A computational study of the stability and growth of Au clusters on MgO.

Supported gold nanoparticles are used for a wide range of catalytic processes. For example, Au/MgO catalysts are used in the low temperature oxidation of CO and the oxidation of alcohols. In this work, the most stable Au clusters of sizes up to 19 atoms on MgO(001) are determined (Aun/MgO, n < 20). The Au clusters are created through the employment of two methods using DFT to minimise the electronic structure: i) a systematic build-up method and ii) unbiased global optimisation methods, i.e. genetic algorithm and basin hopping Monte Carlo. Optimisations are performed using dispersion corrected density functional theory (DFT-D). Trends relating to the structural and electronic properties that are associated with an increase in cluster size, are also evaluated and discussed concluding that as the Au clusters increase in size, they transition from perpendicular planar structures to parallel planar structures with an Au (111) interface configuration with the MgO support. Searches for stable clusters performed with a Monte Carlo global optimisation procedure used empirical potentials to determine the energy. The resulting structures were subsequently optimised using DFT to determine the validity of the method for the Aun/MgO (n < 20) system, and comparisons of the energetic and structural parameters of the clusters made to the findings of the previous search methods employing DFT. However, this methodology was not successful for determining the most stable clusters for the system (where n < 20). Once the most stable Aun/MgO systems were determined, the structures were employed in a study of the Ostwald ripening growth process. We have employed DFT calculations to calculate the barriers associated with ripening, combined with microkinetic simulations to investigate two mechanistic processes, ripening and digestion, relating to Au clusters size changes on a MgO(001) support.