Frontal Cationic Curing of Epoxy Resins SERGIO SCOGNAMILLO, 1 CHRIS BOUNDS, 2 MICHAEL LUGER, 2 ALBERTO MARIANI, 1 JOHN A. POJMAN 2 1 Dipartimento di Chimica, Universita ` degli Studi di Sassari, Sassari, Italy 2 Department of Chemistry, Louisiana State University, Baton Rouge, Louisiana 70803 Received 16 December 2009; accepted 6 February 2010 DOI: 10.1002/pola.23967 Published online in Wiley InterScience (www.interscience.wiley.com). ABSTRACT: We studied the frontal curing of trimethylolpropane triglycidyl ether (TMPTGE) using two BF 3 -amine initiators and two fillers, kaolin and fumed silica. In the case of kaolin, the range of concentrations allowing for frontal polymerization to propagate was dependent on its heat absorption effect whereas in the case of silica it was a consequence of the rheo- logical features of this additive. However, for both systems the velocity and front temperature show the same trends; in all cases front velocities were on the order of 1 cm/min with front temperatures about 200  C. V C 2010 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 48: 2000–2005, 2010 KEYWORDS: cationic polymerization; composites; curing of polymers; epoxy resins; frontal polymerization INTRODUCTION Thermal frontal polymerization (FP) is a localized reaction that propagates through the coupling of thermal diffusion and the Arrhenius dependence of an exo- thermic polymerization reaction. The result is a localized thermal reaction zone that then propagates through the reac- tants as a thermal wave. 1–5 A comprehensive bibliography of frontal polymerization research can be obtained at the web- site listed in reference. 6 Most of the FP systems studied to date have been based on free-radical chemistry. 6–19 Free- radical systems have the advantage of rapid reaction at their adiabatic reaction temperature but low rate of reaction at room temperature. They were ignited by UV 9,20–23 or ther- mal radiation. 24–26 Frontal curing of epoxy–amine systems was first studied by Arutiunian et al. 27 and Surkov et al. 28 and later Mariani et al. 29 modeled the system. Chekanov et al. studied FP of amine–epoxy systems using diglycidyl ether of bisphenol A (DGEBA) as the epoxy. 30 FP with amine-cured DGEBA and clay was studied by Mariani et al. 31 Using a cationic photoinitiator along with a benzoyl peroxide, thermal FP with an epoxy resin was achieved using UV light to ignite. 23 Crivello also considered such hybrid free-radical/ cationic frontal photopolymerizations. 32 Crivello developed photoactivated cationic ring-opening frontal polymerizations of oxetanes. 33–37 The short pot life of amine-cured epoxy systems is a prob- lem. We would like to have a system that was premixed and would remain unreacted for weeks if not months. Latent cat- ionic catalysts provide some hope but if the catalyst and resin will not react at room temperature for a long time, then a high curing temperature is required. Pojman et al. studied a binary frontal polymerization system that con- tained a BCl 3 -amine complex. 38 We chose two BF 3 -amine cat- alysts that have pot lives of several hours and would readily support frontal curing. To suppress buoyancy-driven convec- tion, we added filler. We used kaolin (Polygloss 90 V R ), a typi- cal hydrous alumino-silicate clay 39 and fumed silica (Aerosil). We studied how the front velocity and front temperature were affected by the amount of catalyst and the amount of filler. EXPERIMENTAL Materials The BF 3 -amine initiators (Leecure B-110 and B-950) were obtained from Leepoxy. B-110 and B-950 have a reported gel time at 80  C of 5 and of 11 min, respectively. Fumed silica (Aerosil) was obtained from US Composites Aerosil. The kao- lin was the clay PolyGloss 90 V R from Huber. Trimethylolpro- pane triglycidyl ether (TMPTGE) was used as received from Aldrich. Preparation of Reactant Mixture In a typical run, a mold made of wood (50 mm  20 mm  20 mm) was loaded with a 10 g of TMPTGE and the appropriate amount of initiator calculated in terms of phr (weight parts per hundred resin); these last components were thoroughly mixed to obtain a homogeneous mixture. The proper amount of filler was added to this, and the resulting mixture was scrupulously mixed again. The mold was kept open because our focus was to create a system as much as possible similar to real-world applications. Any- way, the high viscosity of the system prevented significant surface variation. Correspondence to: J. A. Pojman (E-mail: john@pojman.com) Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 48, 2000–2005 (2010) V C 2010 Wiley Periodicals, Inc. 2000 INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA