jection to column 2, we can now propose a number of ways of using the LMCS
device in multidimensional gas chromatography modes.

4.3.3.1 Mode 1: Selective or Targeted
Multidimensional Gas Chromatography

Many approaches have been described in the literature for achieving MDGC.
Perhaps a common theme among these can be summarized as the need to isolate a
small region of the first column separation and pass it to the second column, where
we seek to attain greater separation. The analyst then only has to decide which
choice of column will produce the desired result. Conventionally, however, the
MDGC experiment has been much more involved and complex to implement than
suggested by this simplified description of the process. The classical experiment will
use flow switching valves or pressure-balanced systems that are not trivial to design,
construct, operate and maintain. For these reasons, the number of analysts using
MDGC on a routine basis is still very small, and often restricted to those laboratories
where considerable expertise is available. With reference to Figure 4.1, the experi-
ment will normally involve the following steps: (a) setting up the MDGC system; (b)
running a normal GC analysis with the effluent directed to detector 1 through the
non-retaining transfer line (T) (thus there is no time delay between peaks at the
detector and their entry into the interface (I); (c) setting up the time sequence for
heartcutting selected zones to 2D; (d) deciding if some trapping mechanism is
required at the head of 2D; (e) running the experiment. The stimulus provided to the
interface/value device (I/V) will be the event controlled by the multidimensional
control unit that effects the heartcutting, and may be a valve switch/rotation, a valve