Emulsion polymerization Smith-Ewart theory

In the case of thermal initiation of styrene [79,80], the polymerization rate was found to be proportional to [AIBN] and [KPS] , in good agreement with other data for three- or four-component microemulsions [66,81]. The dependence on AIBN concentration is consistent with the prediction of 0.40 based on the micellar nucleation theory in emulsion polymerization (Smith-Ewart case 2) (see, e.g.. Ref 129). The dependence on KPS concentration lies between this case and the value of 0.5 for solution or bulk polymerization. 

Emulsion Polymerization. Emulsion and suspension reactions are doubly heterogeneous the polymer is insoluble in the monomer and both are insoluble in water. Suspension reactions are similar in behavior to slurry reactors. Oil-soluble initiators are used, so the monomer—polymer droplet is like a small mass reaction. Emulsion polymerizations are more complex. Because the monomer is insoluble in the polymer particle, the simple Smith-Ewart theory does not apply (34). 

Mechanisms. Because of its considerable industrial importance as well as its intrinsic interest, emulsion polymerization of vinyl acetate in the presence of surfactants has been extensively studied (75—77). The Smith-Ewart theory, which describes emulsion polymerization of monomers such as styrene, does not apply to vinyl acetate. Reasons for this are the substantial water solubiUty of vinyl acetate monomer, and the different reactivities of the vinyl acetate and styrene radicals the chain transfer to monomer is much higher for vinyl acetate. The kinetics of the polymerization of vinyl acetate has been studied and mechanisms have been proposed (78—82). 

Based on the Smith-Ewart theory, the number of latex particles formed and the rate of polymerization in Interval II is proportional with the 0,6 power of the emulsifier concentration. This relation was also observed experimentally for the emulsion polymerization of styrene by Bartholomeet al. [51], Dunn and Al-Shahib [52] demonstrated that when the concentrations of the different emulsifiers were selected so that the micellar concentrations were equal, the same number of particles having the same size could be obtained by the same polymerization rates in Interval II in the existence of different emulsifiers [52], The number of micelles formed initially in the polymerization medium increases with the increasing emulsifier concentration. This leads to an increase in the total amount of monomer solubilized by micelles. However, the number of emulsifier molecules in one micelle is constant for a certain type of emulsifier and does not change with the emulsifier concentration. The monomer is distributed into more micelles and thus, the... 

In accordance with the Smith-Ewart theory, the nucleation of particles takes place solely in the monomer-swollen micelles which are transformed into polymer particles [16]. This mechanism is applicable for hydrophobic (macro)mon-omers (see Scheme 2). The initiation of emulsion polymerization is a two-step process. It starts in water with the primary free radicals derived from the water-soluble initiator. The second step occurs in the monomer (macromonomer)-swollen micelles by entered oligomeric radicals. 

The above cited information showed unanimously that, in a mixed-surfactant system of emulsion polymerization, the composition of the mixed surfactant affects the rate of polymerization. Since by Harkins-Smith-Ewart theory, rate of polymerization is proportional to the total number of particles in the system, composition of mixed surfactants seems to affect the efficiency of nucleation. 

Therefore, the nonlinear relationship between rate of polymerization and the total surfactant concentration, as shown in Figure 2, was believed to be caused by a change in micellar size. Thus, the purpose of the present study was to verify the validity of the concept of micellar size effect in emulsion polymerization kinetics. Furthermore, although the Harkins-Smith-Ewart theory of micellar nucleation was proposed in 1948, and has found widespread application ever since, its validity is still challenged even for the case of polymerization of styrene ( ). If micellar... 

Table I shows that the diameters of the polymers from styrene are approximately twice as large as those from 1,4-DVB, and in all experiments with 1,4-DVB at least 6 times more particles have been formed than with styrene. The maximum diameters of spherical particles from 1,4-DVB which can be obtained, are about 500 A. Larger particles are mostly irregularly shaped (see 4.2) whereas in the case of styrene, latex particles of 2000 A and more may be prepared easily. These differences may be explained by a discussion of the SMITH-EWART theory ( ). According to this theory the number N of latex particles formed in the emulsion polymerization of styrene is given by...

Our assumption of prolonged particle formation during the emulsion polymerization of MA can explain the course of v r in mixtures with high emulsifier concentration. For proving this assumption it would be necessary to measure the change in particle number during the course of the emulsion polymerization of MA this has not yet been done. The results of our work on the y-induced emulsion polymerization of MA cannot be interpreted in terms of the Smith-Ewart theory in its simple form (33). Therefore one cannot expect that v r is independent of the monomer-water ratio or proportional to [E] - or to [initator] - (dose rate and initator concentration can be substituted). A quantitative interpretation of the y-induced emulsion polymerization of MA cannot yet be formulated, because of the complexity of the phenomena involved. To make this possible, considerable further work on this subject has to be done. The dependence of on [M] - ... 

