Heterogeneity in Crosslinked Polymer Networks: Molecular Dynamics Simulations

Crosslinked network polymers are a widely used class of chemicals and have been utilized since people used any form of paint, e.g., with proteins (casein, tempura) or unsaturated vegetable oils. Modern high-performance technology relies on different chemistry, such as epoxies, polyurethanes, and rubbers, but the continual search for improvements relies on increasing our understanding of how two or more reactive precursors form the desired network. When these modern technologies were first implemented, considerable advancements could be made rapidly; more recently however, as with all mature technologies, further progress is much more difficult and requires a greater degree of insight. Networks are formed from small molecules that react with other small molecules, and the resulting network topology depends on the specific interactions (chemistry) and the distribution and availability of these precursors, which is governed by their random location and motion. Thus, it is reasonable to expect that one can describe the resulting networks using statistical approaches which have been very successful at describing how representative molecular properties affect macroscopic network properties [1–4]. A simple calculation provides some perspective. We might consider a coating has failed when the density of visible features such as rust spots and blisters reaches a concentration of 1 per square centimeter. A polymer coating 1 cm in area might have a typical thickness of 50 μm and thus a volume of 5 10þ21 Å. If we approximate the volume of most atoms as 1 Å, it is easy to appreciate that, while