Page 102 - Multidimensional Chromatography

94 Multidimensional Chromatography

solutes, and only occurs at reasonably high oven temperature, it behaves almost as
an unretained solute. In this case, the bleed retention profile may reflect the change
in carrier gas flow in the second column as the temperature increases.

4.4 ORTHOGONALITY OF ANALYSIS

Multidimensional analysis methods rely on exploiting different properties or
responses of solutes based on more than one physical/instrumental process, normally
operated in series. Thus, a separation dimension coupled with a spectroscopic
dimension constitutes a multidimensional analysis (27). The usual value of such a
method is that if components are not uniquely identified in one dimension, then they
will hopefully be identified in the second. Provided that the response bases of both
dimensions are sufficiently different, then an orthogonal analysis results. If the
mechanism of the two dimensions are similar, then we might propose that some
degree of correlation exists, and this may reduce the identification power of the mul-
tidimensional analysis. For example, HPLC–diode array detection (DAD) can be
regarded as orthogonal two-dimensional analysis because there is no correlation
between the HPLC retention and the UV spectra of the components. Cortes has dis-
cussed a variety of coupled separation dimensions involving different chromato-
graphic modes (28). An HPLC–high resolution GC system will likewise be
orthogonal (29), and has been applied to such diverse applications such as oil frac-
tions (30) and food and water analyses (31). The HPLC–capillary electrophoresis
experiment involves two separation dimensions, but solutes respond in each dimen-
sion differently–HPLC according to the distribution constant of solutes between the
mobile and stationary phases, and CE according to electrophoretic mobility (arising
from solute size-to-charge properties). We might then conjecture that a multidimen-
sional experiment employing reversed-phase HPLC and capillary electrochromatog-
raphy on a non-polar phase material may possess a degree of correlation, since there
will exist retention parameters strongly dependent upon partitioning processes into
the reversed-phase stationary material which is common to both dimensions. In con-
trast, for the HPLC–DAD analysis the specificity of analysis then relies upon the
components having uniquely identifiable or assignable spectra.
The most valuable tool for routine orthogonal two-dimensional analysis in GC is
the GC–MS technique, with the dimensionality being shown schematically in
Figure 4.11(a) for a GC–MS system where compound identity is defined by its time
on the column (1D) and mass spectra (2D). Choosing unique or characteristic ions
for overlapping components can produce reliable quantitative estimations of each
component. How do we extend this argument to gas chromatography? In GC,
MDGC is a two-dimension separation method. We might initially suggest that since
we have two GC dimensions, then they must be correlated. However, the two dimen-
sions will not be correlated provided that the mechanism of interaction between the
components of a mixture and each column is different. Clearly, two columns of the
same phase type operated at the same temperature will be correlated, and two com-
ponents overlapping on the first column will not be expected to be separated on the