SHIFT IN POLYMER BLEND PHASE-SEPARATION TEMPERATURE IN SHEAR-FLOW

Recent experimental studies of polymer blends in simple shear flow have indicated large "shifts" of the apparent phase-separation temperature. These shifts are examined within the context of a nonequilibrium hydrodynamic theory of phase separation developed by Onuki and Kawasaki. A mean-field version of the hydrodynamic theory indicates that no true shift of the critical temperature T(c) should be observed in high molecular weight polymer blend melts. However, the hydrodynamic theory indicates a large "apparent shift" DELTA parallel-to T(c)(gamma) parallel to the flow direction, DELTA parallel-to T(c)(gamma) approximately gamma(8/15), if the scattering data naively fit to the Ornstein-Zernike function. This spurious "shift" should not be observed in scattering data along the normal to flow direction. These predictions accord qualitatively with experiments on sheared melt blends by Nakatani et al. The apparent shift of T(c) in high molecular weight polymer melt blends is coincidentally similar to the true shift of T(c) observed in small-molecule binary mixtures where DELTA-T(c)(gamma) approximately (gamma)1/3-nu, nu(Ising) = 0.63, is obtained from mode-coupling renormalization group theory. It is argued that a true shift of T(c) should be observed in sufficiently diluted polymer blends in low molecular weight solvents because of a crossover from mean-field to Ising critical behavior upon dilution.