Procedure for the alignment of chromatographic profiles applied in the characterization of gasolines.

Abstract

Gasoline is a mixture of hundreds of hydrocarbons obtained from petroleum by fractional distillation, its composition varies depending on the nature of the crude and the technological refining process. Among the physical chemical parameters of quality control the determination of the composition by high-resolution capillary gas chromatography is required. This analysis provides the total contents of paraffins, isoparaffins, aromatics, naphthenes and olefins and constitute valuable information for the control of the industrial process and the quality of the final product, but they are insufficient to be used in forensic problems, specifically, to detect problems related to adulteration. Chemometric pattern recognition techniques are suitable tools for these purposes; but for the application of the techniques it is essential to minimize in the chromatograms the displacement of the retention times that originate in the different measurements. The objective of the work was to develop a chromatographic profiles alignment procedure of gasoline samples using the transformations: Savitky-Golay filter asymmetric least squares and correlation-optimized alignment. Principal components analysis and partial least squares regression techniques were used to verified that the results are not different from those reported when using the percentages of areas of the 70 most representative peaks of the chromatograms. The variances explained in the principal component analysis models were similar: 92 and 91 %.