Macroemulsion polymerization

Normal emulsion polymerization is sometimes referred to as macroemulsion polymerization because of the large size of monomer droplets (hundreds of microns) compared to those of a microemulsion (tens of nanometer). 

At first, the mechanism of macroemulsion polymerization of vinyl monomers [67] is shortly considered. Emulsion polymerization usually takes place in three... 

To synthesize water-soluble or swellable copolymers, inverse heterophase polymerization processes are of special interest. The inverse macroemulsion polymerization is only reported for the copolymerization of two hydrophilic monomers. Hernandez-Barajas and Hunkeler [62] investigated the copolymerization of AAm with quaternary ammonium cationic monomers in the presence of block copoly-meric surfactants by batch and semi-batch inverse emulsion copolymerization. Glukhikh et al. [63] reported the copolymerization of AAm and methacrylic acid using an inverse emulsion system. Amphiphilic copolymers from inverse systems are also successfully obtained in microemulsion polymerization. For example, Vaskova et al. [64-66] copolymerized the hydrophilic AAm with more hydrophobic methyl methacrylate (MMA) or styrene in a water-in-oil microemulsion initiated by radical initiators with different solubilities in water. However, not only copolymer, but also homopolymer was formed. The total conversion of MMA was rather limited (<10%) and the composition of the copolymer was almost independent of the comonomer ratio. This was probably due to a constant molar ratio of the monomers in the water phase or at the interface as the possible locus of polymerization. Also, in the case of styrene copolymerizing with AAm, the molar fraction of AAm in homopolymer compared to copolymer is about 45-55 wt% [67], which is still too high for a meaningful technical application. 

There are four main types of liquid-phase heterogeneous free-radical polymerization microemulsion polymerization, emulsion polymerization, miniemulsion polymerization and dispersion polymerization, all of which can produce nano- to micron-sized polymeric particles. Emulsion polymerization is sometimes called macroemulsion polymerization. In recent years, these heterophase polymerization reactions have become more and more important... 

In order to adequately discuss miniemulsion polymerization, it will be necessary to review the mechanism of macroemulsion polymerization. 

Macroemulsion polymerization is a complex process. The literature contains extensive reviews of emulsion polymerization theory [6-11]. Only a brief review of the current state of the literature is given here. The theory of emul-... 

All quantitative theories based on micellar nucleation can be developed from balances of the number concentrations of particles, and of the concentrations of aqueous radicals. Smith and Ewart solved these balances for two limiting cases (i) all free radials generated in the aqueous phase assumed to be absorbed by surfactant micelles, and (ii) micelles and existing particles competing for aqueous phase radicals. In both cases, the number of particles at the end of Interval I in a batch macroemulsion polymerization is predicted to be proportional to the aqueous phase radical flux to the power of 0.4, and to the initial surfactant concentration to the power of 0.6. The Smith Ewart model predicts particle numbers accurately for styrene and other water-insoluble monomers. Deviations from the SE theory occur when there are substantial amounts of radical desorption, aqueous phase termination, or when the calculation of absorbance efficiency is in error. 

The distinguishing feature of droplet nucleation as opposed to micellar or homogeneous nucleation is the nature of the particle at birth . Droplets, which are nucleated into particles, begin as nearly 100% monomer. Micellar or homogeneous nucleated particles start out with much lower monomer concentrations and eventually swell to around 60% (for MMA) in the presence of monomer droplets. This fundamental difference may lead to large differences between miniemulsion and macroemulsion polymerizations in radical desorption and/or intraparticle termination during Intervals I and II. 

SE Interval II begins at the cessation of nucleation, or in hght of the nucleation theory just reviewed, when the particle number becomes relatively constant. Most theories developed for this interval assume a constant particle number and use the quasi-steady-state approximation (QSSA) for average number of radicals per particle. The kinetics and mechanisms of Interval II have been some of the most studied aspects of macroemulsion polymerization. SE Interval II ends when the monomer droplets disappear and the monomer concentration in the particles begins to decrease. 

After having described the mechanism of miniemulsion polymerization and how it differs from macroemulsion polymerization in the previous section, this section will focus on the various mechanisms and properties of miniemulsion polymerization. 

The vast majority of miniemulsion polymerizations reported in the literature have been stabilized with anionic surfactants, probably because of the widespread application of anionic surfactants in macroemulsion polymerization, and due to their compatibility with neutral or anionic (acid) monomers and anionic initiators. However, Landfester and coworkers [70, 71] have used the cationic surfactants cetyltrimethyl ammonium bromide (CTAB) and cetyltri-methyl ammonium tartrate for the production of styrene miniemulsions. They report that these surfactants produce similar particle sizes to anionic surfactants used at the same levels. Bradley and Grieser [72] report the use of dodecyltrimethyl ammonium chloride for the miniemulsion polymerization of MMA and BA. 

Following the common practice in macroemulsion polymerization, most miniemulsion polymerizations have been run using water-soluble initiators. However, a number of researchers have looked at the possibility of using an oil-soluble initiator instead. As discussed previously, Schork and Reimers [107] and Asua et al. [108] have used LP as both the initiator and the costabilizer. In addition, Asua et al. used other oil-soluble initiators in conjunction with HD (as the costabilizer) to carry out miniemulsion polymerization of styrene. 

One of the problems with macroemulsion polymerization is the variability of the particle number with initiation rate, monomer quality, inhibition levels, and so on. This is a serious industrial problem, as shown by the fact that a great many industrial macroemulsion polymerizations are carried out as seeded polymerizations in which a known concentration of seed particles are added to the emulsion, and the polymerization is rim under conditions that suppress nucleation of additional particles. The variance in particle number comes about because there is a competition for surfactant between the growth of existing particles (that need additional surfactant to stabihze their growing surface area), and the nucleation of new particles. 

Macroemulsion polymerization relies on the transport of monomer from the monomer droplets to the polymer particles. This transport is driven by the equilibrium swelling of the polymer particles. This presumes rapid (relative to the rate of polymerization) transport of monomer. For most monomers, this is a good assumption. However, for monomers that are very water insoluble (VEOVA [vinyl versatatex] or DOM [dioctyl maleate]), this may not be true. In making this determination, the following assumptions can be made ... 

In the case of nanoencapsulations of solids, or the incorporation of high molecular weight, highly water-insoluble additives (such as polymers, oligomers, alkyds) into polymer particles, macroemulsion polymerization will not work, since the high molecular weight material will remain in the monomer droplet as the monomer is transported out. At the end of the reaction, the additive will remain in the depleted monomer droplets, rather than in the polymer particles. Clearly, these products can only be made via miniemulsion polymerization. 

Macro- and miniemulsion polymerization in a PFR/CSTR train was modeled by Samer and Schork [64]. Since particle nucleation and growth are coupled for macroemulsion polymerization in a CSTR, the number of particles formed in a CSTR only is a fraction of the number of particles generated in a batch reactor. For this reason, their results showed that a PFR upstream of a CSTR has a dramatic effect on the number of particles and the rate of polymerization in the CSTR. In fact, the CSTR was found to produce only 20% of the number of particles generated in a PFR/CSTR train with the same total residence time as the CSTR alone. By contrast, since miniemulsions are dominated by droplet nucleation, the use of a PFR prereactor had a negligible effect on the rate of polymerization in the CSTR. The number of particles generated in the CSTR was 100% of the number of particles generated in a PFR/CSTR train with the same total residence time as the CSTR alone. 

Observation (i) above can be understood in terms of droplet nucleation and the lack of competition between nucleation and growth. A mechanistic understanding of observation (ii) above was provided by Samer and Schork [64]. Nomura and Harada [136] quantified the differences in particle nucleation behavior for macroemulsion polymerization between a CSTR and a batch reactor. They started with the rate of particle formation in a CSTR and included an expression for the rate of particle nucleation based on Smith Ewart theory. In macroemulsion, a surfactant balance is used to constrain the micelle concentration, given the surfactant concentration and surface area of existing particles. Therefore, they found a relation between the number of polymer particles and the residence time (reactor volume divided by volumetric flowrate). They compared this relation to a similar equation for particle formation in a batch reactor, and concluded that a CSTR will produce no more than 57% of the number of particles produced in a batch reactor. This is due mainly to the fact that particle formation and growth occur simultaneously in a CSTR, as suggested earlier. 

