A. Vázquez

Study of the kinetic and crosslinking reaction of novolak with epoxy resin

Novolak and epoxy reaction was studied by means of differential scanning calorimetry. Two reactions can take place in the mixture: one between OH groups of the novolak and epoxy groups, and the other between secondary OH groups (coming from the opening of the epoxy ring by the first reaction) and epoxy groups. Since only one peak was detected in dynamic thermograms obtained at different heating rates, both reactions occurred simultaneously. A global kinetic model with two kinetic constants was developed. It agreed with dynamic and isothermal experiences. The reaction heat was determined by an average of several runs as 23.1 kcal equiv. −1 (96.7 kJ/equiv.epoxy). The glass transition temperature was determined for different novolak/epoxy stoichiometric ratios from 0.04 to 2.2 The maximum glass transition was obtained for the 1:1 ratio and it was 99°C. The activation energy obtained from gel—time measures was 15.8 kcal equiv. −1 (66 kJ equiv. −1 ). A statistical structural model was used in order to describe the epoxy—novolak cure. Equal reactivity between OH groups of the novolak, equal reactivity between epoxy groups and no-substitution effect are taken into account. Intramolecular reaction in the pre-gel stage is neglected. Expression of the average molecular weight and gel conversion are derived. It is shown from the model that for the stoichiometric ratio of OH groups to oxirane groups, the predicted gel conversion agreed with the experimental value.

The first stages of an epoxy homopolymerization initiated by piperidine

Abstract Reactions taking place in the phenyl glycidyl ether (PGE)-piperidine system were followed by size exclusion chromatography using ultra-violet detection at 254 nm. The first step was the addition of piperidine to PGE to give a tertiary amino alcohol (TAA). This reaction took place by both catalytic (E = 50.7 kJ mol−1) and non-catalytic mechanisms. No other reaction could be detected before the piperidine concentration was depleted. The reaction of PGE with TAA was followed at 120°C. The main product was a TAA-PGE addition compound, following a pseudo-first-order kinetic rate equation. After this reaction, the PGE oligomerization takes place at room temperature.

Analysis of kinetic parameters of an urethane–acrylate resin for pultrusion process

An urethane–acrylic resin for a pultrusion processing application was studied. The concentration of Perkadox 16 and methyl methacrylate (MMA) was changed in the formulation mixture. A calorimetric study was performed in a DSC equipment. Isothermal runs from 42 to 60°C were performed to obtain a kinetic model for the polymerization reaction. Conversion of vitrification as a function of temperature was determined and the total heat of reaction as a function of MMA content was also measured. A general kinetic model was applied. An autocatalytic model and master-curve approach with an order of reaction of n + m = 2 and an activation energy of 95.9 kJ/mol were found. By the application of the Kissinger model for dynamic runs, an activation energy of 88.8 kJ/mol was obtained.

Effect of etherification reaction on extraction and photoelastic and dynamic mechanical behavior of diepoxide–diamine networks

Sol fraction and equilibrium photoelastic and dynamic mechanical behavior of epoxide networks based on bisphenol A diglycidyl ether (DGEBA) and poly(oxypropylene diamine) (Jeffamine® D-400) with four initial molar ratios of epoxy (E) and amine (A) groups, rE = [E]0/(2[A]0) = 1.2, 1.5, 2.0, and 2.3 were investigated. Networks with different extents of total epoxy group conversions (including etherification), αE, were prepared for each rE value. Both the ratio rE and the conversion αE affected the value of the equilibrium modulus, G, and the weight fraction of the gel, wg. As expected, decreasing the rE ratio (at constant αE) and increasing αE (at constant rE) were accompanied by an increase in the modulus, G, and gel fraction, wg. The stress optical coefficient, C, is independent of αE decreasing with increasing rE. The frequency–temperature superposition could be performed for all networks; the temperature dependence of the horizontal shift factor, aTo, satisfied the WLF equation. The temperature and time positions of viscoelastic functions predominantly depend on the overall concentration of elastically active network chains νe, regardless of the values of rE and αE. While the shape of viscoelastic functions at the beginning of the main transition region depended on the detailed structure of the chain (number and length of pendant chains), the shape at the end of the transition was determined mainly by the concentration of elastically active network chains. An unexpected universal increase was found in the half-width of the maximum in the dependence of the superimposed loss compliance, J″p, on reduced frequency ωaT with increasing crosslinking density.