Optimisation of data handling and detection conditions for automated chromatographic assay software

Summary The principle of automated chromatographic peak detection and analysis software is summarized, and critical steps are systematically studied. As the only parameter to be entered is the acquisition frequency, evaluation of its effect on software performance is discussed. In the case of relatively noisy chromatograms, it is shown experimentally that numerous points per peak have to be taken, leading to quite fast computer acquisition procedures. The use of discrete Fourier transform filtration techniques can modify peak shapes and a comparative study evaluates the relative errors induced in the shapes and characteristics of the chromatographic profiles. Optimisation of filtering conditions is achieved and it is shown that for a filter position only 2% of the Nyquist frequency no deformation occurs in the chromatographic profile. Detection of the start and finish of chromatographic peaks is optimized according to a simple four step iterative procedure. In the case of simulations, the difference between the values used to simulate peaks and those calculated by the software are less than 1%.