Application of factor analysis for quantitative analysis of mixed xylene from fourier transform infrared spectra

Abstract

This research involves a quantitative analysis of mixed xylene that consists of three isomer: 0-, m-, and p-xylene by Fourier transform infrared (FT-IR) spectroscopy. FT-IR spectroscopy is easy to operate, fast and accurate. But the infrared spectra of xylene showed the deviation from Beer-Lambert Law, which caused the conventional quantitative infrared analysis technique inapplicable and caused the error. The mathematical algorithm called factor analysis was applied to solve this problem. In this research the ability of factor analysis was studied with the simulated spectra that generated under various experimental conditions. The factor analysis program was developed for the calculation. The program was tested with the mixture of three isomers of xylene, which know the composition in order to proof the ability of program in the actual application. Finally, the program was used with the unknown mixed xylene. The comparison between the proportions of o-, m-, and p-xylene obtained by this technique was compared with that from gas chromatography (GC) technique. The result indicates the feasibility of factor analysis for quantitative analysis of mixed xylene from FT-IR spectra.