Linear baseline interpolation for single-process DSC data-Yes or no?

Theoretical simulations were used to evaluate performance of the linear baseline interpolation for the consequent kinetic analysis of single processes under different circumstances including the effects of varying height of the heat capacity step between the reactants and products, or varying asymmetry of the interpolated kinetic peak. In general, height of the Cp step was found to have larger distorting influence than the peak asymmetry. The model-free kinetic methods were found to be robust enough not to be significantly affected by the distortions caused by using linear interpolation (the performance was compared to that of the most accurate and physically meaningful tangential area-proportional interpolation) - the largest errors in activation energy did not exceed approx. 3%. On the other hand, model-based analysis has shown that usage of the linear interpolation leads to a significant distortion of the shape of the kinetic peaks and large deviations in the integrated area of the kinetic peak (corresponding e.g. to specific heat associated with the process studied by differential scanning calorimetry). Distortions of the peak shape consequently led to often largely deviated values of kinetic exponents determined for the correctly preselected known kinetic models. Model-based kinetic analysis of real-life experimental data treated via linear interpolation could in such case lead even to a selection of an incorrect kinetic model.