In 1995, Philip Anderson wrote, “The deepest and most interesting unsolved problem in solid state theory is … the nature of glass and the glass transition” (Science 1995, 267, 1615). One year earlier, Keddie, Jones and Cory (Europhys. Lett. 1994, 27, 59) discovered that a 15-nm-thick polystyrene film supported on silica exhibits a glass transition temperature, Tg, that is reduced by more than 20 K relative to bulk Tg. Since then, dozens of studies have characterized how Tg values of polymers are modified by nanoscale confinement. However, most studies characterized only average Tgs across films and focused on a single polymer -- polystyrene.
In 2000, Pierre de Gennes (Eur. Phys. J. E 2000, 2, 201) challenged the glass transition and confinement research communities by stating, “Future experiments should aim not at the determination of a single Tg, but at a distribution of Tgs.” In response to this challenge and based on our interests in contributing to the understanding of the important effects of confinement in applications ranging from membranes and microelectronics to nanocomposites, we developed a simple fluorescence/multilayer method to determine distributions not only of Tg but also of physical aging and diffusion coefficients in polymer films and nanocomposites. We have found that perturbations to Tg at a free surface or polymer-substrate interface (with attractive interactions) can propagate several tens to hundreds of nanometers into a film and that the strength of the gradient in Tg is strongest at the free surface or interface. We have also determined that the length scales over which physical aging (slow relaxation toward equilibrium in the non-equilibrium glassy state) is affected by confinement can exceed those associated with changes in Tg. Finally, we have discovered that diffusion coefficients of dye molecules dispersed in polymer can be reduced by as much as a factor of 1000 with confinement of the polymer film.
Our studies have led to understanding why the effects of surfaces and interfaces in perturbing Tg depend strongly on polymer species. The Tg reduction associated with free surface effects is enhanced when the polymer is altered to increase its requirement for cooperativity in the segmental mobility associated with Tg. Whan H-bond formation occurs between chain segments and hydroxyl groups on nanofiller or substrate surfaces, increasing the strength or density of attractive interactions results in increasing enhancements of Tg and increasing suppressions of physical aging below Tg. The latter discovery suggests a new application for nanocomposites --producing glassy materials with properties that are stable over long-term use.
Finally, we shall describe behavior that result in new questions related to the fundamental nature of glass-forming materials. Using multilayer films, we have found that an ultrathin layer of one polymer can have its glass transition behavior strongly modified by or even “slaved” to that of neighboring domains of other polymers. Thus, multilayer films of different, immiscible polymers may yield properties that are not possible with conventional polymer blends.