Automation of gas chromatography instruments. Part I. Automated peak identification in the chromatograms of standard test mixtures

The evaluation of the chromatogram recorded during the gas chromatographic analysis of a standard test mixture is of importance in the assessment of the performance of the underlying GC system. Automated evaluation is essential in the development of software that can be used in systems that can perform totally automated analysis. In this study, an algorithm for automated peak identification in the chromatogram of the standard test mixture has been developed. In the proposed method, a specified test mixture is analyzed and the resultant chromatogram is stored. Peak identification for this experimentally determined chromatogram, which it is assumed will incorporate all recurring problems that might exist for all analyses, is accomplished by matching the measured peaks with peaks in the reference chromatogram. The proposed algorithm is partly based on information theory and partly on chemical knowledge about the standard test mixture. The algorithm may be used to identify highly distorted peaks, peak retention time shifts (both positive and negative with respect to the reference chromatogram), missing peaks, and irregular peak retention times (shifts in both directions for different peaks). The program gives the user the ability to display the chromatogram and calculate four peak parameters from the experimental results of the test mixture under consideration: (i) peak retention time, (ii) the peak area counts, (iii) the peak width, and (iv) the asymmetry factor of the peak. Satisfactory performance is obtained using a set of simulated data that exhibit a range of problems that commonly occur in gas chromatograms as analyses proceed over time.