In this thread the development of a new model for the reactive scattering of molecular hydrogen from metal surfaces in which phonon effects are taken into account is presented.
When studying the reactive scattering of hydrogen from metal surfaces, most theory is based on using models in which the atoms of the metal surface do not move. The surface atoms are placed at their ideal lattice positions during these simulations. Despite previsously made arguments concerning the mass-mismatch between the hydrogen atoms and the metal surface atoms, say copper, the lack of a temperature dependance in the reactive scattering of hydrogen on coper in experimental data and the renowned normal-energy scaling law observed for the reactive scattering of hydrogen on copper, recent advanced and state-of-the-art six-dimensional quantum dynamics simulations show discrepancies when compared to the available experimental data. Experimentally accessible parameters like the rotational quadrupole allignment, the rotational in-elastic diffraction probabilities and the reaction probabilities calculated using these advanced dynamical methods on highly accurately calculated potential energy surfaces based on density functional theory still do not reproduce the observed qualitative behaviour in experiments well. It is currently unclear what causes the miss-match and it is often attibuted to two assumptions in the models; 1) neglect of phonon effects and 2) the neglect of electron-hole pair excitations.
Currently on this Leiden Classical BOINC project, the first steps are being undertaken to include surface atom motion effects into our models. To this end a new application is being developed based on previous work. The hydrogen on copper application already running on this project is being altered by a MSc student Mark Wijzenbroek. He will adapt the existing code to allow for simulations of the reactive scattering of hydrogen on a static copper surface but with the surface atoms being displaced from their ideal lattice positions. The surface atoms will be displaced according to the observed average displacement as a function of the temperature of the metal. Currently this change in the model does not allow for energy excange during the scattering event of hydrogen, but it does introduce extra corrugation and anisotropy into the dynamics. It is believed that this extra corrugation and anisotropy has a large effect on the calculated experimentally accesible parameters discussed above.
To this end the existing model potential describing the interactions of hydrogen with a copper surface is being modified and test calculations on the system with the surface atoms displaced are to be compared to simulations in which the atoms have not been displaced. If large differences are found, the static but corrugated surface (the 'wrinkled surface') effects are important and should be included in more sophisticated models and methods. To this end a large number of classical trajectory calculations are needed and BOINC will help us with that!
More news and information on this project will be made available in this thread.