Francisco Martínez-Veracoechea and Fernando A. Escobedo. School of Chemical and Biomolecular Engineering, Cornell University, 120 Olin Hall, Ithaca, NY 14853
Block copolymers often form ordered phases that posses unique optical, rheological, and mechanical properties, making them attractive components in novel fibers, coatings, nanocomposites, and electronic materials. Consistent with the recent advances in synthetic techniques that produce novel mesogenic building blocks, our goal is to use molecular simulations to map out the still uncharted phase behavior of systems containing them. In this work, we use a coarse-grained description of the copolymer chains (i.e., dissipative particle dynamics fluid), together with continuum-space Monte Carlo and Molecular Dynamics methods, to study systems of diblock copolymers melts that have been “filled” with selective additives (e.g., solvent particles, homopolymer, and nanoparticles). Approximate phase boundaries were found via free-energy calculations and great care was taken to enact the commensurability of system size with the unit-cell dimensions of distinct candidate phases. We focus on the stabilization of bi-continuous phases and the strikingly different phase behavior observed when the nature of the selective filler is changed. Our results elucidate the origins of the packing frustration that limits the viability of the gyroid, double-diamond, and plumber's nightmare phases and provide insights for overcoming it. In addition, a novel phase is observed wherein the minority component forms cylinders of two different diameters that alternate in a square lattice. Attention is also focused on directly determining the areas of phase diagram where macro-phase separation occurs. Moreover, we compare the particle-based simulation results with a implementation of self-consistent field theory that allowed us to map out the phase diagram over wider ranges of parameter space. Finally, combination of theory and simulation were found to be synergistic given that the former allows for fast exploration of parameter space together with direct determination of free energy values, while the latter incorporates fluctuation effects and requires no initial guesses for the phases to form.