Modeling Of Particle Size Distribution In Butyl Acrylate Emulsion Polymerization

Abstract

The goal of the present work is to develop, solve and validate dynamic models for butyl newlineacrylate emulsion polymerization in batch and semi-batch reactors. The polymerization system newlinethat the model simulates include iso-thermal, well-mixed batch and semi-batch reactors with newlinebutyl acrylate (monomer), potassium per sulfate (initiator), sodium dodecyl sulfate (emulsifier), newlinewater as a dispersion medium and polybutyl acrylate (polymer). The model describes and newlinepredicts the particle size distribution in this polymerization system. The population balance newlineequations are solved efficiently using orthogonal collocation. The model is validated against the newlineexperimental data taken from literature. newlineThe model s predictions for the batch reactor were in close agreement with the newlineexperimental data of Coen et al. [12] and Sajjadi and Brooks [78]. The model s reliability was newlinedemonstrated by successfully simulating the experimental data for number of variables which newlineincluded the number of particles formed at different emulsifier and initiator amounts, variation of newlineparticle diameter with time and the variation of conversion with time at one initiator amount and newlineone emulsifier amount, number average, volume average and weight average diameters at newlinedifferent emulsifier, initiator and monomer amounts. Also, the model s predictions for the full newlineparticle size distribution and the variables characterizing the broadness of the distribution viz. newlinestandard deviation, polydispersity index and range were discussed. newlineFor semi-batch reactor, initially a parametric sensitivity study was conducted. Three newlineparameters were identified to influence the nucleation and growth kinetics. They were initiator newlinedecomposition efficiency factor (f), the area occupied by an emulsifier molecule on the surface newlineof micelles and particles (aem=aep) and the propagation rate constant (kP). It was found that newlineslower kinetics was predicted by lowering the values of these variables keeping all the other newlinevariables at the base values as reported in Table 3.1. The model s predictions