Page 78 - Multidimensional Chromatography

Multidimensional High Resolution Gas Chromatography 69

The analysis of persistent organic pollutants presents particular problems to the
analyst. While instrumental methods exist for determining the bulk chlorinated con-
tent, the variability in toxicity effects means that these are of little use in health
impact assessment. A fully speciated analysis of each individual congener is
required, and the analysis is hampered not only by target species complexity, but the
more concentrated and equally complex sample matrix. In both soils and atmo-
spheric samples, the overwhelming organic background is that of toxicologically
insignificant aliphatic species. The isolation of organochlorine compounds against
this background requires considerable sample preparation prior to analysis, in com-
bination with selective detection (often using electron-capture detectors (ECDs) or
MS) and a high resolution separation. Ideally, a single column would be used for the
universal separation of all congeners of interest; however, after many years of opti-
mization this seems unlikely to emerge (49, 50). Extensive multilaboratory studies of
retention behaviour on a variety of stationary phases (51) have highlighted that it is
not possible to determine all species under a single set of single-column
conditions. A number of approaches to deconvolution have been undertaken, includ-
ing the use of (i) parallel columns to increase probability of isolation on at least one
column, (ii) mass spectrometric deconvolution, (iii) serially coupled columns, and
(iv) two-dimensional GC. While option (iv) requires significant apparatus and
further development, it is the most reliable method on offer at the present time.
Work by Kinghorn et al. has demonstrated a two-dimensional separation of the
PCB, Aroclor 1254. This separation used a non-polar primary column with selected
cuts to a secondary chiral selective column (52). This system utilized a combination
of a Deans switch transfer with a cryogenically cooled intermediate capillary column
which was used to refocus the analytes prior to secondary column analysis. Poorly
resolved single-column peaks were well resolved on application of a second separa-
tion in combination with the refocusing step, with the exception being congener 138,
which may be resolved only through the use of very polar cyanopropyl or liquid
crystal phase columns (53, 54). Chromatograms from this work are shown later in
Chapter 13 (see Figure 13.1).
Of the 209 PCB congeners, 78 are known to exist in two chiral forms. Rather than
chirality based around a central carbon atom, asymmetric substitution of both phenyl
rings leads to axial chirality of all non-planar conformations. Many of these have low
energies of transformation and are able to interconvert by rotation about the central
C–C bond and form racemic mixtures. There are however 10 chiral tri- or tetraortho-
substituted PCBs which have rotational energy barriers sufficiently high to be con-
formationally stable and thus will not undergo racemization–these are known as
atropisomers. There has been recent evidence to show that under certain conditions
the biodegradation of PCBs favours one enantiomer and the relative ratio of the
respective enantiomers may be used to study this phenomenon (since the physico-
chemical and transport properties will not affect the enantiomeric ratios).
It is in the study of this phenomenon where two-dimensional GC offers by far the
most superior method of analysis. The use of chiral selector stationary phases, in
particular modified cyclodextrin types, allows apolar primary and atropisomer selec-
tive secondary separation. Reported two-dimensional methods have been successful