Glass transition of polymer melts: test of theoretical concepts by computer simulation

摘要: Abstract Polymers are good glass formers and allow for the study of melts near transition in (meta-)stable equilibrium. Theories imply such an equilibrium can, hence, be tested by polymer melts. After a brief summary basic experimental facts about polymers, main theoretical concepts reviewed: mode coupling theory (MCT), entropy theory, free-volume idea growing length describing size cooperative regions, etc. Then, two coarse-grained models polymers described, which have been developed aiming at test these concepts. The first model is bond-fluctuation on simple cubic lattice; second bead-spring continuum. While studied kinetic Monte Carlo methods, molecular dynamics simulation: issue addressed aspects results model-dependent generic should universally apply, including real materials. Attempts to include chemical detail order describe materials discussed, too. It shown that idealized MCT description onset slow relaxation (‘cage effect’) discussed melts, but singularities critical temperature rounded, correct close calorimetric remains controversial. configurational melt does decrease strongly when approached, evidence presented ‘entropy catastrophe’ Gibbs Di Marzio artifact inaccurate approximations. Simulations confined thin film geometry also reviewed, emphasizing question extent they shed light correlation length. Finally, we discuss implications motion larger scales. short (non-entangled) chains Rouse stays essentially valid, friction coefficient reflecting slowing down as described α-relaxation, while small-scale motions (e.g. β-relaxation regime MCT) connectivity not very relevant. We conclude theories some vitrification phenomena correctly, unified puts all into one coherent framework, yet exist.