In a study aimed at process optimization of caprolactam polymerization, particular reference has been paid to the competing role of water in caprolactam hydrolysis and endgroup polycondensation. The dependence of the apparent equilibrium constant for polycondensation on water concentration indicated that there is a strong effect of the medium on the activities of the reacting species which can not be neglected in a kinetic study of the polycondensation reaction. The effect of a variation of the medium was taken into account by using a polycondensation rate constant which included a function of the water present at any given time. With the aid of analog computer curve‐fitting techniques, good agreement with second‐order kinetics was found. The validity of a second‐order mechanism was confirmed in a kinetic study of the chain amide linkage hydrolysis. The hydrolysis of caprolactam follows substantially different kinetics, where the generation of carboxyl groups reduces the activation energy of the reaction, which follows predominantly a third‐order mechanism.

The system water–caprolactam–polymer at equilibrium is regarded as a solution consisting of two solvents (water and caprolactam) and one solute (polymer). The activities of water and caprolactam in equilibrium at 270°C in the range of 2–10 wt‐% total water content have been determined by vapor‐pressure measurements. Water shows large negative deviations from Raoult's law, as a consequence of the different size of water and polymer molecules. The partial molar free energies of mixing are compared with the expressions derived from the Flory‐Huggins theory of polymer solutions; the results are not conclusive, but seem to indicate a qualitative agreement with the theory. The increase in vapor pressure during polymerization in sealed systems and the water dependence of the polycondensation equilibrium are discussed and explained in terms of water activity changes.

The rational synthesis of the ‘cyclic’ tetranuclear p‐cresol novolak (VIII) from the corresponding resole (VII) which was synthesized from 3‐bromo‐2:2′‐dihydroxy‐5:5′‐dimethyldiphenylmethane (I) has been accomplished in accordance with the scheme given below. A molecular model of the resole (VII) indicated that intramolecular hydrogen bonding of the hydroxyl groups would assist the formation of a configuration which could undergo facile cyclization to give the required novolak (VIII). The infrared absorption spectrum of the ‘cyclic’ tetranuclear novolak showed only the band corresponding to 1:2:4:6‐tetra‐substitution and absence of that due to 1:2:4‐tri‐substitution which was present in the resole (VII). Confirmation of structure was obtained by conversion into a tetra‐acetate which gave a molecular weight determination of the correct order and whose infra‐red absorption spectrum also showed only 1:2:4:6‐tetra‐substitution.

No abstract

No abstract