Theoretical frameworks for the development of surface reaction mechanisms

• Main Author: Kraus, Peter
• Other Authors: Lindstedt, Peter ; Beyrau, Frank
• Published: Imperial College London 2016
• Subjects: 621
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.739596

A class-based framework for generation of heterogeneous reaction mechanisms has been proposed. The framework consists of a transition state theory method for estimating adsorption and desorption rate parameters, two-dimensional collision theory for homogeneous surface reactions and the unity bond index–quadratic exponential potential (UBI–QEP) for estimation of barrier heights. The framework has been developed to address the reliance of customary approaches on experimental sticking coefficients, and provide a fully self-consistent method for mechanism development on novel metals. Two different implementations of transition state theory have been considered, with the variational approach (VTST) showing improved quantitative results and higher robustness at negligible additional computational expense. Mechanisms prepared using this framework have been validated against a wide range of experimental data. On platinum, the VTST-based mechanism was applied for hydrogen, syngas, ethane and methane combustion. A new mechanism has been also developed for methane combustion on rhodium. The mechanisms have been applied under fuel-lean and fuel-rich conditions, various geometries, pressures and inlet velocities. The performance is generally better or as good as the optimised collision theory based determinations. The compatibility of the framework with high-accuracy density functional theory (DFT) data has been established using a case study of ethane adsorption on oxygen-covered platinum. At temperatures below 1500 K, the DFT study predicted considerably slower rate constants than the VTST approach. However, for the studied cases, the mol fraction profiles calculated with the VTST based rate determinations were in acceptable agreement with the experimental data. The developed method reproduces experimental data without the reliance on sticking coefficients, and facilitates the efficient generation of novel heterogeneous reaction mechanisms.