Reaction Models from Reactive Molecular Dynamics and High-Level Kinetics Predictions

Outline

The need for understanding networks of reaction pathways arises from various fields of science and industry, such as combustion technology (EXC236) or polymerization (SFB985). For the optimal design of engines and reactors several global properties of the relevant reaction mechanisms have to be determined. In the field of combustion technology one of the most important global properties is the time the fuel requires to start burning, after ignition was triggered. Determining the delay between trigger and ignition is done experimentally e.g. by shock tube experiments. Deriving the parameters of the underlying kinetic model from this global quantity is an inverse problem. Although additional experiments could aid solving this inverse problem, experiments are expensive, time-consuming and require a minimum amount of fuel. The speed of ignition, i.e. the reciprocal of the ignition delay time, is dominated by the speed of reactions of the underlying mechanism. Investigating the reaction rates of the mechanism yields information about the fuel's combustion chemistry. In contrast, focusing on a single global property offers only a narrow insight on the fuels behavior. Therefore, the best way for predicting the combustion behavior of a fuel is to determine its reaction kinetics. Unfortunately, parameters of the mechanisms used to predict global properties still lack quality and quantity in reproducing reality. This is due to the vast amount of parameters to be determined by only few global properties. Even for the most simple combustion reaction equation, H2 + ½O2 à H2O, the mechanism of reaction includes eight different species and nineteen different reactions [1]. Such a small system might be investigated experimentally only. However, for larger reaction networks the number of unknown parameters exceeds the number of available independent experiments by an order of magnitude at least. Further, some of the species are short-lived and therefore hard to measure. The difficulties in obtaining the parameters of a mechanism from experiments in combination with their costs necessitate the use of computational methods. Ab initio calculations yield information about possible reaction pathways and reaction rates. Although the accuracy of ab initio calculations is often considered to be lower than the accuracy of experiments, results of computational investigations are even more reliable in some cases [2].

[1] Li, J., Zhao, Z., Kazakov, A., Dryer, F. L.: An Updated Comprehensive Kinetic Model of Hydrogen Combustion. International Journal of Chemical Kinetics, 36, 566-575, 2004

[2] Noble, B.B., Coote, M.L.: First Principles Modelling of Free-Radical Polymerization Kinetics. Internal Reviews in Physical Chemistry, 32 (3), 467-513, 2013