First principles ammonia synthesis and decomposition over a Ru-based catalyst
Honkala, Karoliina
Finland

A complete characterization of a catalytic reaction, based solely on first principles, has been a long-lasting dream in the surface science and catalysis communities. Using the NH3 synthesis and decomposition as an example, we have shown that the reaction rate over a nanoparticle Ru catalyst under industrial conditions can be calculated directly using DFT calculations together with the kinetic model, and that the results are in good agreement with rate measurements performed over a wide range of industrially relevant conditions.
Our starting point for calculating ammonia synthesis process is that N2 dissociation is the rate-limiting step and that we only need to consider dissociation along the step sites where the special B5 sites exist. The rate of the rate-limiting step will in general depend on the local environment of the step site where N2 is dissociating. We therefore calculate the total rate as the sum of the contributions of all different configurations weighted by their statistical probability depending also on the interactions between adsorbates.
The first principles model for NH3 synthesis on an unpromoted Ru-based catalyst consists of energetics calculated within the DFT framework, coverages and conditional probabilities obtained from Monte Carlo (MC) simulations, and macroscopic properties, such as partial pressures and flows provided by a kinetic model. The basic shape and number of active sites for the nano-sized Ru particles were determined using DFT and an atomistic Wulff construction. The only link between the catalyst material and the model is the size distribution of the Ru particles, which was determined from transmission electron microscopy. All trends in activity are captured and the absolute value of the ammonia synthesis/decomposition productivity is predicted to within a factor of 1-100 depending on the conditions. The good agreement between theory and experiment can be associated with the fact that the total rate is less sensitive to systematic inaccuracies in the calculated energies than the rate of the individual elementary reaction steps. This offers hope for computer-based methods in the search for new catalysts.
back