Adaptive control of the molecular weight-distribution in living anionic polymerization processes

Abstract

A model reference control strategy is discussed for the production of monodisperse polymers and polymers of any pre-defined molecular weight distribution fran ‘living’ anionic polymerizations carried out in a small-scale tubular reactor system. Initially it was assumed that when a fixed (monomer)/(initiator) ratio was introduced into the reactor that it was possible to produce monodisperse polymer with a predictable average molecular weight and further that by perturbing the (monomer)/(initiator) ratio in a prescribed monomer with time that an integral product of any molecular weight distribution (MWD) could be achieved. Experimental results indicate that the limitations of this simple approach were (i) the (initiator) was severely altered by adventitious side reactions, (ii) axial dispersion in the reactor and a finite reaction initiation rate resulted in an undesirable broadening effect on the MWD. To overcome these limitations a spectrophotometer has been incorporated into an input adaptive loop to control the living chain end concentration in the reactor and the axial dispersion effects on the MWD as deduced from gel permeation chromatography measurements have been used to adjust the input feed rate of the reactants; both modifications are currently being tested and are outlined here. In principle perturbing the feed to a polymerization reactor can be applied to any polymerization process to alter the MWD envelope of the product but the efficacy of this approach is sensitive to the mechanism of the polymerization reaction and the reactor configuration.