Novel Acceleration Methods and Improved Transition State Finding Approaches for the Automatic Exploration of Reaction Networks

Auteurs

• Lukas Krep

Paramètres

• Physique
• Sciences naturelles

En savoir plus sur le livre

Reaction models are crucial for engineering tasks like optimizing internal combustion engines and chemical production lines. ChemTraYzer, an automated reaction space exploration method, aims to streamline the often lengthy development of reaction models. It employs reactive Molecular Dynamics (rMD) simulations to uncover the fundamental reactions involved. By linking rMD outputs to quantum-mechanical reoptimizations of reactant, product, and transition state (TS) geometries, as well as transition state theory, ChemTraYzer generates kinetic and thermochemical data with low uncertainties, essential for various applications. This thesis addresses two key challenges of the ChemTraYzer method. First, it introduces two acceleration techniques for rMD simulations: pressure-accelerated Dynamics (pAD) and ChemTraYzer-Temperature-Accelerated Dynamics (ChemTraYzer-TAD). These techniques extend rMD's applicability to longer reaction processes while requiring minimal prior knowledge of the reaction systems. The pAD method utilizes elevated pressure and temperature to hasten reaction events without biasing the reaction network. ChemTraYzer-TAD builds on this, enabling higher simulation temperatures and achieving boost factors of 108. Both methods are tested on low-temperature ignition reaction processes. Second, the thesis enhances the automatic TS search performance by introducing a recovery method for failed searches, improving the succes

Achat du livre

• Année de publication: 2023