A grand challenge facing the chemical kinetics science is automated generation of predictive kinetic models, e.g., for predicting drug shelf life or for optimizing alternative fuel mixtures. Predictive models are crucial for decision making and allow to affordably and feasibly predict the reactivity of numerous chemical systems. Often, there are orders of magnitude more parameters required for a detailed kinetic model than available from experimental data; the properties of many reactions are challenging to measure under conditions of practical interest, and the capacity to conduct experiments is naturally lower than the immense demand for data. At the same time, computational chemistry has already advanced to the stage where it has quantitative predictive capability for rate coefficients of complex reactions, making it an indispensable tool for the development of reliable kinetic models.<br/><br/>We have developed a high-level routine, The Tandem Tool (T3) for automated kinetic model generation and refinement, that iteratively calls a model generation software, in this case the Reaction Mechanism Generator (RMG), and an automated ab-initio refinement software, in this case Automated Reaction Calculator (ARC). T3 simplifies the required input from the user, automatically selects important parameters from a kinetic model, passes them to automated refinement, and feeds the information back to the modeling software. Both tools are written using the adapter/factory design pattern which abstracts away specific implementations of key external packages and facilitates high modularity.<br/><br/>In this talk we will present T3 and demonstrate its applicability. Selected challenges in the field such as automated identification of non-physical species proposed by the computer will be discussed, and efforts toward building an open-source community database for ab-initio computations in the context of chemical kinetics (TCKDB) will be presented.

Tandem tool (T3) for automated kinetic model generation and refinement

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.