(428i) Atomistic Simulation Studies of the Substitution Effect in Amine-Cured Epoxy Thermosets

Recent atomistic-level simulation studies of epoxy thermosets have demonstrated the feasibility of using simulation methods to make accurate predictions of experimentally-measured properties of these cured resin materials. Thus, for example, it has been shown that the effect of the molecular architacture of the epoxide component on the small strain elastic constants of crosslinked materials cured with the same amine (4,4'-Diaminodiphenylsulfone) can be predicted with a high degree of accuracy and precision [1]. Additional simulation measurements of the actual gel point in a second epoxy system, which has been studied experimentally [2a], have also been found to agree well with the experimental data [3]. Moreover, delay of the gel point relative to the predictions of the well-known Flory-Stockmayer theory of network forming polymerization, observed experimentally to occur in this and other network forming systems and attributed to the occurrence of intramolecular reactions which do not contribute to formation of a network, has also been demonstrated in the atomistic model systems.

It has long been understood that a number of factors may potentially affect the gelation conditions, structure and resulting properties of such systems. Thus, as studied extensively by a number of groups, and summarized by Dusek [4], factors affecting network topology include the chemical functionality of the reactants, relative reactivities of the groups involved in the reaction, composition of reactants, and details of the various reactions involved in crsslinking.

The role of relative reactivities, most commonly characterized by the ratio of rate constants of the species involved in curing - also referred to as induced unequal reactivities, and as the substitution effect - has an effect on the gel point resembling that of intramolecular reactions, tending to increase the extent of reaction at gelation by a few percent. Experimentally however the complexities of the reaction chemistry, which translate into uncertainty in the effective rate constants, have led to differing estimates of the magnitude of the substitution effect, typically ranging from 0-4 percent delay in gel point [2b, 5].

Although cross-linked resin formation inevitably involves random and possibly inhomogeneous systems on both the macroscopic and atomistic length scales, our previos studies of gelation in diglycidyl ether of bisphenol A systems cured with poly(oxypropylene) diamine have demonstrated that computing model gel points with precision somewhat below 4% is quite feasible provided one samples a modest number of independent models [3]. Moreover, varying the relative probability of bond formation between epoxide and primary or secondary amine groups is computationally straightforward. Accordingly, in this presentation we will discuss the effect of induced unequal reactivity on the gel point and, in addition, will present comparisons of the small strain elastic constants of thermoset models created both with and without the assumption of unequal reactivities.

References

1. Rigby, D., Saxe, P., Freeman, C. and Leblanc, B., in Advanced Composites for Aerospace, Marine and Land Applications, S.T. Sano, T.S. Srivatsan and M.W. Paretti (eds), John Wiley (2014).
2. Tanaka, Y., Stanford, J.L. and Stepto, R.F.T., Macromolecules 45, (a) 7197, (b) 7186 (2012).
3. Rigby, D., Saxe, P.W., AIChE Annual Meeting, San Francisco, CA Nov 16, 2016.
4. Dusek, K. in Rubber-Modified Thermoset Resins, C.K. Riew and J.K. Gilham (eds), ACS Adv. Chem. Ser. 208, 3 (1984).
5. Rozenberg, B.A., Adv. Polym. Sci. 75, 113 (1985).