Computer modelling of the degradation of linear polymers

Abstract

Mechanisms of degradation of linear polymers are considered with the help of a computer program, which uses a pseudo-random number generator to carry out a Monte Carlo-like simulation of the processes. Randomly chosen bonds are degraded either at random positions along a molecule or by a number of specific systematic processes, such as breaking the molecule in half or slicing off a fixed number or percentage of units. The resulting molecular weight distribution (MWD) changes are compared to experimental measurements reported in the literature and generated in the authors' own laboratory. Totally random processes do not explain observed movements of the MWD to lower molecular weight or the formation of new peaks. Both processes require specific systematic scission of the polymer chain, such that, for instance, the length of one of the degradation products falls within the range of the original MWD. New peaks require the splitting off of fragments of a systematic length and/or recombination of fragments with each other or with other molecules in the distribution. Peak movement requires the preferential scission of the longest molecules. Random processes tend to follow standard first-order kinetic plots of reciprocal DP vs time, provided all the products of degradation are included in the DP calculation. Systematic scission processes tend to introduce curvature.