(217ac) Dynamic Modeling and Optimization of An Industrial Semibatch Acrylonitrile-Vinyl Chloride Copolymerization Reactor

Dynamic Modeling and Optimization of an Industrial Semibatch Acrylonitrile-Vinyl Chloride Copolymerization Reactor

Copolymers of acrylonitrile and vinyl chloride are used in the manufacture of flame retardant modacrylic fibers. Modacrylics have a wide range of uses in textiles, wig materials, carpeting and other applications that require flameproof fibers. It is known that such copolymers can be produced by emulsion polymerization [1,2]. However apart from couple of patents, there is no analytical model available for emulsion copolymerization of acrylonitrile and vinyl chloride in the open literature.

This work reports a mathematical model for emulsion copolymerization of acrylonitrile (AN), and vinyl chloride monomer (VCM) in a semibatch reactor. The dynamic model predicts average properties such as monomer conversion, copolymer compositon, and the polymerization rate. The modeling equations describe the concentrations of the monomers, radicals and other components in aqueous, particle and monomer droplet phases. Free radicals form in the water phase by redox initiation using the persulfate initiator. Kinetics are governed by free radical polymerization mechanisms. Micellar and homogeneous nucleation mechanisms have been considered for the generation of particles and the capture of radicals by the micelles and particles are modeled after the literature [3,4]. It is assumed that thermodynamic equilibrium exists between the VCM droplet and the growing polymer particles. However the more soluble monomer AN does not form any droplets and is all soluble in the water phase. Mass transfer limitations from aqueous phase to particle phase are also considered for each monomer, and the corresponding mass transfer coefficients are estimated by parameter estimation techniques. Semibatch reactor is initially loaded with all the VCM and some amount of AN. Next the acrylonitrile monomer is continuously fed in a semi-batch mode until the end of the batch. The important model parameters such as partition and mass transfer coefficients were estimated and the theoretical model was validated against experimental data. Experimental data consisted of AN and VCM conversion values measured under different AN feeding strategies implemented on an industrial research laboratory scale reactor. Model predictions were found to be in close agreement with experimental data. Next dynamic optimization is performed to maximize AN conversion subject to AN to VCM ratio in the polymer close to one. Optimization computes the initial optimal amount of AN placed in the reactor and the best AN feeding policy i.e. AN feed rate (g/min) throughout the reaction time. When implemented on the experimental system the optimal policy resulted in AN conversion above 90% and VCM conversion around 60%. These conversion levels meet the desired operational targets and the product fiber prepared by this comonomer was found to have desired product specifications such as limiting oxygen index, LOI. The dynamic model developed here is suited for real-time optimization and control purposes.

1. Process for the production of acrylonitrile-vinyl chloride copolymers. United States Patent 3936511.
2. Process for the production of acrylonitrile-vinyl-chloride copolymers with improved whiteness. United States Patent 4118556.
3. Hansen, F. K. And Ugelstad, J. “Particle nucleation in emulsion polymerization. I. A theory for  homogeneous nucleation”. J. Polym. Sci., Polym. Chem. Ed. 16, 1953, 1978.
4. Washington , I.D, Duever, T.A. and Penlidis, A. “Mathematical Modeling of Acrylonitrile-Butadiene Emulsion Copolymerization: Model Development and Validation”. Journal of Macromolecular Science, Part A: Pure and Applied Chemistry, 47, 747–769, 2010.