Study of the correctness of the solubility parameters obtained from indirect methods by molecular dynamics simulation

Conformational energies of models of polyethylene (PE), isotactic-polypropylene (i-PP), atactic-polypropylene (a-PP) and polyisobutylene (PIB) with different chain lengths in their unperturbed liquid state (El), unperturbed isolated state (Eiu) and perturbed isolated state (Eip) at 190 °C were computed using molecular dynamics simulation. The computed El, Eiu and Eip were then used to calculate the solubility parameters of the polymers (δ). Our results indicate that there exists a cutoff chain length (∼20 backbone carbons) above which the differences between the solubility parameters (Δδ) computed based on (Eiu−El) and (Eip−El), respectively, become significant. In addition, it was found that the higher the degree of shrinkage of the polymer in vacuum is, the larger the Δδ is. Since PE and PIB exhibited considerable shrinkage in vacuum, their Δδ are much higher than those of the PPs. It seems that, at 190 °C, vacuum acts more or less like a Θ solvent for both types of polypropylenes but a bad solvent for PE and PIB. Our results also suggest that the characteristic ratio of the polymer has little effect on Δδ and that it was the long range attractive interactions between the monomers distant along the backbone contour of the same molecule that led to the formation of fairly compact globule in the cases of PE and PIB. Since δ of the high molecular weight models computed from (Eiu−El) agree with experiment better than those from (Eip−El), it implies that δ obtained from indirect measurements correspond to a hypothetical vaporization process where the conformations of polymer coils do not undergo significant changes. In other words, δ of a polymer that is determined by the currently available indirect methods may deviates significantly from their ‘true’ value if the polymer exhibits a high degree of swelling or shrinkage in vacuum at the temperature of the experiment.