Nanostructure of tetrafunctional epoxy resins and composites: Correlation to moisture absorption properties.

Abstract

The effect that changes in network topology, while maintaining a constant network polarity (i.e. thermodynamic driving force was kept constant), had upon the moisture absorption properties of an aerospace grade tetrafunctional epoxy (TGMDA) cured with multifunctional amines were investigated. Utilizing Positron Annihilation Lifetime Spectroscopy (PALS) to characterize the nanoscale structure of these epoxies, it was found that as the static hole volume (a measurement of packing defects at 0K) increased so did the equilibrium uptake. PALS studies of one of these resins cured to varying extents, found that this static amount increased with degree of cure indicating that the network becomes more open as a direct consequence of crosslinking. Polar groups, which are the attractive force for diffusion, are in the vicinity of these crosslinks, therefore it is believed that the increase in static hole volume results in exposing more polar groups for absorption. The diffusion coefficient, which is representative of the kinetic aspect of diffusion, was also investigated. It was discovered that the amount of nanohole volume in the polymer; whether the total, the static, or dynamic (i.e. thermally activated) does not correlate to the diffusion coefficient in anyway. Furthermore, at an isotherm the diffusion coefficients for all these materials were relatively constant. From this it is hypothesized that it is the similar sub-T$\rm\sb{g}$ motions of these resins which is the rate limiting step in diffusion. This was bolstered by the fact that the activation energy for diffusion and for the sub-T$\rm\sb{g}$ motions for these epoxies are of the same order of magnitude. The nanostructure of fiber reinforced epoxy composites (i.e. a boron/epoxy and a graphite/epoxy) were probed with the bulk PALS technique as well. It was observed that for the graphite/epoxy composite and its flash (i.e. no fibers present) cured under identical conditions, that the nanoholes in the composite were larger than those present in the flash at temperatures below the epoxy's T$\rm\sb{g}.$ Curiously the boron/epoxy composite and its flash showed an opposite trend. Several potential explanations were examined. The only viable explanation for the observed nanostructural differences between the flash and the resin in these composites utilizes a micromechanics approach involving the CTE mismatch between the fibers and the matrix material. In this approach it is proposed that the fibers in the composite act as a constraint, preventing the nanohole from freely contracting (upon cooling through T$\rm\sb{g})$ in the axial direction, while Poisson's ratio forces the holes to contract more in the transverse direction than the unrestrained hole in the flash. Therefore the resultant nanoholes in the composite maybe elongated in the fiber direction and shortened in the transverse direction when below the curing temperature. When the PALS technique probed these elongated holes it averaged their dimensions (but weighted the shortest dimension more heavily), thereby yielding the observed results. Despite slightly smaller static holes in the boron/epoxy composite than its flash, no difference in equilibrium uptake was noticed. The diffusion coefficient for the epoxy resin in this composite was found to be an order of magnitude higher than its flash. Nanostructure is not believed to be the cause of this but rather the glass fiber scrim cloth utilized in the processing of the prepreg.