Application of a Multiple Linear Regression Model to Fixed Bands IR Detector Data in GPC-IR Analysis of Polyolefins

An infrared detector based on a set of narrow band optical filters was coupled to a high temperature Gel Permeation Chromatograph (GPC) producing continuous chromatograms of absorbance after the molar mass fractionation. A multiple linear regression (MLR) model was established to relate the measured absorbance to the average octene weight percent in industrial ethylene-octene copolymer samples. This method is compared to univariate and multivariate band ratio models. The application of these models to produce molar mass compositional distributions is also outlined.