Characterization of epoxy and polyimide cure by UV-visible and fluorescence spectroscopy: Azochromophoric labeling approach

Characterization of crosslinked polymers remains one of the challenging problems in polymer science. This dissertation describes a quantitative methodology developed to probe cure reactions in network polymers. The approach is based on a reactive dye labeling technique in which azochromophoric dye mimics the curing agent and its absorption and emission behavior provide information on the curing process.^ In epoxies, p,p$\sp\prime$-diaminoazobenzene (DAA) is used to mimic the behavior of diaminodiphenylsulfone (DDS). The epoxy systems investigated include two diepoxides, diglycidyl ether of bisphenol A (DGEBA) and diglycidyl ether of butanediol (DGEB); and a tetraepoxide, tetraglycidyl diphenyl methane (TGDDM). All of these were cured with DDS. As the polymer is cured, the $\lambda$ max of the $\pi$ to $\pi$* transition corresponding to the azo bond of DAA red-shifts in a way that provides spectral discrimination of the cure products (cross-linkers, branch points, linear chains, chain ends and unreacted diamines). We quantified the products, determined the kinetic parameters, and made some structure-property correlations. Distinction of the cure species has not been possible before and the results obtained give us insight into the cure mechanism. Fluorescence due to the DAA label increases sharply due to the increasing fluorescence quantum yield of the cure products rather than due to viscosity changes. Thus fluorescence intensity was used to estimate cure composition based on the kinetic differential equations.^ The kinetics and mechanisms of thermal imidization of polyamic acid made from a conjugated diamine (DAA) and a non-conjugated dianhydride (6F-DA) were investigated both in dilute solution and in solid state. We quantified the products, obtained kinetic parameters, and proposed a much better defined mechanism for the solid film and dilute solution reactions. If degradation is assumed negligible, then dilute solution results are adequately modeled by assuming a first-order, two-step imidization reaction. On the other hand, solid film results seem to be reasonably simulated by assuming a second order reaction for the first step, followed by a first-order reaction for the second step to account for the decrease in the concentration of the conformationally favorable catalytic amic acid group.