Chemometrics In The Study Of Polymer Degradation And Life Prediction By Spectroscopic Methods

The correlation of polymer structure with properties has been greatly assisted by the application of statistical methods — chemometrics — to spectroscopic and physical multi-variate data. Chemometrics is a multivariate methodology that allows the information from entire spectra to be used and can be considered as a maximal data compression technique. The extracted “factors” can then be regressed against property data and, given a characterised set of “training” samples, can then be used in a predictive mode. For example, the relationship between the near infrared (MR) spectra of polyolefins and their density and melt flow index (MFI) is already being used for on-line process analysis and control of powder and pellet grades.We will show that simple spectroscopic methods, such as Raman, infrared and near infrared techniques, when coupled with chemometrics, are capable of providing molecular weight and distribution information, and we will illustrate this with reference to poly(dimethylsiloxanes) (PDMS) and cellulosic materials. We will further show the application of the method to tracking structural and property changes in synthetic and natural polymers due to degradation.