Dynamic Heterogeneity in Fully Miscible Polymer Blends through Molecular Simulations

SUMMARY: The effect of self-concentration and intermolecular packing on the dynamics of polymer miscible blends is examined by extensive atomistic simulations. Direct information on local structure of the blend system allows a quantitative calculation of selfand effective composition terms at various length scales that are introduced to proposed models of blend dynamics. Through a detailed statistical analysis, the full distribution of relaxation times associated with reorienation of carbon-hydrogen bonds was extracted and compared to literature experimental data. A direct relation between relaxation times and local effective composition is found using a composition-dependent length scale. The pertinent dynamic length scale that best describes the slow segmental dynamics in miscible blends relates to both intraand intermolecular contributions. INTRODUCTION: The dynamics of polymer mixtures remains an area of intense research for nearly two decades due to their complex dynamical and rheological behavior. It is well established that even thermodynamically miscible blends, can retain distinct individual mobilities in the mixed state that are separate from the pure components. A critical parameter in the observed behaviour is the dynamic asymmetry, controlled by the difference in the glass transition temperatures (Tg) of the constituent homopolymers. In this aspect several open questions pertaining to linking molecular details to model parameters; see refs. 1-6 and references within. Several models combine concentration fluctuations and contributions from chain connectivity to provide a framework that rationalizes the observed experimental behavior. Concentration fluctuations are expected to be present in mixtures and depending on their lifetime they can promote a distribution of segmental relaxation times. Besides that the effect of self-concentration can also be of great importance for describing the dynamics of miscible polymer blends. According to this concept (Lodge-McLeish, LM, model), each segment of a specific component A is experiencing an environment that is enriched to A due to chain connectivity. In the present work, we examine whether the dynamics in fully miscible polymer blends as described by atomistic simulations, can be predicted by employing the concept of self-concentration combined with composition fluctuations which lead to a distribution of relaxation times. Two different binary systems were examined: (a) Oligomers of polyisoprene (PI)/polystyrene (PS), and (b) Blends of polystyrene with oligostyrene. SIMULATION METHODS: Detailed atomistic molecular dynamics (MD) simulations were performed in the NPT ensemble maintaining a pressure of 1 bar at four different temperatures T: 413, 443, 473, and 503 K. All bonds were kept constant and the time step in the integration of the equations of motion was 1fs. For mixtures at least 200ns trajectories were generated, far beyond the relaxation time of these oligomers. More details about the all atom MD simulations employed in this study and the equilibration procedure are given elsewhere. The (atactic) PS model is based on a fully atomistic description that is described in the literature. We verified that the conformational and thermodynamic properties are reproduced faithfully in agreement to available experimental data. PI oligomers are also modeled using an all-atom model. The chain dimensions as well as the structure of PI bulk systems are in good agreement with available experimental data. Note also that the molecular lengths of both PI and PS were chosen in order to be very similar to PI/PS blends studied before through experiments. Blends with longer PI or PS chains will lead to strong miscibility problems. RESULTS AND DISCUSSION: Structure and Self Composition. We first discuss how components distribute within the oligomer melts. The LM model employs self, φself, and effective, φeff, local volume, or mass, fractions to correlate compositions around a polymer segment to the observed dynamic behavior.