Polymer properties, such as their mechanical strength, barrier properties, and dielectric response, can be dramatically improved by the addition of nanoparticles. This improvement is thought to be because the surface area per unit mass of particles increases with decreasing particle size, R, as 1/R. This favorable effect has to be reconciled with the expectation that at small enough R the nanoparticles must behave akin to a solvent and cause a deterioration of properties. How does this transition in behavior from large solutes to the solvent limit occur? We conjecture that for small enough particles the layer of polymer affected by the particles (“bound” polymer layer) must be much smaller than that for large particles: the favorable effect of increasing particle surface area can thus be overcome and lead to the small solvent limit with unfavorable mechanical properties, for example. To substantiate this picture requires that we measure and compare the “bound polymer layer” formed on nanoparticles with those near large particles with equivalent chemistry. We have implemented a novel strategy to obtain uniform nanoparticle dispersion in polymers, a problem for many previous works. Then, by combining theory and a suite of experimental techniques, including differential scanning calorimetry and positron annihilation lifetime spectroscopy, we show that the immobilized poly(2-vinylpyridine) layer near 15 nm diameter silica particles (∼1 nm) is considerably thinner than that at flat silica surfaces (∼4 to 5 nm), which is the limit of an infinitely large particle. We have also determined that the changes in the polymer’s glass-transition temperature due to the presence of this strongly interacting surface are very small in both well-dispersed nanocomposites and thin films (<100 nm). Similarly, the polymer’s fragility, as determined by dielectric spectroscopy, is also found to be little affected in the nanocomposites relative to the pure polymer. While a systematic study of the dependence of the bound polymer layer thickness on particle size remains an outstanding challenge, this first study provides conclusive evidence for the hypothesis that the bound polymer layer can be significantly smaller around nanoparticles than at chemically similar flat surfaces.

No abstract

The molecular mobility of chemically cross-linked and thermoplastic elastomers is of utmost importance in establishing physical properties for uses ranging from automotive tires and shoe soles to more sophisticated aerospace and biomedical applications. In many of these applications, network chain dynamics under external stresses/strains are critical for determining ultimate performance. It is well established that T g (or T R , the dynamic T g ) and the breadth of the segmental (R) relaxation time distribution increase with increasing cross-link density for chemically cross-linked polymers. For elastomers, there are considered to be two contributions important in determining the change in T R upon uniaxial extension: a reduction in conformational entropy, encouraging an increase in T R , and a volume increase on uniaxial deformation, leading to a reduction in T R (nonideal networks are compressible to some degree). 1,2Considering the importance of mechanical deformation in elastomer applications, there have been relatively few previous investigations of mechanical strain on chain dynamics, e.g., 3,4 even fewer using broadband dielectric spectroscopy, and none on thermoplastic elastomers. [5][6][7][8] This is despite the fact that modern broadband dielectric spectroscopy is a powerful tool for the investigation of material dynamics over very wide frequency and temperature ranges.In this Communication, we report the findings of our initial dielectric spectroscopy investigation of the role of uniaxial extension on the relaxation behavior of cross-linked polyisoprene and segmented polyurethane elastomers. A Novocontrol GmbH Concept 40 broadband dielectric spectrometer was used to measure dielectric permittivity and loss over a broad range of temperatures and frequencies.Synthetic polyisoprene [PI, NIPOL-IR2200 (Zeon Chemicals), 98.5% cis-1,4 content] was cross-linked with dicumyl peroxide (Sigma-Aldrich) at 180°C for 4 min in an aerated oven. Crosslink density was determined by measuring the swelling ratio in toluene and the Flory-Rehner equation.9 A poly(tetramethylene oxide) (PTMO, MW ∼ 1000) soft segment polyurethane (PU) with 32.5 wt % of hard segments [4,4 0 -diphenylmethane diisocyanate (MDI) and 1,4-butanediol (BDO)] was supplied by AorTech Biomaterials (Scoresby, QLD, Australia) in film form.For the uniaxial deformation experiments, the stretching rig shown in Figure 1 was developed in order to allow convenient loading of deformed specimens directly into the dielectric spectrometer. This rig has three essential parts: sample clamper, electrode, and stretching knob. A sample strip (10 mm  80 mm) was placed along the groove on the rig and drawn to a specific extension ratio (λ). An electrode was positioned in the middle of the strained sample and clamped. The strained sample-electrode assembly was then separated from the rig and dried in vacuum for 24 h before measurement.The influence of uniaxial extension on the dynamics of the cross-linked PI (cross-link density = 6.1  10 -5 mol/cm 3 ) is displayed in Fi...