Department of Analytical Chemistry, Umeň University, S-901 87 Umeň, Sweden
Great errors are involved in the measurement of asymmetric, overlapping, chromatographic peaks using the perpendicular- drop or tangent skimming algorithms incorporated in most integrators .
Most chromatographic peaks are characterized by some degree of tailing (see Fig. 1) and can be described by the exponentially modified gaussian (EMG) function ;
Deconvolution of this function using the following relationship in which the first derivative of the chromatogram is used (Fig. 2) [3, 4];
recovers the input function (ideal Gaussian) as seen in Fig. 3. The time constant () must be determined before deconvolution is possible. Calculations using empirical equations for total variance (M2) and standard deviation ( G) can be employed .
If semi-enclosed tubular atomizers are utilized, as for the coupled gas chromatography - quartz tube atomic absorption spectrometry (GC-QTAAS) system in this study, the time constant may be theoretically determined .
Three organolead species were separated and detected by a GC-QTAAS system using non-optimal conditions, yielding the chromatogram shown in Fig. 4(a). Following deconvolution, the peak shapes were improved (as seen in Fig. 4(b)), and the accuracy of quantitation was much better (Fig. 5).
|Fig. 4 (a) Original and (b) deconvoluted chromatograms. Elution order is triethylpropyllead, triethylbutyllead and diethyldipropyllead|
|Fig. 5 Relative errors (%) in the determination of the concentrations of (1) triethylpropyllead, (2) triethylbutyllead and (3) diethyldipropyllead from the chromatograms in Fig. 4, using (a) peak area and (b) peak height signal evaluation, with (red bars) and without (blue bars) deconvolution|
The results of this study clearly demonstrate the benefits of deconvolution prior to data evaluation of overlapping, asymmetric, chromatographic peaks.
|Return to the Conference Presentations page||Return to the Atomic Spectrometry Group home page|