Abstract Quantitative experimental evidence on (i) the stress in stretched rubbers, (ii) temperature coefficients of the stress, (iii) cyclization equilibria, and (iv) thermodynamic activities of solvents in the Henry's law range reveal no discemible effects of dilution that could be attributed to dispersal of chain order postulated to exist in the undiluted polymer. The results in support of this statement have been obtained by diverse methods applied in many different laboratories. They appear to be confirmed by neutron scattering studies now in progress. The values of 〈r2〉0 deduced for PDMS from (iii) and of d ln〈r2〉0/dT from (ii) for several bulk polymers are in excellent agreement with measurements in dilute solutions. The body of evidence as a whole is compelling to the effect that the configurations of macromolecular chains in amorphous bulk polymers are random, and that they differ imperceptibly from the configurations occurring at infinite dilution (apart from the effect of excluded volume in dilute media). Scarcely less compelling is the evidence for local intermolecular order furnished by optical anisotropies observed by depolarized light scattering and by strain birefringence as discussed above. These observations do not directly contradict those cited in the paragraph above. We have pointed out earlier that intermolecular correlations need to be carefully distinguished from intramolecular correlations; it is the latter that are closely bound to the molecular configuration. Several aspects of the evidence for correlations obtained from optical anisotropy should be carefully noted. First, the range of intramolecular correlation of the group polarizability tensors is short. This is apparent from the fact that 〈γ2〉#x002F;n for the isolated chain increases by a factor of only ca. 1.3 for PM chains CH3—(CH2)n—H from n=1 to ∞. It is almost as if the individual group tensors αCH2 were mutually independent, in which case 〈γ2〉/n would be constant with n. In contrast, the characteristic ratio 〈r2〉0/n increases about eightfold (at ordinary temperature) from n=1 to ∞. The apparent optical anisotropy is therefore more vulnerable to intramolecular effects as judged by relative changes in this quantity. Secondly, intermolecular correlative effects manifested in the optical anisotropy do not necessarily denote parallelization of chains. In the case of n-alkane chains, for example, the same enhancement of the apparent 〈γ2〉 could be achieved by alignment of the vectors that are normal to the planes defined by the carbon atoms in trans sequences occurring in neighboring chains. Finally, the ‘bundle’, ‘micelle’, and mesomorphic models so often advocated for bulk amorphous polymers would require the optical anisotropy to be at least an order of magnitude greater than observed. The effects observed are small. They cannot be reconciled with the degree of order postulated. They are comparable in magnitude to the correlative effects identified in liquids consisting of small molecules. The concatenation of units comprising the polymer chain evidently does not of itself greatly enhance intermolecular correlations in the amorphous state.
Read full abstract