Guest Seminar by Department of Chemical and Biological Engineering - Artificial intelligence and Automation in Chemical Kinetic Modeling

The chemistry of simple molecules can be extremely complex. Thermal decomposition, combustion, and pyrolysis processes often involve hundreds of intermediate species and hundreds of thousands of elementary reactions between those species. In the last decade, computer-aided kinetic modeling software has been developed to deal with those large kinetic models. Two examples of such software are (1) the open-source Reaction Mechanism Generator (RMG) developed in the Green Group at Massachusetts Institute of Technology and (2) Genesys developed at the Laboratory for Chemical Technology at Ghent University. Both have demonstrated success in automatically developing kinetic models for the pyrolysis of oils, high-temperature pyrolysis of natural gas, combustion of biofuels, etc. The number of significant species and reactions in gas-phase kinetic models increases exponentially with the number of heavy atoms in the fuels. One of today’s challenges for automatic kinetic modeling software, and one of the highlights of this talk, is how to deal with detailed elementary-step kinetic models for molecules with more than ~6 heavy atoms, for molecules with heteroatoms and for surrogate mixtures.

These kinetic models contain so many thermodynamic and kinetic parameters (e.g. k’s, Keq’s) that they cannot all be determined experimentally. Instead, most of those parameters are determined automatically using structure-activity relationships. The few thermodynamic and kinetic parameters that are sensitive towards the concentration of certain desired products, are typically refined with high-level theoretical calculations or experimental measurements. Estimates of thermodynamic and kinetic parameters are available in a user-friendly form on rmg.mit.edu. Recent advances in artificial intelligence and machine learning for applications in chemical engineering have opened a new route for the fast and more accurate prediction of such chemical properties. This talk will center around the application of machine learning using message passing neural networks for the fast predictions of thermodynamic and kinetic properties in the framework of computer-aided chemical kinetic model development. More specifically, this talk will focus on the use of hybrid machine learning models for the prediction of solubility limits.