Size-exclusion chromatography of poly(ethylene 2,6-naphthalate).

A solvent mixture of 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) and dichloroacetic acid (DCAA) is used to dissolve difficultly soluble poly(ethylene 2,6-naphthalate) (PEN). Solutions can be diluted and analyzed in a common size-exclusion chromatography (SEC) eluent, HFIP. The HFIP/DCAA mixture is better at dissolving PEN than either solvent individually and it is easier and safer to work with than phenolic and strongly acidic eluents. Dissolution temperatures between 50 and 60 °C are sufficiently low to minimize hydrolytic degradation of the polyester. PEN does not dissolve in the solvent mixture if the water concentration is greater than 0.76 wt%, and preferably the water content should be less than 0.13 wt% to eliminate minor prepeak artifacts. The procedure is suitable for PEN that is less than 48% crystalline, including prepolymers, oriented films and some solid-state polymerized materials. Highly crystalline polymers can be melt-quenched into a more amorphous state to render them soluble. The dilute solution conformational properties of PEN are compared to PET in HFIP, and molar mass-intrinsic viscosity scaling constants and unperturbed dimensions are calculated from SEC data.

Size-exclusion chromatography of perfluorosulfonated ionomers.

A size-exclusion chromatography (SEC) method in N,N-dimethylformamide containing 0.1 M LiNO(3) is shown to be suitable for the determination of molar mass distributions of three classes of perfluorosulfonated ionomers, including Nafion(®). Autoclaving sample preparation is optimized to prepare molecular solutions free of aggregates, and a solvent exchange method concentrates the autoclaved samples to enable the use of molar-mass-sensitive detection. Calibration curves obtained from light scattering and viscometry detection suggest minor variation in the specific refractive index increment across the molecular size distributions, which introduces inaccuracies in the calculation of local absolute molar masses and intrinsic viscosities. Conformation plots that combine apparent molar masses from light scattering detection with apparent intrinsic viscosities from viscometry detection partially compensate for the variations in refractive index increment. The conformation plots are consistent with compact polymer conformations, and they provide Mark-Houwink-Sakurada constants that can be used to calculate molar mass distributions without molar-mass-sensitive detection. Unperturbed dimensions and characteristic ratios calculated from viscosity-molar mass relationships indicate unusually free rotation of the perfluoroalkane backbones and may suggest limitations to applying two-parameter excluded volume theories for these ionomers.

Viscoelasticity of randomly branched polymers in the vulcanization class.

We report viscosity, recoverable compliance, and molar mass distribution for a series of randomly branched polyester samples with long linear chain sections between branch points. Molecular structure characterization determines tau=2.47+/-0.05 for the exponent controlling the molar mass distribution, so this system belongs to the vulcanization (mean-field) universality class. Consequently, branched polymers of similar size strongly overlap and form interchain entanglements. The viscosity diverges at the gel point with an exponent s=6.1+/-0.3, that is significantly larger than the value of 1.33 predicted by the branched polymer Rouse model (bead-spring model without entanglements). The recoverable compliance diverges at the percolation threshold with an exponent t=3.2+/-0.2. This effect is consistent with the idea that each branched polymer of size equal to the correlation length stores k(B)T of elastic energy. Near the gel point, the complex shear modulus is a power law in frequency with an exponent u=0.33+/-0.05. The measured rheological exponents confirm that the dynamic scaling law u=t/(s+t) holds for the vulcanization class. Since s is larger and u is smaller than the Rouse values observed in systems that belong to the critical percolation universality class, we conclude that entanglements profoundly increase the longest relaxation time. Examination of the literature data reveals clear trends for the exponents s and u as functions of the chain length between branch points. These dependencies, qualitatively explained by hierarchical relaxation models, imply that the dynamic scaling observed in systems that belong to the vulcanization class is nonuniversal.