Automated generation of large reaction mechanisms for heterogeneous catalysis
A major goal in catalysis is to predict the performance of novel materials. Accurate prediction of the reactivity and selectivity of catalytic materials requires a detailed list of elementary surface reactions, or a microkinetic mechanism. The microkinetic mechanism is the critical bridge between the electronic potential energy surface and the catalyst’s performance under industrially relevant conditions. Unlike more traditional models used in kinetics (e.g. pseudo-steady state, quasi-equilibration, irreversibility), microkinetic models do not make a priori assumptions to simplify the chemistry. The process of building such a detailed mechanism is, however, time consuming and error-prone.
The open-source software RMG-Cat predicts elementary reactions in the gas phase and on the surface, including estimates for the thermodynamic properties of all the reactants, products, and intermediate species, and kinetic parameters for all the elementary reactions. The user specifies the starting species, the conditions of interest, and some tolerance settings which determine the final model size, and the software grows a model, following the pathways that are fastest at the specified conditions.
Developments under ECC
In this part of the work we enhance and extend our recently developed Reaction Mechanism Generator for Heterogeneous Catalysis, RMG-Cat, which builds microkinetic models for catalysis in an automated manner, and couple RMG-Cat to the other software tools and databases.
- Richard H. West (Northeastern)
- C. Franklin Goldsmith (Brown)
- Emily Mazeau (Northeastern)
- David Farina (Northeastern)
RMG-Cat is currently a fork of the open-source project RMG (Reaction Mechanism Generator). We plan to merge it back into the upstream master version of RMG, so that all RMG users can build models including heterogeneous catalysis and so that other improvements made to RMG will immediately benefit our work. But for now, RMG-Cat’s home is on the ‘cat’ branch of cfgoldsmith’s fork of RMG-Py, accessible on GitHub. You will also need the corresponding RMG-database from this GitHub site . You can also get a snapshot of the v.1.0 release here and here.