Flash Point and Cetane Number Predictions for Fuel Compounds Using Quantitative Structure Property Relationship (QSPR) Methods

In the present work, we report the development of models for the prediction of two fuel properties: flash points (FPs) and cetane numbers (CNs), using quantitative structure property relationship (QSPR) approaches. Compounds inside the scope of the QSPR models are those likely to be found in alternative jet and diesel fuels, i.e., hydrocarbons, alcohols, and esters. A database containing FPs and CNs for these types of molecules has been built using experimental data available in the literature. Various approaches have been used, ranging from those leading to linear models, such as genetic function approximation and partial least squares, to those leading to nonlinear models, such as feed-forward artificial neural networks, general regression neural networks, support vector machines, and graph machines. Except for the case of the graph machine method, for which the only inputs are the simplified molecular input line entry specification (SMILES) formulas, previously listed approaches working on molecular descriptors and functional group count descriptors were used to build specific models for FPs and CNs. For each property, the predictive models return slightly different responses for each molecular structure. Thus, final models labeled as "consensus models" were built by averaging the predicted values of selected individual models. Predicted results were compared with respect to experimental data and predictions of existing models in the literature. Models were used to predict FPs and CNs of molecules for which to the best of our knowledge there is no experimental data in the literature. Using information in the database, evolutions of properties when increasing the number of carbon atoms in families of compounds were studied.