The Influence of Stoichiometry and Cure Conditions on Surface and Interfacial Properties of Epoxy Resin Barrier Coatings

Abstract

In the transport of corrosive, small molecule cargoes, ship-tanker holds are often covered with (2+ layer) amine-cured epoxy resin coatings to prevent damage to the tanker steel. However, over time these coatings can experience interlayer delamination or adhesion failure, and while the associated financial implications can be large, ultimately, relatively little is known about the causes of these adhesion failures. This thesis addresses the potential causes of interlayer adhesion failure in 2-layer epoxy systems by characterising and comparing system properties as a function of stoichiometry and cure condition.
Variation in cure conditions (temperature, relative humidity (RH) and delay time between epoxy component mixing and film casting, “induction time”) and stoichiometry significantly altered the surface properties of the ambient-cured, first coat of the epoxy resin system (Dow Epoxy Novolac, (D.E.N. 431), resorcinol diglycidyl ether (RDGE), and 4,4-diaminodicyclohexylmethane (PAC-M)). Gravimetric analysis showed that increasing induction time significantly reduced surface layer formation (carbamation) of cured epoxy resin coatings at 80% RH, but had no measurable effect at 40% RH and below. At lower RH, the likelihood of carbamate formation decreases as water catalyses carbamation. Therefore, to reduce carbamation, it is advised that films are cured at lower RH (<40%). RMS surface roughness increased with increasing RH and decreased with increasing induction time and ambient cure temperature, at two stoichiometric extremes. However, the net change in surface area arising from these conditions was not sufficient to significantly alter the equilibrium contact angles or wetting regime. The variation in surface wettability appears to be predominantly influenced by variations in surface chemistry rather than roughness. Stoichiometry emerged as the most influencial factor affecting surface wettability, average void volume and fractional free volume, while cure temperature significantly influenced the extent of cure at both stoichiometries. Off-stoichiometry formulation and elevated ambient cure temperature significantly increased system average void volume, while fractional free volume decreased.
Interlayer diffusion, and the formation of an interphase, was shown to occur in bilayer ambient cured epoxy resins systems and is likely to be significant regarding interlayer adhesion. This was achieved by mapping the ingress of probe molecules into ambient cured epoxy resins after overcoating, using Raman spectroscopy. This process was shown to occur over tens of microns during a 3-day period. Extent of diffusion depth was significantly increased by decreasing ambient cure temperature, overcoating interval and RH. Most impactfully, high carbamate formation almost completely suppressed second coat ingress suggesting that carbamate contributes to adhesion failure by inhibiting the interdiffusion of sequential layers.
Novel approaches, using micron-resolution, spatially-resolved solvent ingress measurements (Raman mapping; PIXE), were developed, and showed that the crosslink density homogeneity of films polymerised using predominantly amine addition was increased when the cure temperature was increased from 25 °C to 35 °C. However, little difference was observed between bulk and interphase crosslink densities. Moreover, neutron reflectometry (nanometre-resolution), showed that cure temperature, stoichiometry and overcoating interval all affected film crosslink density at the nanometre scale. In particular, increasing the cure temperature from 25 to 35 °C generally increased crosslink density homogeneity, as regions of increased crosslink density were observed near the surface of films cured at 25 °C. This temperature-dependent effect is likely attributed to the accelerated reaction kinetics at higher temperatures, promoting more uniform crosslinking throughout the material and reducing the formation of localised high-density regions observed at lower cure temperatures. However, throughout all films, no obvious signs of early-stage adhesion failure were observed. Given delamination is a fatigue associated issue, this result may imply that obvious signs of delamination are not intrinsically present in the film from synthesis, and instead are caused by the repetitive cycles of solvent exposure, followed by film drying, in industrial application.
Solvent ingress studies showed that epoxy resin system stoichiometry significantly affected ultimate ingress more than cure temperature. The time-dependence of solvent ingress indicated a behaviour neither purely Fickian (linear dependence of penetration of diffusant with the square root of time) nor Case II (linear dependence with the time), but appeared to be intermediate between these extremes.