Effects of experimental variables on the determination of kinetic parameters with differential scanning calorimetry. II. Calculation procedure of Freeman and Carroll

Abstract

The effects of sample mass, heating rate and particle size on the determination of reaction order, activation energy and pre-exponential factor with the calculation procedure of Freeman and Carroll in differential scanning calorimetry was studied with factorial designs. Sodium bicarbonate, three organic hydrates (potassium oxalate hydrate, mercaptopurine hydrate, and sodium citrate dihydrate) and two substances giving solid-solid transformations (potassium nitrate and hexamethylbenzene) were studied. It was found that the extent of these effects depended on the nature of the decomposition and on the evaluated section of the transition peak. It was not justified to use data from the complete peak in the calculation, as different processes contribute to the total solid-state reactions. Arrhenius plots with such data were curved. In the initial section of the peak heat transport is the main process; high values of the kinetic parameters are then obtained. Lower values, with smaller error, were found for the middle section where mass transport is the predominant process. Narrow peaks gave rise to very high activation energies and large error. Even negative reaction orders were then found which are indicative of the explosive character of the transition. As the kinetic parameters differ considerably during the peak, it is not justified to attribute any physical meaning to them.