Diffusion-induced growth of compositional heterogeneity in polymer blends containing random copolymers.

The compositional relaxation in random copolymer systems on a macroscopic scale is considered in theory. A set of diffusion equations is derived that describes the motion of chains of different composition and then converted into coupled equations for statistical moments of the compositional distribution. Several ways to solve the closure problem for these equations are discussed. The simplest is the situation when the shape of the transient compositional distribution can be predicted a priori, for example, a bimodal distribution is kept during interdiffusion of two copolymers that are not very close in composition. For a general case, it is shown that the cumulant-neglect closure based on the truncation of high-order cumulants is an effective method to get an approximate solution in terms of the time-dependent local mean composition and its dispersion. This method is applied to non-homogeneous compatible polymer systems, such as a random copolymer AB of a composition varying in space, a bilayer of Bernoullian copolymers AB of different composition, and a bilayer of homopolymers A and B, in which an autocatalytic polymer-analogous reaction A --> B takes place, with possibility of the neighbor group effect. It is found that the interdiffusion can lead to a substantial broadening of the local compositional distribution, which, in turn, accelerates the system dynamics and promotes chemical reactions.