High End Computing for the Characterisation of Structure and Reactivity of Complex Oxide Catalysts

We propose to continue the development and optimisation of computer codes for embedded cluster (mixed quantum/classical) calculations. The developments proposed target High End Computing platforms, and involve optimisations an algorithm changes needed to make full use of machines with hundreds or thousands of processors. The result will be codes capable of performing more accurate calculations of the structure and reactivity of condensed phase systems, such as the active surface sites of systems which perform heterogeneous catalysis. Such systems are often mixed metal oxide materials, often with additional transition metal atoms or clusters providing the reaction site.The proposed developments will allow high accuracy quantum methods to be applied to a larger part of the system, deal with larger and more complex reacting sites, and also to allow more advanced theoretical methods, for example moving from reliance on Density Functional Theory (DFT) to multiconfiguration methods (MCSCF), perturbation theory (MP2) and coupled cluster (CCSD(T)).We will test out the developments on a number of catalytic systems of industrial interest, such as(i) the surfaces of ZnO and TiO2 (catalysts for the production of hydrogen by ethanol steam reforming), (ii) the interface between MoS21AI203 interface - for hydrodesulphurisation of gas oils(iii) catalysts based on microporous siliceous and aluminium silicate materials (including zeolites) for partial oxidation, hydroxylation and epoxidation of chain and aromatic hydrocarbonsThe project will be suitable for a PhD student who is also expected to complete an MSc in High Performance Computing. This course wll be taught as a series of modules during the studentship, providing training in the underlying technology, e.g. MPI programming of parallel computers that is needed for the software developments.