No abstract

No abstract

synopsisIsoprene was polymerized in batch reactors by use of bottle polymerization techniques a t 30°, 40°, and 5OoC a t concentrations from 1 to 5 molar. Butyllithium concentration was varied from 0.005 to 0.03 molar. Isoprene and n-butyllithium conversions and molecular weight distributions were determined for different reaction times. Rate equations for the initiation and propagation reactions are presented. The importance of the association reactions in obtaining a narrow molecular weight distribution is illustrated. SCOPE AND SIGNIFICANCEI n the design, optimization, or control of polymerization reactors, it is desirable to have a mathematical model which adequately represents the process. The objective of this work was to study the total polymerization of isoprene in hexane with n-butyllithium and develop such a model by use of the experimental data. A secondary objective was t o gain an insight as to the effect of the association reactions on the molecular weight distribution. This system was chosen because a cis-1,4 polyisoprene similar to heva natural rubber is produced. EXPERIMENTALSeven runs consisting of 40 samples were conducted in batch reactors at 30°, 40°, and 50°C. At least two runs were conducted a t each temperature.The isoprene was double distilled and refluxed over sodium ribbon before use. Baker instraanalyzed hexane was stored for several days over sodium ribbon before use. n-Butyllithium was purchased from Foote Mineral Company in a 1.6 molar solution of hexane. This solution was diluted in hexane to approximately 0.3 molar and anaIyzed by use of disulfide cleavage and subsequent titration with silver nitrate. This procedure has been presented by Koltoff and Harris' and Uranek et a1.2 The polymerization bottles were dried and rinsed with butyllithium solution and rerinsed with purified hexane before use. Numerous precautions were taken to insure that no air or water con- 1805

SynopsisSytrene was polymerized in a 0.8-liter vessel which was operated as an isothermal and a nonisothermal batch reactor. Styrene and n-butyllithium conversions were determined for different reaction times. Rate equations were developed by use of the isothermal data and then used to estimate the conversions for the nonisothermal experiments. The importance of using nonisothermal data in the development of rate equations is illustrated.

Using recently developed coaxial line methods values of permittivity and dielectric loss have been determined over the frequency range 0.5 to 7 GHz for a series of reactionbonded silicon nitride specimens in which the degree of nitridation has been varied. For fully nitrided material (having a weight gain of 62% and a volume porosity of 19%) the measured permittivity was 4.60 and was almost independent of frequency; fitting both the permittivity and loss data to the Universal Law of dielectric response confirmed that the limiting condition of lattice loss applied with n = 0.98 + 0.02. Reduction of the degree of nitridation caused progressive increases in permittivity and loss, both of which closely approached the quoted values for pure silicon at weight gains below about 40%.