Evaluating kinetic models for preferential CO-oxidation catalysts using optimization-based parameter estimation

We adapt a general-purpose optimization-based parameter estimation technique previously described in the literature [1] to evaluate the suitability of a number of common kinetic models for the representation of key performance characteristics (conversion and selectivity) of catalysts used for the preferential oxidation of CO in the presence of H2. We find that, for process engineering applications, there is no clear practical advantage to using mechanistically based kinetic models (e.g. Langmuir-Hinshelwood) unless the precise chemical mechanism is known. Empirical rate models are found generally to provide equivalent or better simulations of key performance variables for a diverse group of catalyst formulations. Furthermore, we demonstrate that the water-gas-shift (WGS) reaction is relevant within PROX reaction systems under conditons containing high fractions of CO2 and H2, confirming the expectations of Choi and Stenger (2004) [2]. Finally, we attempt to identify any emergent trends in kinetic parameters among catalysts sharing similar active metal or metal oxide components. Unfortunately, apart from confirming that the activation barrier for CO oxidation is generally less than the barrier for H2 oxidation (an expected relationship for PROX catalysts), no such trends are found.