Spatially Localized Catalysis Imprints Global Network Topology during Thermoset Curing

Thermosetting networks retain a structural memory of how they were cured. Cure temperature, diffusion, and catalyst loading are recognized levers on the gel point; catalyst spatial distribution at a fixed loading has received less attention. We introduce a catalyst-centered domain (CCD) construction that precures a separate domain around each Zn catalyst before assembly, applied to bisphenol A dicyanate ester (BADCy) curing. At matched loading and chemistry, five replicas per protocol give mean gel-point conversions of 51.5 ± 4.5% for CCD curing and 44.3 ± 5.3% for conventional single-cell curing, providing suggestive evidence of delayed gelation. Raw simulated Tg values are 629 ± 39 K and 567 ± 13 K, respectively; agreement of the CCD estimate with experiment is obtained only under the selected WLF extrapolation. Local structure matches at conversion, while the giant cluster diverges globally: the CCD giant is smaller, spectrally better-connected, and near-spherical, whereas the conventional giant is a larger, rod-like spanning cluster. Reaction-population analysis shows that inter-cluster reactions dominate the analyzed curing stages. These results link the compared network-construction protocols to differences in gelation behavior and network topology.