Polymerization of styrene in alcoholic media in the presence of poly(jV-vinylpyrrolidone) (PVP) produces monodisperse latex particles from 1 to 18 µ in size. The results are rationalized in the context of a mechanism involving grafting of polystyrene chains to PVP steric stabilizer. Increasing the initial styrene concentration from 5 to 40% by volume increases the particle size obtained. Comparison of this effect to that of added toluene suggests that the primary effect of monomer concentration is a solvent effect: the more soluble the polystyrene is in the reaction medium, the larger the particles. The influence of PVP concentration and molecular weight was found to obey the approximate relationship dso = 90.4[PVP]~°'31MW~°•18. The power law dependence of diameter on initiator concentration was found to be 0.39, in good agreement with prior work. The addition of a chain-transfer agent produced polydisperse particles with molecular weight dispersities (Mw/Mn) of ca. 100, modifying the power law dependence of size on initiator concentration to a slope of 0.19. A maximum in particle size was found in a series of solvents varying from 80% ethanol/water to ethanol and from methanol to decanol, with 4-µ particles obtained in butanol and pentanol, 1.2-µ particles in 80% ethanol/water, and 1.6-µ particles in decanol. The molecular weights correlated inversely with particle size and could be semiquantitatively rationalized on the basis of the locus of polymerization shifting from particle capture of growing oligomers from the continuous phase in the case of smaller particles to particle capture of dead polymer in the case of large particles.

An integrated model for dispersion polymerization is developed to predict particle size from first principles. The key components of this model are a multibin kinetic model for unstabilized particle coalescence, the grafting mechanism of stabilization, and the radius of gyration of the grafted stabilizer chains. A critical point is defined where similar-sized particles stop coalescing with one another because the graft available equals the minimum graft required to stabilize the particles. Examples indicate that the particle count is determined within the first 1% of conversion, in agreement with experiment, and that the critical size particles are larger than 0.1 pm-significantly bigger than nuclei. In cases where full adsorption of graft is likely, absolute prediction of particle size is remarkably consistent with experiment. More generally, however, partial adsorption of graft is likely to affect the predicted scaling of diameter with stabilizer, initiator and monomer concentrations, and stabilizer molecular weight. Above the critical size, the size distribution narrows because the growing particles continue to capture dead polymer formed in solution. Observed loss of monodispersity by stabilization of secondary particles or coalescence of large particles is readily explained with this model. A critical factor influencing the final size distribution is the locus of polymerization and how it changes during the reaction.During dispersion polymerizations latex particles are formed from an initially homogeneous reaction mixture by polymerization in the presence of a suitable steric stabilizer p01ymer.l-l~ In hydrocarbon solvents, poly(dimethylsiloxane), poly(isobutylene), poly(l2-hydroxystearic acid), and poly(2-ethylhexyl methacrylate) have been employed as stabilizers for polymerization of methyl metha~rylate.~-" Styrene monomer has been polymerized in alcohols with steric stabilizers such as (hydroxy-propy1)cellulose (HPC), poly(acry1ic acid) (PAA), or poly(N-vinylpyrrolidone) (PVP).1s4s6J2-17 In favorable circumstances, these particles can have a very narrow, or even monodisperse, size distribution.Our previous work on the dispersion polymerization of styrene in alcohols had focused primarily on HPC and, more recently, on PVP as steric stabilizer.l* These efforts continued work initiated by Ober, Lok, and Hair (HPC and PAA),13-15 Almog e t al. (PVP and others),17 and Vanderhoff and El-Aasser et al. (PVP).16 Although a costabilizer (anionic or nonionic surfactant) was employed with P V P in t h e investigations by Vanderhoff and El-Aasser, recent evidence suggests that this additive has no influence on the outcome of the reaction in most of the parameter space inve~tigated.~ The gross features of the HPC-and PVP-stabilized reactions are substantially similar and are comparable to reactions stabilized by PAA, poly(viny1 butyral), and poly(N-vinylpyridine).18The preceding work by us and others has shown the mechanism of dispersion polymerization to be complex and poorly understood. It is known that the particle c...

No abstract

SynopsisChanging solvent has a dramatic effect on the outcome of HPC stabilized dispersion polymerization of styrene in polar solvents. In methoxyethanol/ethanol mixtures, particle size varies from essentially infinity for unstabilized reactions (below 30% ethanol) to about 3 microns in ethanol. In a series of n-alcohols, there was a maximum in particle size at intermediate chain length particle size increased from 2 microns in methanol to 8.3 microns in pentanol, then decreased again to 1 micron in octadecanol. These results were rationalized in terms of the three-component Hansen solubility parameters. The largest particles were obtained in solvents with Hansen polarity and hydrogen bonding terms closest to HPC. The generality of the three-component solubility parameter approach was examined by reactions performed in eight mixed solvents with the same values of all three Hansen terms. Seven of the eight solvents gave particles of similar size and molecular weight. The two critical effects of solvent on both particle size and molecular weight appear to be: ( a ) the solubility properties of the grafted HPC-PS formed, and ( b ) the partitioning of monomer and initiator between solution and particle phases subsequent to nucleation.

No abstract