A procedure based on the power spectrum (PS) model of a multicomponent chromatogram is introduced by which the number m of detectable components (or single-component peaks) and the parameters of the single-component peak such as standard deviation and asymmetry factor can be evaluated. In essence, when fitted to theoretical models, the experimental PS-expressing the chromatographic response variance dependence on the time distance-provides the Information necessary to accept or reject the model and to give the necessary parameter estimations. The procedure is tested by using computer-generated multicomponent chromatograms with Poissonian retention time distribution and random and uncorrelated peak heights, in which density, asymmetry and height distribution are widely varied. How to obtain unbiased PS numerical determination by windowing is also discussed. It Is shown that unbiased parameter estimations are obtained, the only procedure limitation being the approximation made In the evaluation of the single-component peak height dispersion. An example Is given of how a retention time distribution other than the Poissonian can be detected. © 1990, American Chemical Society. All rights reserved.
Fourier Analysis of Multicomponent Chromatograms. Numerical Evaluation of Statistical Parameters
PASTI, Luisa;DONDI, Francesco
1990
Abstract
A procedure based on the power spectrum (PS) model of a multicomponent chromatogram is introduced by which the number m of detectable components (or single-component peaks) and the parameters of the single-component peak such as standard deviation and asymmetry factor can be evaluated. In essence, when fitted to theoretical models, the experimental PS-expressing the chromatographic response variance dependence on the time distance-provides the Information necessary to accept or reject the model and to give the necessary parameter estimations. The procedure is tested by using computer-generated multicomponent chromatograms with Poissonian retention time distribution and random and uncorrelated peak heights, in which density, asymmetry and height distribution are widely varied. How to obtain unbiased PS numerical determination by windowing is also discussed. It Is shown that unbiased parameter estimations are obtained, the only procedure limitation being the approximation made In the evaluation of the single-component peak height dispersion. An example Is given of how a retention time distribution other than the Poissonian can be detected. © 1990, American Chemical Society. All rights reserved.I documenti in SFERA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
