## Final Activity Report Summary - RDPOLSYSHETCAT (Reaction Dynamics of POLyatomic SYStems and HETerogeneous CATalysis)

The RDPOLSYSHETCAT project corresponded to an ongoing research project by Dr Fermín Huarte Larrañaga. The overall goal of the project was the theoretical simulation of the reaction dynamics of two families of processes with high technological interest, namely the gas phase reactivity of polyatomic systems as a way towards the understanding of simple organic reactions and the reactivity in heterogeneous processes, in particular heterogeneous catalysis. This was a long term running project which started thanks to the Marie Curie action and the obtained results during this first year were expected to be extended in the near future.

Concerning the study of polyatomic gas phase reactivity, the theoretical simulation of the H + C2H2 greater than C2H + H2 reaction was started. Apart from the importance of the reaction, which was involved in the combustion of natural gases, this work represented a challenge in itself. Nowadays, few reactions involving more than four atoms have been accurately studied. The first steps towards the simulation of this reaction were taken, such as the study of the topology of the Potential energy surface (PES) and its stationary points. An adequate topological study of the PES was crucial for our work since it allowed us to define the dividing surface which best separated reactants from products. The rather particularly odd topology of the PES which was employed up to now allowed us to choose a dividing surface which could be used for the reaction rate calculation.

Most of our effort during the period of contract duration was devoted to the development of a method for the calculation of initial state-selected reaction probabilities employing flux correlation functions and the MCTDH scheme. Nowadays, it is fairly straightforward to calculate, for a wide range of triatomic reactions, the probability for a selected initial reactant state to react and evolve into a selected final product state. However, whenever the system contains a larger number of atoms or in case these atoms are heavy, such a detailed calculation can no longer be obtained. In this case, reaction probabilities selecting only the initial reactant state turn up as a useful tool. Most of the methods applied up to date obtain the reaction probability by constructing a wave packet corresponding to the initial state in the reactant asymptotic region and propagating this wave packet, across the transition state region, until the product asymptotic region.

Our scheme, which was developed within the RDPOLSYSHETCAT project, adopted a different perspective for the calculation. The wave packet was generated on top of the reaction barrier and did not correspond to a particular initial state but was an ensemble of flux eigenstates. We believed that this type of scheme could be very useful for the calculation of the reactivity of a set of initial states. The method was validated studying the H + H2 reaction. As it was described in the project proposal, the development of this method was also of key importance for the study of reactions in the heterogeneous phase, which was the other main goal of our project.

Concerning the simulation of processes in the heterogeneous phase, the previously described method was applied to the dissociative adsorption of methane (CH4) on a nickel (Ni) surface. In order to do this, a three-dimensional model for the CH4-Ni system was used as a start-up basis, where the metal surface was simplified as a single Ni atom and CH4 as a H-X pseudo-diatomic molecule. The Hamiltonian corresponding to this reactive system was developed and, by the time of the project completion, work was in progress to converge the numerical parameters of the simulation.

As detailed in the management section of the project report, two graduate students participated in the project, Mr Marc Moix and Mr Xevi Biarnés. Both students actively worked in the development and application of dynamics' applications. Mr Moix worked in the first objective of our project developing a wave packet propagation code for the study of simple gas phase triatomic reactions. The task performed by Mr Biarnés focussed on the other project objective, carrying out molecular dynamics' simulations for the physisorption of molecular hydrogen on carbon nanostructures.

Concerning the study of polyatomic gas phase reactivity, the theoretical simulation of the H + C2H2 greater than C2H + H2 reaction was started. Apart from the importance of the reaction, which was involved in the combustion of natural gases, this work represented a challenge in itself. Nowadays, few reactions involving more than four atoms have been accurately studied. The first steps towards the simulation of this reaction were taken, such as the study of the topology of the Potential energy surface (PES) and its stationary points. An adequate topological study of the PES was crucial for our work since it allowed us to define the dividing surface which best separated reactants from products. The rather particularly odd topology of the PES which was employed up to now allowed us to choose a dividing surface which could be used for the reaction rate calculation.

Most of our effort during the period of contract duration was devoted to the development of a method for the calculation of initial state-selected reaction probabilities employing flux correlation functions and the MCTDH scheme. Nowadays, it is fairly straightforward to calculate, for a wide range of triatomic reactions, the probability for a selected initial reactant state to react and evolve into a selected final product state. However, whenever the system contains a larger number of atoms or in case these atoms are heavy, such a detailed calculation can no longer be obtained. In this case, reaction probabilities selecting only the initial reactant state turn up as a useful tool. Most of the methods applied up to date obtain the reaction probability by constructing a wave packet corresponding to the initial state in the reactant asymptotic region and propagating this wave packet, across the transition state region, until the product asymptotic region.

Our scheme, which was developed within the RDPOLSYSHETCAT project, adopted a different perspective for the calculation. The wave packet was generated on top of the reaction barrier and did not correspond to a particular initial state but was an ensemble of flux eigenstates. We believed that this type of scheme could be very useful for the calculation of the reactivity of a set of initial states. The method was validated studying the H + H2 reaction. As it was described in the project proposal, the development of this method was also of key importance for the study of reactions in the heterogeneous phase, which was the other main goal of our project.

Concerning the simulation of processes in the heterogeneous phase, the previously described method was applied to the dissociative adsorption of methane (CH4) on a nickel (Ni) surface. In order to do this, a three-dimensional model for the CH4-Ni system was used as a start-up basis, where the metal surface was simplified as a single Ni atom and CH4 as a H-X pseudo-diatomic molecule. The Hamiltonian corresponding to this reactive system was developed and, by the time of the project completion, work was in progress to converge the numerical parameters of the simulation.

As detailed in the management section of the project report, two graduate students participated in the project, Mr Marc Moix and Mr Xevi Biarnés. Both students actively worked in the development and application of dynamics' applications. Mr Moix worked in the first objective of our project developing a wave packet propagation code for the study of simple gas phase triatomic reactions. The task performed by Mr Biarnés focussed on the other project objective, carrying out molecular dynamics' simulations for the physisorption of molecular hydrogen on carbon nanostructures.