Results from the polymer-costabilized miniemulsion polymerizations are shown in Table 2. Droplet sizes were found to vary between 115.1 and 121.0 nm. These are in accord with measurements made by Fontenot [140] for MMA miniemulsions stabilized with hexadecane. The sizes of the particles in the final products were close to the sizes of the droplets, ranging from 102.6 to 108.1 nm, with polydispersities ranging from 1.011 to 1.027. The ratio of the number of particles to the number of droplets (N /N ) was found to be between 0.95 and 1.08. Therefore, the majority of the droplets were nucleated to form polymer particles. Droplet nucleation led to polymerization rates comparable to those for the corresponding macroemulsions. For equal concentrations of initiator, 0.01 Maq, the rates are 0.199 and 0.233 gmol/min L q for the mini- and the macroemulsion polymerizations, respectively. 

The effects of the (water-soluble) initiator concentration on the polymerization of polymer-stabilized miniemulsion are shown in Table 2. An increase in the initiator concentration does not change the number of particles, but does increase the rate of polymerization. This is due to an increase in the number of radicals per particle. However, the number of radicals per particle ranged from just 0.5 to 0.8, indicating that the kinetics (after nucleation) are still essentially Smith Ewart Case II. The number of particles was found to be proportional to the initiator concentration raised to the power of 0.002 0.001. Macroemulsion polymerizations, in contrast, show a dependence of 0.2 and 0.4 for methyl methacrylate and styrene, respectively [141]. The fact that the exponent approaches zero indicates that all or nearly all of the droplets are being nucleated. 

A water-phase retarder (sodium nitrite) was added to both the mini- and macroemulsion polymerizations. The rate of polymerization was reduced with increasing level of retarder, as would be expected. However, the number of particles increased with increasing retarder concentration. This result would only be expected with an oil-soluble retarder. The reason for this anomaly is... 

Fig. 12 The effect of a water-phase retarder on the number of particles in macroemulsion polymerization (from [95])...

Macroemulsion polymerizations carried out in the presence of an oil-phase inhibitor (DPPH) resulted in an increase in the number of particles. Presumably initiator radicals that enter droplets are terminated by the inhibitor, resulting in dead particles. These particles do not grow, and hence do not consume surfactant to stabilize their increasing surface area, until they absorb another radical. The surfactant not adsorbed by dead particles is available to... 

One of the most unique properties of miniemulsion polymerization is the lack of monomer transport. Recall from Fig. 1 that with macroemulsion polymerization, the monomer must diffuse from the monomer droplets, across the aqueous phase, and into the growing polymer particles. In contrast, in an ideal miniemulsion (nucleation of 100% of the droplets), there is no monomer transport, since the monomer is polymerized within the nucleated droplets. This lack of monomer transport leads to some of the most interesting properties of miniemulsions. For most monomers, macroemulsion polymerization is considered to be reaction, rather than diffusion limited. However, for extremely water insoluble monomers, this might not be the case. In this instance, polymerization in a miniemulsion might be substantially faster than polymerization in an equivalent macroemulsion. For copolymerization in a macroemulsion, where one of the comonomers is highly water insoluble, the comonomer composition at the locus of polymerization might be quite different from the overall comonomer composition, resulting in copolymer compositions other than those predicted by the reactivity ratios. 

The Mayo Lewis equation, using reactivity ratios computed from Eq. 18, will give very different results from the homogenous Mayo Lewis equation for mini-or macroemulsion polymerization when one of the comonomers is substantially water-soluble. Guillot [151] observed this behavior experimentally for the common comonomer pairs of styrene/acrylonitrile and butyl acrylate/vinyl acetate. Both acrylonitrile and vinyl acetate are relatively water-soluble (8.5 and 2.5%wt, respectively) whereas styrene and butyl acrylate are relatively water-insoluble (0.1 and 0.14%wt, respectively). However, in spite of the fact that styrene and butyl acrylate are relatively water-insoluble, monomer transport across the aqueous phase is normally fast enough to maintain equilibrium swelling in the growing polymer particle, and so we can use the monomer partition coefficient. 

Typically, copolymer composition can be manually adjusted by slowly feeding the more reactive monomer in throughout the reaction but this may not be helpful when trying to overcome monomer transport limitations. Therefore, Reimers and Schork [ 102] performed identical copolymerization experiments in miniemulsions, where monomer transport is less significant, in order to determine what effect this would have on the evolution of the copolymer composition. Data on the MMA/VS (and other) copolymerizations indicate that the Schuller equation (and not the Samer adaptation) fits the copolymer composition data. This points to the effect of extremely low monomer water solubility on copolymer composition in macroemulsion polymerization, and the relative insensitivity of miniemulsion polymerization to this effect. 

---