文摘
We show that the time corresponding to the peak of the segmental relaxation time distribution andthe mean time of that distribution for the components of several miscible polymer blends are strongly affectedby both chain connectivity and concentration fluctuations. These two measures of characteristic segmental relaxationtimes differ from the time corresponding to the mean composition experienced by a segment, with these differencesbeing emphasized for blends with large glass transition contrast, lower temperatures, or increased concentrationfluctuations on the nanometer scale. These findings are in contrast to self-concentration models, which generallyassume that concentration fluctuations affect neither the mean nor the peak segmental relaxation times and areonly relevant for determining the distribution of relaxation times. Going further, we show through the inclusionof self-concentration and concentration fluctuation effects that segmental dynamics are only affected by a localenvironment of size ~1 nm surrounding a test monomer. This length scale is only weakly temperature andcomposition dependent, even near Tg. This estimate of a relevant dynamic length scale is in good agreement withthe conjecture on which the Lodge-McLeish self-concentration model is based but is contrary to the ansatz usedby many concentration fluctuation-based models which assume that this local environment size diverges in thevicinity of the glass transition.