Dynamics of miscible polymer blends: predicting the dielectric response

Abstract

We present a predictive model for the dynamics of miscible polymer blends by assuming that local composition variations determine segmental dynamics. We then properly incorporate the distributions of intrachain (“self-concentration”) and interchain (“concentration fluctuations”) contributions to the compositions surrounding a polymer segment. In a manner similar to our past work we then derive the distribution of relaxation times for both components. These distributions depend explicitly on the size of the volume in which fluctuations are sampled by the test segment. We find by fitting our model to an extensive body of data on various miscible blends (PI/PVE, PBO/PVE, PVME/PS, and PVME/P2ClS) that an essentially composition-independent value of the correlation volume reproduces all available segmental relaxation data at temperatures above the glass transition of the low-T_g component. We have shown that the apparent temperature dependence of the size of the correlation volume is very weak at all temperatures above the T_g of the low-T_g component. The size of the correlation volume is found to be comparable to the Kuhn segment length of the chains. While our approach thus resembles the model of Lodge and McLeish, it is important to emphasize that we include the distributions of compositions experienced by a test segment instead of invoking a mean field. The appropriate choice of the size of the correlation volume and proper incorporation of concentration fluctuations are vital to simultaneously model the peak segmental time and the width of the relaxation time spectra. By comparing PI/PVE with PBO/PVE and comparing PVME/PS with PVME/P2ClS, we show that the size of the correlation volume is apparently a polymer-chain-specific property that does not depend on the blend composition and blend partner.