Closed-loop in-silico control of a two-stage emulsion polymerization to obtain desired particle morphologies

Abstract Particle morphology strongly affects the performance of waterborne polymer dispersions, which are products-by-process materials produced by emulsion polymerization, a process prone to suffer run-to-run irreproducibility. Therefore, on-line control of particle morphology would be tremendously beneficial, but this is an unsolved issue because particle morphology is neither on-line measurable nor observable from available online monitoring techniques. This article explores in-silico the possibility to overcome this problem by using a mathematical model able to calculate the particle morphology from the monomer conversion and reactor temperature as a “soft sensor”. These estimates are used in combination with an economic nonlinear model predictive controller. It is shown that this controller is able to achieve a good control of the particle morphology even in the presence of serious disturbances such as the presence of a strong inhibition or the failure of the reactor cooling system. Under these circumstances, neither open-loop control nor tracking of off-line optimized reaction trajectories are able to control the particle morphology tightly. The limits of the soft sensor are investigated by considering non-detectable disturbances such as the use of a seed of unknown glass transition temperature and the benefits of a hypothetical online morphology measurement available for control is explored.