330 THE BELL SYSTEM TECHNICAL JOURNAL, MARCH 1952 



frequencies approached the glass-into-rubber relaxation times. Clearly, 

 again, mdividual interaction of chain-like chemical units and not any 

 micellar or other special aggregation of them, predominates polymer 

 mechanics. 



It still remains, however, to sepai'ale interactions of the basic; units 

 within and between chains. Most likely, the model plastic vs rubber 

 h quids just discussed differ in the high frequency region substantially 

 only because of m/er-chain forces between phenyl vs methyl groups. 

 However, especially in the high-frequency region, questions of intra- 

 chain structure, such as the steric hindrance of adjacent pairs of methyl 

 groups in polyisobutylene, restricted rotations about bonds, etc., come 

 in. Obviously where configurational or quasi-configurational displace- 

 ments are important, as in all cases of elongation >20 per cent (this is 

 certainly an upper limit), flexibility of single chains needs to be under- 

 stood. This is built deeply into chemical structure; plasticizers pre- 

 sumably may change over-all configuration as well as modify interaction, 

 but they are impotent to vary flexibility. Accordmgly, problems of rub- 

 ber, usable in the Arctic, and of wire and cable insulation bendable at 

 low temperatures always come back to whether the polymer chain bonds 

 have free rotation. Some examples of the combinations of effects within 

 and between chams can indeed be shown m several other polymer liquids 

 which are rubber models. 



This influence of small changes m chemical structure is compactly 

 illustrated by comparing a few other hydrocarbon polymer liquids with 

 polyisobutylene. Also, rather dilute dipolar groups have been introduced 

 in the linear polyester liquid polypropylene sebacate, whose structure is 

 otherwise like that of hydrocarbons.'*^'' In Table VI, liquids of the given 

 structure with some (unknown) distribution of molecular weights, were 

 studied with shear waves at 77 and 142 kc at a temperature where each 

 had the same steady flow viscosity. The figure chosen was 700 poises, 

 and the temperature range required to adjust to it in the series was 

 10.9° to 85°C, meaning that the liquids had comparable consistencies at 

 ordinary temperatures. 



Despite these similarities under steady stress, the retardation times, 

 t', vary three-fold, with the highly substituted hydrocarbon chains, 

 polyisobutylene and polypropylene, the highest. Despite the intermo- 

 lecular action of the dipoles in polypropylene sebacate, the low polymer 

 has a short retardation time, although its "brittle point" with decreas- 

 ing temperature is far above that of polybutadiene or even polyiso- 

 butylene. Presumably the flexibility around C — — C bonds rather 

 compensates for mcreased dipole interaction. Where both low polarity 



