Quantitative Analysis in Pharmaceutical Analysis - Chiral Separations by Capillary Electrophoresis - page 332

in aqueous phosphate buffer, the isomers were evaluated in nonaqueous systems.
Preferably heptakis-(2,3- O -diacetyl-6- O -sulfo)- -CD) was used,

which resulted in very good resolution in a rather short analysis time. The mixtures
of diastereomers and enantiomers of a given compound (1/2 and 3/4) can be per-
formed in one run (Figure 11.8a) [34].
Nevertheless, the more robust aqueous method, which was employed for the
determination of small amounts of racemic trans substance in racemic cis substance,
was validated according to ICH guidelines [21] and Ph. Eur . [17] with respect to
specii city, linearity, and range, as well as LOD and LOQ, precision, accuracy, and
robustness. With respect to specii city, all four isomers were well separated (Figure
11.8b). By spiking the samples with the individual racemic substances the peaks of
the electropherogams were assigned. In order to determine the trans isomer as the
diastereomeric impurity in samples of the cis isomer, a calibration was performed
using eight concentration levels in the range of 5.0-50.0

β -CD (HDAS- β

g/mL which corresponds
to a 0.25%-2.50% content of the trans substance in a cis sample of 2.0 mg/mL. Each
standard was injected three times randomly and the mean corrected peak area of
each enantiomer was evaluated. Calibration curve was obtained by plotting the per-
centage of diastereomeric impurity against the peak area ratios. The good calibra-
tion data are demonstrated by the linear equation y = 0.0097 x + 0.0002 and showed
a linearity coni rmed by the coefi cient of determination r 2 = 0.9989. Figure 11.9
shows the residuals plotted against the calculated y -values from the regression line.
It can be seen that there is only random l uctuation in the residuals (i.e., no trends
are observed) corroborating the linearity of the method. For the highest concentra-
tions, a slightly larger elevation can be seen. This is most probably caused by random
errors in the sample preparation.

μ

Residual analysis

0.0006

0.0004

0.0002

0

0 0.005 0.01 0.015 0.02 0.025 0.03

-0.0002

-0.0004

-0.0006

y (calculated)

FIGURE 11.9 The plot of the residuals against the calculated y -values from the regression
line of the determination of aziridine diastereomers.