A multiscale scheme for simulating polymer Tg.

All-atomistic (AA) molecular dynamics (MD) is considered as one of the desirable methods for studying glass transition temperatures (Tg) of specific polymers. However, heavy computational efforts are generally required, and the simulated Tg values are not always in good agreement with the experimental data. In this work, a multiscale scheme is proposed: first, the structural and volumetric properties based multiscale modeling is employed to parameterize the coarse-grained (CG) potentials against the AA simulations of an oligomeric melt; with the CG potentials, MD simulations are then carried out on a serial of oligomer bulks and polymer systems of interests, for which the dynamical Tg values are determined. With poly(ethylene oxide) and poly(methyl methacrylate) as typical examples, the simulated dynamical Tg values of the oligomeric bulks exhibit a linear relation with the empirical values, which is used to determine the "actual Tg" for the polymer bulk. The so-obtained Tg is found to compare very well with the experimental data. Such a computational framework can be quite promising in investigating the effects of various complex factors on polymer Tg. Graphical Abstract The actual Tg for a polymer can be reliably predicted by rescaling the simulated dynamical Tg.