Radiotagged acrylamide was polymerized in a solution containing untagged polyacrylamide in order to study branch formation. Polymerizations were carried out at 50°C. and at 78°C. with a persulfate‐bisulfite catalyst. At 78°C., polymer with from one to seven branches was formed, while at 50°C. even with larger amounts of catalyst, the polymer was almost exclusively linear. The intrinsic viscosity of branched polyacrylamide, in water at 30°C., was compared with the predictions of the Zimm and Stockmayer equation for the intrinsic viscosity of trifunctionally branched molecules. This equation gave values which agreed well with measured values of the number of branches.

Polymerization of acrylonitrile in dilute aqueous solution initiated by the chloratesulfite redox was studied kinetically and with the aid of the electron microscope. Polymerization rate depended on the [ClO 3−] [H2SO3] product, the monomer content and the sodium lauryl sulfate concentration. The rate was depressed by agitation and by salts; it was accelerated by Fe++ and by Cu++ and was only moderately dependent on temperature. Particles of polymer separated from the start as spheroids less than 200 A. across and grew rather uniformly to 2000 to 3000 A. After a short time no new particles formed unless more initiator was added. Rate of polymerization per particle was reasonably constant. Salts and agitation led to aggregation of particles and polymerization rate was reduced. Sodium lauryl sulfate and Methocel had the opposite effect. The system bears a general resemblance to typical emulsion polymerization, and thus treatments based on homogeneous kinetics appear to be inadequate.

Previous kinetic studies of acrylonitrile polymerization in bulk disclosed unexpected features that were associated with the heterogeneous nature of the system. In the present work, polymerization was carried out in homogeneous solution, principally in dimethyl formamide. Polymerization rate in dimethylformamide was dependent on the square root of the azo initiator concentration and approximately on the three‐halves power of the monomer concentration. Over‐all activation energy was 21 kcal./mole. Rates were slow in this solvent and molecular weights were low. Chain transfer activity of several liquids was estimated by substituting them for part of the dimethylformamide. Contrast between results in the homogeneous and heterogeneous cases is explained as due to removal by solvent of any opportunity for radical burial or catalyst enrichment. Standard kinetic treatments do not explain quite adequately the details of rate and molecular weight dependence on monomer. It appears that sluggish radicals are formed by chain transfer to solvent. A kinetic scheme is proposed to account for these observations.

Kinetics of the chlorate-sulfite reaction in water were investigated at 0 and 20°at pH between 1 and 3.5. The rate of disappearance of sulfite was followed by iodometric titration and the rate of formation of chloride ions was determined ampevometrically with silver nitrate. The data show that the reaction is second order and that the rate is pH dependent. A mechanism is proposed involving the reaction between H2S03 molecules and chlorate ions as the rate-determining step. In units of liters, moles and seconds, the reaction rates were formulated as -d[C108~]/d¿ = 3.3 X ID7 e-nmlRTlC10s-]• [H2S03]. _ The reaction is not much influenced by traces of metals, even though these metals strongly accelerate polymerizations initiated by this system. The exact nature of the free-radical intermediates is thus still in doubt, but it seems likely that sulfoxy radical-ions and hydroxyl radicals are involved. Hydrogen ion concentration determines the H2S08 level and thus the reaction rate.

The slow rates of polymerization and the low molecular weights of polymer obtained by the free radical‐initiated polymerization of vinyl benzoate is noted. The kinetics have been investigated and indicate that addition of the growing chain to the aromatic nucleus to form a stable, nonpropagating radical is the probable reason. The chain length and extent of branching are consistent with the observed rates and also with the methyl affinities of similar compounds.