Simulation and non-linear optimization of Šesták-Berggren kinetics

Abstract

The effect of procedural conditions (density of points, initial value of conversion rate) on the accuracy of Šesták-Berggren (AC) simulations was studied. Newly developed software programmed in R-language and using the LSODA initial-value-problem-solver was used to simulate sets of AC kinetic data-sets with increasing precision. Evaluation of the kinetic data-sets was done using the benchmark commercial software, Thermokinetics (Netzsch), to consolidate uniformity of the data treatment. Whereas the points density used during the simulation had no effect on the resulting shape of the kinetic peak (and the corresponding kinetic description), the initial value of conversion rate was found to crucially influence the shape/parameters of the resulting kinetic peak. Importance of utilizing low enough initial values of conversion rate in the AC simulations dramatically increases for positively asymmetric kinetic peaks. Based on the present results, transferability of the kinetic parameters evaluated for the generic/experimental data by different softwares was discussed.