(653c) Single Particle Modeling of Seed Polymerization Process | AIChE

(653c) Single Particle Modeling of Seed Polymerization Process

Authors 

Lin, W. - Presenter, Carnegie Mellon University
Biegler, L. - Presenter, Carnegie Mellon University
Jacobson, A. - Presenter, Carnegie Mellon University


Seed polymerization technology is recognized as a promising field for novel polymer material development. Its advantages in particle size control and versatile customized properties make it a competitive choice in many areas of application. However, due to the complexity of the seed particle growth phenomenon, many experimental observations remain unexplained, and the relationships between process modification and polymer characteristics are challenging to identify. Modeling intricate particle growth and morphology, which is significant for the design and operation of particulate processes, has not yet been developed.

In order to facilitate the seed polymerization progress, a comprehensive dynamic single particle model is discussed in this paper. To depict the special physical process of seed polymerization, the interactions of the outer mass transfer and the particle internal diffusion are considered with polymer kinetic models. A tie between macro-scale feeding strategy and single particle evolution is built for further purposes of optimal control. In contrast to traditional single particle models, the particle size growth is modeled in a more rigorous way, where the seed property and the dynamic density change are taken into account. More over, dynamic functions are adopted for polymer apparent kinetic rates, which reveal more insight about the morphology formation of these complex polymeric particles.

The model is consisted of a set of differential algebraic equations, and the numerical simulation is performed for a free radical seed polymerization process. The comparison of simulation results and experimental measurements shows consistent agreement both in time and space evolution of the particle growth. Further more, parameter estimation and process control problem are investigated based on the model formulation. Efficient algorithm could be applied with a direct nonlinear programming transcription method, showing its promising capacity for further industrial application interest.