On the Immobilized Polymer Fraction in Attractive Nanocomposites: Tg Gradient versus Interfacial Layer

We study the slowing down of polymer dynamics in canonical nanocomposites made of strongly interacting mixtures of poly-2-vinyl pyridine and silica nanoparticles (radius R = 7 ± 2 nm) by means of low-field NMR relaxometry. We demonstrate that this technique enables the accurate quantification of the fraction of immobilized (irreversibly adsorbed) polymer on to the filler surface, in a manner that complements data from broadband dielectric spectroscopy (BDS) and differential scanning calorimetry (DSC). We rationalize the slowing down using two previously developed approaches to model the immobilized layer. The first one is based on a glass-transition temperature (Tg) gradient, while the second one assumes a single interfacial component with a distribution of relaxation times. While both models convincingly fit the NMR data providing us with robust and complementary conclusions, the former approach is found to fit more accurately previously published DSC experiments. Quantitatively, NMR results indicate that the segmental relaxation of the immobilized layer is ca. 1 decade slower than in the bulk polymer, while earlier BDS analyses were equivocal on this point, indicating either 1 or 2 decades. These discrepancies highlight the difficulties and potential model dependencies in quantifying the dynamics of the minority polymer fraction in the immobilized layer, necessitating a systematic multi-technique approach to properly characterize dynamical heterogeneities in polymer-based materials.