What does the Tg of thin polymer films really tell us?

Abstract

The disparate results for Tg shifts in ultra-thin polymer films with the same film thickness, substrate, and chemistry raises the question of whether Tg is a useful way to characterize thin films. Underlying many Tg measurements is the assumption that Tg estimates used for bulk polymer are helpful to infer the dynamical changes in thin films. We use molecular simulations of a coarse-grained polymer film supported on a substrate with variable interaction strength with polymer to test this assumption. As expected, the slowing of polymer dynamics near a highly attractive substrate forms a bound layer that effectively insulates the reminder of the film from strongly interacting substrate. As a result, thermodynamic estimates of Tg is relatively insensitive to the polymer-substrate interaction strength ε when the substrate interactions exceed those of the polymer-polymer interactions. Due to this insensitivity, thermodynamically defined measures of Tg decrease with decreasing film thickness, regardless of ε. Dynamically defined measures of Tg also decreases with decreasing thickness when ε is relatively weak; however, the dynamic Tg increases when ε exceeds the polymer-polymer interaction strength. Our results show that these qualitatively different trends in Tg reflect differing sensitivities to the mobility gradient across the film. We emphasize that using single measure of Tg to infer changes in the dynamics of polymer thin films is limited.