The insulating effect of the low thermal conductivity of epoxy on the resolution of the heat of reaction during polymerization by differential scanning calorimetry

Abstract

It is demonstrated that the heat of epoxy cure as measured by isothermal differential scanning calorimetry (DSC), as commonly calibrated by the heat of fusion of an indium standard, is consistently low as compared to methods “directly” calibrated by Joule heating [i.e., isothermal microcalorimetry (IMC)]. The discrepancy between measurement techniques is shown to be on the order of 15% for low thermal conductivity epoxy thermosets. In addition to direct comparisons between DSC and IMC measurements during epoxy polymerization, indium samples embedded in cured epoxy were studied to determine if the DSC was able to accurately capture the total heat of fusion through the polymeric insulating layer. It is found that the indium heat of fusion measured by DSC is lower when the indium is embedded in epoxy than when the indium is in direct contact with a steel sample pan. The fraction of the indium heat of fusion detected through the epoxy insulation by the DSC cell is comparable to the fraction of the heat of reaction detected by DSC during epoxy cure, as determined from the DSC-IMC comparison. It is concluded that the heat flow detected by DSC during epoxy cure must be scaled by a factor of 1.18 in order to accurately portray the full heat of reaction under the conditions used in this work. It is argued that the difference found between DSC and IMC is in reasonable agreement with qualitative heat transfer calculations and with previous thermal conductivity measurement of polymers via DSC. The specific scaling factor is anticipated to depend on the material studied, on the pan type, on the sample geometry, on the purge gas details, and on the instrumentation. Instruments used in this work include a Q2000 DSC and a TAM Air IMC (both TA instruments). The epoxy was diglycidyl ether of bisphenol A (DGEBA) cured with diethanolamine (DEA).