DECONVOLUTION OF OVERLAPPING SKEWED PEAKS

Asymmetrical chromatographic peaks can be well modeled by the exponentially modified Gaussian model (EMG). In this study, an analytical expression is derived to give the model for the sharpened overlapping peaks when the deconvolution is performed by another EMG peak as a sharpening function. By means of this model, it can easily be explained what kind of peak shapes can be expected after deconvolution. The validity of the model derived was tested by numerical deconvolution methods-the relaxation and the Fourier deconvolution. The model derived here proves that it is impossible to conserve the peak shape during deconvolution. The integration of the peak areas is almost always burdened with error. The model gives an explanation of why side lobes and spurious peaks emerge during the deconvolution procedure.