JOURNAL OF APPLIED POLYMER SCIENCE VOL. 16, PP. 835-847 (1972) Effect of Stirring on the Emulsion Polymerization of Styrene MAMORU NOMURA, MAKOTO HARADA,* WATARU EGUCHI,* and SHINJI NAGATA, Department of Chemical Engineering, Kyoto University, Kyoto, Japan Synopsis The purpose of this paper is to clarify the effect of stirring on the course of emulsion polymerization of, for example, styrene. It establishes the existence of an optimum range of stirring speed and three important factors which must be considered in carrying out emulsion polymerization. (1) Stirring significantly affects the course of reaction in the presence of an imperfectly purified nitrogen atmosphere. Consequently, the number of polymer particles produced and the polymerization rate per particle will be affected. (2) At higher stirring speeds, polymer particles coagulate and coalesce. At lower stir- ring speeds, the reaction rate is controlled by the monomer transport rate from monomer droplets to the aqueous phase. (3) Stirring contributes to the reduction of the number of micelles because emulsifier molecules are adsorbed onto the surfaces of monomer droplets finely dispersed by the stirring. At low emulsifier concentrations near the critical mi- celle concentration, this effect cannot be neglected. INTRODUCTION Stirred tank reactors are widely used for emulsion polymerization on an industrial scale. It is often observed that the reaction rate and the quality of the polymer produced are affected by the stirring conditions, but as this effect is very complicated, it has not been well understood. Therefore, from the standpoint of reactor design and reactor scale-up, it is very im- portant to know what kind and degree of stirring is required for emulsion polymerization. Shunmukham studied the effect of stirring on the emulsion polymeriza- tion of styrene and concluded that violent agitation diminished the poly- merization rate. (2) pyrogallol solution; (3) HaSOr; (4) CaClz; (5) electric furnace; (6) voltage regulator; (7) feeder for initator; (8) reflux condenser; (9) float; (10) sampling cock; (11) thermometer; (12) baffle; (13) reaction vessel; (14) pressure regulator. was purified by passing nitrogen gas (industrial cylinder nitrogen of 99.9% purity) through an alkaline pyrogallol solution, and the other, by passing it through both an alkaline pyrogallol solution and an electric furnace with copper gauze. Polymerizations carried out with the former were called A, those with the latter were called B, and those with high-purity cylinder nitrogen (better than 99.99% purity) without further purification were called C. The number of polymer particles, the monomer weight fraction in the polymer particles, and the average degree of polymerization were deter- mined in the same way as described in the previous paper.6 Monomer con- versions were determined gravimetrically . When a monomer layer sepa- rates from the emulsion phase, the ratio of monomer to water is not always uniform throughout the reactor, so that samples withdrawn from the bot- tom of the reactor do not represent the mean composition of the reaction mixture. Thus, monomer conversion may be determined from (a/b)/A!, where a and b represent the amounts of polymer and aqueous solution (water + emulsifier + initiator) in the sample, and Mo is the initial ratio of monomer to aqueous solution. In these experiments, the values of a and b were determined by the following procedure. Monomer was added to the sample vrithdrawn from the reactor. After the polymer particles in the sample were fully saturated with the monomer, monomer droplets in the sample were separated by a centrifuge. The amount of the polymer, a, in the sample was determined gravimetrically. The total weight of poly- mer and monomer in the sample was known to be 2.33 X a because the polymer particles were saturated with monomer.6 The amount of aqueous solution, b, in the sample was found by subtracting 2.33 X a from the weight o f the sample. The average diameter of the emulsified monomer droplets was deter- mined by the following procedure. Monomer droplets in a sample with- drawn from the reaction mixture were separated as a cream by centrifuga- tion. The emulsifier concentration in the aqueous phase of the sample was measured by the Epton method.s The average diameter of the monomer droplets was determined from the decrease in emulsifier concentration in the aqueous phase assuming that the emulsifier molecules were adsorbed on 838 NOMURA ET AL. Fig. 3. Effect of stirring on the course of emulsion polymerization under nitrogen atmos- pheres having different purities A, B, and C. monomer droplets in a monomolecular layer and using a, equal to 35 X 10-l6 cm2/molecule, the area per adsorbed emulsifier mlecule. RESULTS AND DISCUSSION Effect of Stirring in the Presence of Imperfectly Purified Nitrogen Emulsion polymerization was carried out under nitrogen at