Monthly Archives: June 2014

Cure Kinetics of Thermosets with DSC Analysis, part 1

For people who work with thermoset polymers, cure information is important from manufacturing control to safe storage and transportation to cure process.  In general, cure information of thermosets is characterized by DSC experiment and analysis.  For most purposes a single DSC scan (i.e., with a 10C/min heat ramp) will suffice, which will give users its overall reactivity behavior.

However, if there is a need in numerical modeling for prediction purposes of thermal management, a lot more work is required so that cure kinetic parameters can be generated.  It would involve several DSC scans of various heat rates and/or several isothermal DSC scans at various temperature values.

The chemical reaction rate depends on two factors, the reaction rate constant and the conversion.  The reaction rate constant uses the Arrhenius expression where the pre-exponential factor and activation energy need to be determined from DSC experiment.  The pre-exponential factor and activation energy are assumed constant over the entire range of conversion (from 0 to 1).  The conversion expression can be expressed as a mathematical equation were one is assumed, for example, n-th order model or autocatalytic model.  If so, additional kinetic parameters such as the reaction order and other reaction constants that are expressed in the mathematical equation would be determined from these DSC experiment as well.  The beauty of using this approach is that everything is expressed in an elegant equation and parameters are constants.  The downside however, is that the fit is at times not good enough.

Instead of using a specific equation to describe the conversion expression, one can apply the differential isoconversional method, which is now commonly referred to as the “model-free” approach.  With model-free approach, only the pre-exponential factor and activation energy are the parameters to be determined.  However, these parameters are no longer constant and they can vary over the range of conversion.  It then becomes a surface-fitting exercise to determine these two parameters with respect to conversion.  With powerful computing resources and commercially available software nowadays, it is becoming easier to analyze kinetic data using this approach.