Chemistry Reference
In-Depth Information
y-intercept (based on regression analysis with zero not included) is recommended as
it is a better indicator of the QL at low concentrations than averages derived at higher
concentrations from the residual standard deviation.
Determining the QL is a two-step process. Regardless of the method used to deter-
mine the QL, the limit should be first documented and supported, and then an appropri-
ate number of samples are analyzed at the QL to fully validate the method performance
at the QL. That is, a candidate QL is estimated, such as by signal-to-noise ratio or the
slope of a calibration curve, and once this value is found, it needs to be confirmed by
demonstration of the appropriate response for samples formulated at the LOQ (limit
of quantitation).
When determining limits chromatographically, the efficiency and lifetime of
the column can play a significant role. Figure 4.5 shows two separations performed
on two different columns under equivalent conditions. The columns used are both
designed for the application; however, one (Figure 4.5a) is more efficient (either from
being packed better, or from having fewer secondary interactions leading to peak
broadening) than the other. Lower efficiency leads to broader peaks and lowers the
signal-to-noise ratio. A loss of efficiency can also occur as a column ages. Depending
on the mobile-phase conditions, it is common over time for the base silica to slowly
dissolve, or it is possible for some of the surface ligands to be stripped off, either
0.0005
(a)
0.0
5.0
10.0
15.0
Time in Minutes
(b)
FIgure 4.5
PDA detector response of the chromatographic analysis of an API and impu-
rity at the 0.01% level. Insets show an expanded view of the chromatogram in the region
where the impurity elutes, and the PDA-derived UV spectra for both the API and the impurity.