According to O Donnell et al. [130], the emulsion polymerization of vinyl acetate follows the Smith-Ewart theory of emulsion polymerization [131] because the rate of polymerization is independent of the total amount of monomer present, the rate is a function of the 0.6th power of the emulsifer concentration, and the rate of emulsion polymerization is a function of the 0.7th power of the initiator concentration instead of the expected 0.4th power. In this work poly(vinyl alcohol), 88% hydrolyzed with a medium molecular weight (i.e., Du Font s Elvanol 52-22), was used as the only externally added emulsifier. Light-scattering studies indicated that this emulsifier formed no aggregates in the aqueous solution. These latter observations may, however, have been made at room temperature and not at the reaction temperature [1]. The conversion versus time curve was essentially linear up to 80% conversion. 

Okamura and Motoyama [132] showed that the emulsion polymerization of vinyl caproate, a monomer of low water-solubility, followed the same pattern as styrene did as far as the Smith-Ewart theory is concerned. 

That there is no generally acceptable model for the emulsion polymerization was emphasized by Min and Ray in their extensive discussion of mathematical modeling of emulsion polymerizations [145]. They list five deviations of the emulsion polymerization of vinyl acetate from the Smith-Ewart theory, which may be a bit different from the points made by Nomura and co-workers [143] and by Friis and Hamielec [144]. These points are as follows ... 

A priori, it seems logical to apply the accepted concepts of conventional emulsion polymerization (with water-soluble initiators) to inverse emulsions using oil-soluble initiators. In fact, only few attempts have been made to apply the Smith-Ewart theory [26,36-38]. The determination of n is difficult here because of the ill-defined stages of the reaction, the unusual kinetics and the broad particle size distribution. The kinetic studies of Vanderhoff et al. [26,29] and Visioli [37] are examples of applying the Smith-Ewart theory to the polymerization of acrylamide and p-vinylbenzenesulfonate in xylene initiated with benzoyl peroxide. The data unexpectedly followed Smith-Ewart Case 1 (n 0.S). It was postulated that radicals were generated in, or enter particles pairwise due in the enhanced water solubility of the benzoyl peroxide by the presence of monomo-. 

Figure 3.1. A schematic representation of the emulsion polymerization system that follows the Smith-Ewart theory. The symbols [S]o and [l]o represent the concentrations of surfactant and initiator, respectively, initially present in the reaction system.

The concise Harkins-Smith-Ewart theory [9-16] delicately describes the key characteristics of emulsion polymerization. However, the difference in colloidal properties (e.g., composition, size, surface charge density, and particle surface area occupied by the adsorbed surfactant) between the monomer-swollen micelles and particle nuclei was not taken into account in the derivation of Eq. (3.4). The probability for micelles or particle nuclei to capture oligomeric radicals in the continuous aqueous phase is simply assumed to be proportional to their total oil-water interfacial area. 

As discussed above, the well-known Smith-Ewart theory predicts that the number of particle nuclei per unit volume of water generated at the end of Interval I (Np,i) is proportional to the 0.6 power of the surfactant concentration and to the 0.4 power of the initiator concentration [Eq. (3.3)]. Accordingly, the rate of polymerization in Interval II (Rp Np, ) is expected to behave in an identical fashion. These predictions very often form the basis of a test used to verify the validity of the Smith-Ewart theory in emulsion polymerization. The early experimental results mostly obtained from emulsion polymerization of styrene were reported to be in reasonable agreement with the theory under adequate conditions. However, deviations between the experimental data and the Smith-Ewart theory were also observed [13,50,51]. 

Sutterlin [46] studied the effect of the polarity of various monomers (styrene, acrylate ester monomers, and methacrylate ester monomers see Table 3.1) on the particle nucleation mechanisms involved in emulsion polymerization. When the surfactant concentration is above its CMC, the emulsion polymerization of styrene follows the Smith-Ewart theory (Npj 5o ) except those experiments with relatively low levels of surfactant. The exponent x in the relationship Npj So decreases with increasing monomer polarity when the surfactant concentration is above its CMC. This trend is attributed to the increased tendency of agglomeration of particle nuclei with monomer polarity. The emulsion polymerizations of less polar monomers deviate significantly from the Smith-Ewart theory (x 0.6) if the surfactant concentration is reduced to a level just below its CMC. This implies that some mechanisms other than micellar nucleation (homogeneous nucleation or coagulative nucleation) must operate in these emulsion polymerization systems. 

