Coalescene and re-dispersion frequency of dispersed droplets in an continuous suspension polymerization stirred tank reactor

To explore the possibility of semi-continous operations and polymer particle size controllings in the suspension polymerization, it is important to clarify the behavior of swarms of droplets in the continuous stirred tank reactor in a comparatively low polymerization region. In this study the following three experiments were carried out: [Exp. I] (1) Two separate flows having the same flow rate; one was a styrene flow with no initiator, and the other was a toluene flow involving a certain amount of initiator, were supplied as dispersion phase. At 360 to 1000 rpm the yield and degree of polymerization were measured in the continuous suspension polymerization. (2) Equi-volume mixture of styrene and toluene was fed as dispersion phase, and the same experiment as mentiond above was carried out. [Exp. II] A colored and a colorless dispersion flow were fed, and the degree of color for each droplet at the effluent was determined by a microscopic photo analysis. [Exp. III] The mixing degree and the effective volume fraction of the dispersed phase in the heterogeneous continuous stirred tank reactor were determined from the tracer response test of the dispersion phase.The following results were obtained; (1) In a heterogeneous continuous operation, the effective volume fraction of the dispersion phase might be reduced when the dispersion phase is lighter than the continuous phase. (2) The effects of revolution rate on coalescence and re-dispersion frequency per droplet were negligible although a marked effect was predicted. (3) Polyvinylalcohol (PVA), a stabilizer for droplets in the suspension polymerization, has a decreasing effect on droplet size. However, np effect from PVA was detected on the coalescence and re-dispersion frequency of the droplets in the low polymerization region. (4) The frequency per droplet of coalescence and re-dispersion was estimated to be about 0.01 min^{-1} from simulation of the polymerization results using the Monte Carlo method