Multiple response optimization of styrene–butadiene rubber emulsion polymerization

A multiple response optimization of styrene–butadiene rubber (SBR) emulsion batch polymerization is proposed. Several properties of latex and rubber were optimized to obtain a particular grade of SBR, namely 1712. Artificial neural networks (ANNs) were employed for the modelling of the following properties: solid content of latex, Mooney viscosity and polydispersity. The training was done by feeding the ANNs with experimental data obtained from a central composite design in which the concentration of some of the polymerization reagents (initiator, activator and chain transfer agent) was varied. The one-dimensional desirability function was used for optimization, in order to obtain a single set of reaction conditions for the multiple responses. With optimum conditions, polymerization experiments were carried out and good agreement was found between predicted and experimental values of the required properties.