Nomura et al. [74,75] proposed an experimental method to study the competitive particle nucleation mechanisms (micellar nucleation versus homogeneous nucleation) in a given emulsion polymerization system. This approach involves the emulsion copolymerization of relatively hydrophobic styrene with relatively hydrophilic monomers such as methyl methacrylate or methyl acrylate. The composition of copolymer produced during the very early stage of polymerization (far lower than 1% monomer conversion), which reflects the characteristic of copolymer at the locus of particle nucleation, is then determined. Emulsion copolymerization of styrene with methyl methacrylate (or methyl acrylate) was carried out, where sodium dodecyl sulfate was used to stabilize the emulsion polymerization system and where the weight ratio of styrene to methyl methacrylate (or methyl acrylate) was kept constant at 1 1. The experimental results show that the compositions of copolymers obtained from emulsion polymerizations in the presence and absence of monomer-swollen micelles are quite different. This provides supporting evidence of the generally accepted Smith-Ewart theory that micellar nucleation controls the particle nucleation process in the emulsion copolymerization of styrene with... 

Capek and Chudej [87] studied the emulsion polymerization of styrene stabilized by polyethylene oxide sorbitan monolaurate with an average of 20 monomeric units of ethylene oxide per molecule (Tween 20) and initiated by the redox system of ammonium persulfate and sodium thiosulfite. It is interesting to note that the constant reaction rate period is not present in this polymerization system. The maximal rate of polymerization is proportional to the initiator and surfactant concentrations to the -0.45 and 1.5 powers, respectively. The final number of latex particles per unit volume of water is proportional to the initiator and surfactant concentrations to the 0.32 and 1.3 powers, respectively. In addition, the resultant polymer molecular weight is proportional to the initiator and surfactant concentrations to the 0.62 and -0.97 powers, respectively. Some possible reaction mechanisms may explain the deviation of the polymerization system from the classical Smith-Ewart theory. Lin et al. [88] investigated the emulsion polymerization of styrene stabilized by nonylphenol polyethoxylate with an average of 40 monomeric units of ethylene oxide per molecule (NP-40) and initiated by sodium persulfate. The rate of polymerization versus monomer conversion curves exhibit two nonsta-tionary reaction rate intervals and a vague constant rate period in between. 

Sajjadi [47] developed two mechanistic models for the particle nucleation process involved in the semibatch emulsion polymerization of styrene under the monomer-starved condition. In the first model, Smith-Ewart theory was extended to take into account the particle nucleation under the monomer-starved condition. The number of latex particles per unit volume of water is proportional to the surfactant concentration, the rate of initiator decomposition, and the rate of monomer addition, respectively, to the 1.0,2/3, and -2/3 powers. The second model considers the aqueous phase polymerization kinetics and its effect on the efficiency of free radical capture by the monomer-swollen micelles. This model is capable of predicting some features of the particle nucleation process. 

The Smith and Ewart-Stockmayer-O Toole treatments [48-50] (see Chapter 4) that are widely used to calculate the average number of free radicals per particle (n) are based on the assumption that the various components of the monomer-swollen latex particles (e.g., monomer, polymer, free radicals, chain transfer agent, etc.) are uniformly distributed within the particle volume. A latex particle in emulsion homopolymerization of styrene involves uniform distribution of monomer and polymer within the particle volume except perhaps for a very thin layer near the particle surface. In the case of free radicals, this uniform distribution would only hold in a stochastic sense. However, as illustrated in Eq. (8.1), free radicals are not distributed uniformly in the latex particles when water-soluble initiators are used to initiate the free radical polymerization. The assumption of uniform distribution of free radicals in the latex particles would be valid only if the particles are very small or chain transfer reactions are the dominate mechanism for producing free radicals. If such a nonuniform free radical distribution hypothesis is accepted, the very basis of the Smith and Ewart-Stockmayer-O Toole methods might be questioned. Despite this potential problem, the Stockmayer-O Toole solutions for the average number of free radicals per particle have been used for kinetic studies of many emulsion polymerization systems. The theories seem to work reasonably well and have been tested extensively with monomers such as styrene. 

Recently, Durant et al. [55] developed a mechanistic model based on the classic Smith-Ewart theory [48] for the two-phase emulsion polymerization kinetics. This model, which takes into consideration complete kinetic events associated with free radicals, provides a delicate procedure to calculate the polymerization rate for latex particles with two distinct polymer phases. It allows the calculation of the average number of free radicals for each polymer phase and collapses to the correct solutions when applied to single-phase latex particles. Several examples were described for latex particles with core-shell, inverted core-shell, and hemispherical structures, in which the polymer glass transition temperature, monomer concentration and free radical entry rate were varied. This work illustrates the important fact that morphology development and polymerization kinetics are coupled processes and need to be treated simultaneously in order to develop a more realistic model for two-phase emulsion polymerization systems. More efforts are required to advance our knowledge in this research field. 

The Smith-Ewart theory was developed for monomers such as styrene, with very low water solubility. Monomers such as acrylonitrile, with appreciable water solubility (on the order of 10%), may undergo significant homogeneous initiation in the aqueous phase. In some emulsion systems, the particles flocculate (coalesce) during polymerization, not only making a kinetic description difficult, but also sometimes badly fouling reactors. Good reviews of this subject are available [8-11], as well as complete books [7,12-15]. 

---