Role of Effective Composition on Dynamics of PEO−PMMA Blends

Abstract

We use molecular simulation to compare component dynamics of a poly(ethylene oxide) (PEO) and poly(methyl methacrylate) (PMMA) blend with that of a diblock copolymer of the same overall composition. The blend and the copolymer have different intermolecular packing, which leads to a difference in compositions defined over local length scales. These directly impact dynamic behavior through the chain connectivity model for blend dynamics, which is based on a single controlling length scale for dynamics equal to the Kuhn length. By comparing the change in dynamics expected on the basis of different concentrations with the actual change observed in the simulations, we find that this idea is quantitatively accurate for the PMMA component. For PEO, the controlling length scale for dynamics varies with the size of the observation volume. In the concentration fluctuation model, 3,6,7 the local composition differs from the bulk value because of fluctuations related to thermodynamics. The magnitude of these fluctuations depends on proximity to the critical point and depends on component molecular weights, the use of deuterium labels, and the interaction parameter, � . The relevant dynamic length scale is the cooperative volume as described by Donth for the glass transition. 8,9 This varies but can be regarded as on the order of 10-30 nm. In the chain connectivity model, 10 the local composition differs from the bulk value because of covalent bonding along the chain backbone. In this case, the controlling volume for dynamics is the Kuhn length, lk, of each component, on the order of 10 A. This volume is sufficiently small that chain connectivity will significantly alter its composition compared to bulk values. In what follows we consider two systems in which the local composition varies over length scales of 5-25 A, while keeping other system variables (bulk composition, identities of the two components) constant. This is accomplished by comparing a blend of two homopolymers with a diblock copolymer in which the same two polymers form the two blocks, with lengths that result in the same overall composition. We use molecular simulation, with which the local composition is easily characterized as a function of length scale. Because of the small size of the simulated chains, concentration fluctuations related to thermodynamics are minimal, thus isolating the role of local composition arising from local packing and connectivity. The two systems present different intermo- lecular packing, resulting in different values of the local composition throughout the length scales mentioned above. These length scales are not relevant to concentration fluctuations but provide a useful test of the chain connectivity model. Specifically, it will address the question of the appropriate controlling length scale for mixture dynamics. In the chain connectivity model, the local composition of each component is defined via an effective concentration: 10