Classification of polymer batch variations by dynamic headspace/capillary gas chromatography/multivariate data analysis

Publisher Website
Google Scholar
Abstract
Gas chromatographic profiles have been generated from different batches of a polypropylene/polyethylene copolymer. The profiles generated from polymer pellets have been obtained by dynamic headspace/capillary gas chromatography analysis. Initially 10 to 40 peaks were chosen at random from the quantitative reports and transferred to a data table. After appropriate scaling the table has been analyzed by a multivariate statistic program, SIMCA (Soft Independent Modeling of Class Analogy) a pattern recognition technique. The method has been used to differentiate batches according to sensory qualities of the final packaging product and changes in polymer peilet production.