Site-specific labeling and the distribution of free volume in glassy polystyrene

Free volume environments at different sites along the polystyrene molecule have been evaluated by site-specific labeling with azobenzene chromophores. The sites investigated are the regions around the chain ends, the chain sides, and the chain centers along the backbone. Also investigated was the free volume of the least constrained regions between the chains by the use of free probes not attached to the polymer molecule. Trans - > cis photoisomerization kinetics of the labels and the free probe were studied at 70, 80, and 90 deg C as a function of annealing time following cooling from the melt. The photoisomerization kinetics suggest that each of the labels and the free probe can find itself in one of two different environments. In one environment, accounting for a fraction alpha of each of the chromophores, photoisomerization was rapid; in the other environment, photoisomerization was very slow. The fractions of rapidly photoisomerizing chromophores were found to decrease with aging time, with the change being similar, though not identical, for each site. The magnitudes of the rapid fractions always remained in the following descending order, however: free probe, chain ends, side groups, and chain center. The Robertson--Simha--Curro (RSC) theory was used to compute the populations of regions having a greater than critical amount of free volume, and these were then compared with the experimentally measured fraction ( alpha ) of the fast photoisomerization. Except for recovery at 30 deg C below the glass transition temperature, at which temperature and below the computational approach appears to be inadequate, the computed fit to the data was very satisfactory. This suggests that photoisomerizable labels and, to some extent, the free probe are able to ascertain the magnitude of the local free volume at various locations in bulk polymer. It suggests also that the basis of the RSC theory, involving a distribution of free volumes arising from thermal fluctuations, is correct. 35 ref.--AA