(233a) Polymer Characterization By Advanced Chromatographic and Rheological Methods

To infer long-chain branching level from rheological data, we are developing theory and computer software based on the “tube model” to predict linear viscoelastic properties for mixtures of unbranched and branched polymers with arbitrary lengths and positions of branch points. To test and improve these predictive methods, model anionic polymers with one or two branches (i.e., star or H polymers) of the same or different lengths are synthesized and thoroughly characterized by both size exclusion chromatography (SEC) and temperature gradient interaction chromatography (TGIC). It is found that even carefully synthesized polymers that show nearly monodisperse molecular weight distributions by SEC nevertheless contain impurities that can be identified and quantified by TGIC.  While these impurities cannot be completely eliminated by fractionation, their effect on the rheology can be accounted for by theory.  The accuracy of the theory is demonstrated not only by prediction of the rheology of the final product branched polymer, but also by its prediction of the rheology of blends of this polymer with linear or star-shaped “impurities” deliberately introduced. To illustrate the power of rheology to detect tiny levels of long-chain branching, we formulate a series of metallocene polymers with nearly the same weight averaged molecular weights, Mw = 115,000 and 116,400, respectively, and nearly the same polydispersity, Mw/Mn = 2.2, but with levels of long-chain branching ranging from 15 down to less than 0.3 long chain branches per million backbone atoms. We find that we can detect and predict differences in the rheology produced by less than branch per million backbone atoms, and show how to account for variations in rheological properties due to both the long-chain and short-chain branches, the latter produced by different choices of co-monomers.