(589d) Automated Kinetics Via Transition State Theory for Improved Mechanism Generation

An automated tool to calculate reaction rate parameters via classical transition state theory has been implemented in Reaction Mechanism Generator (RMG)[1]. The tool uses a group additive approach and distance geometry to derive good estimates of transition state geometries, and automatically optimizes the reactants and saddle point using DFT quantum calculations before calculating the kinetics.

During automatic mechanism generation, group additive methods have typically been used to estimate thermodynamic and kinetic parameters, with varying degrees of accuracy. The accuracy particularly suffers when there is little data from which to derive required group values, or when non-nearest-neighbor effects such as steric hindrance or ring strain become important. As kinetic group data are sparse, many of the used estimates contain several orders of magnitude error. Models in RMG are expanded to include reactions that meet a user-defined flux, so the large kinetic errors lead to unfavorable model expansion, and may lead to important pathways being excluded from the model if their rates are incorrectly underestimated. Our automated TST-based kinetic calculator should reduce these cases, improving models generated in RMG.

Kinetic calculations require reactant and transition state geometries, so we have developed a group additive approach to estimate distances between reacting atoms at the transition state. The distance estimates are used to provide good transition state geometry approximations, which are optimized using semi-empirical or DFT quantum chemistry, then validated via a reaction path analysis. Properties for the reactants and products are determined automatically in a similar manner[2]. Reactant and transition state partition functions are calculated from their molecular properties, while the tunneling correction is calculated via the Eckart model. Point groups are found[3] in order to assign molecular symmetries, allowing the kinetic parameters to be calculated via classical transition state theory.

These steps have now been integrated in RMG to allow automated kinetic calculations for reactions with activated complexes. This has been shown to improve the kinetics used in models generated by RMG, and the overall model generation process. 

References

[1]  Reaction Mechanism Generator - RMG-Py http://greengroup.github.io/RMG-Py

[2]  Magoon GR, Green WH Design and implementation of a next-generation software interface for on-the-fly quantum and force field calculations in automated reaction mech- anism generation. Comput Chem Eng.2012;52:35–45.

[3]  Patchkovskii S, 2003; http://www.cobalt.chem.ucalgary.ca/ps/ symmetry/