Study of the linear range in HPLC analyses with UV detection: Methodology and experimental application to the influence of the analyte UV spectrum

Determination of the linear range is one of the main concerns in validation of an HPLC analysis method. It is particularly important since single poin t calibration will be then used routinely. We proposed an iterative methodo logy to handle this problem. The idea was, at each step, to test statistica lly whether the following point belonged to the same regression line. The m ethodology was then used to evaluate quantitatively the effect on linear ra nge of a shift in detection wavelength or of the detector bandwidth. Althou gh experimental results were globally in accordance with spectroscopic theo ry, magnitudes observed were rather large. So the linear range could vary b y a factor of over 2 with changes in conditions that remained within the ra nge of current practical values. Changes in detection wavelength were limit ed to about fifteen nm around lambda(max) and the detector used was conside red to be representative of modern high-performance UV detectors. The quest ion of how to take consequences in method validation into account was raise d. The solution proposed recommended that the validation was undertaken in conditions as close as possible to those where the method would be conducte d routinely. This means with the same instrumentation and on the product of interest for analyses.