Discrimination of ethylene/vinyl acetate copolymers with different composition and prediction of the vinyl acetate content in the copolymers using Fourier-transform Raman spectroscopy and multivariate data analysis

Abstract Fourier-transform (FT) Raman spectra have been measured for twelve kinds of ethylene/vinyl acetate (EVA) copolymers with vinyl acetate (VA, the comonomer) varying in the 7 to 44 wt% range. Principal component analysis (PCA) has been carried out for the Raman spectra pretreated by multiplicative scatter correction (MSC) or normalization to discriminate the twelve samples. They are clearly discriminated from one another by a score plot of the PCA factor 1 and factor 2 based upon 36 Raman spectra in the series. Principal component (PC) weight loadings for factor 1 show that the discrimination relies largely upon the bands due to CO stretching and OCO bending modes of VA and those assigned to CC stretching modes of the ethylene unit. A peak corresponding to the CH 2 wagging mode, whose intensity is known to be sensitive to the crystallinity of polyolefines, appears strongly in PC weight loadings for factor 2, indicating that the score on factor 2 reflects the crystallinity of EVA. Partial least squares regression (PLSR) has allowed us to propose a calibration model which predicts the VA content in EVA with the correlation coefficient and standard error of prediction of 0.997 and 0.95 wt%, respectively.