Dynamics of the photo-induced desorption and oxidation of CO on Ru(0001) with different (O, CO) coverages

  1. TETENOIRE, AUGUSTE LOUIS JEAN MARTIN CAMILLE ETIENNE GERMAIN
Dirigida por:
  1. Joseba Iñaki Juaristi Oliden Director/a
  2. Maite Alducin Ochoa Director/a

Universidad de defensa: Universidad del País Vasco - Euskal Herriko Unibertsitatea

Fecha de defensa: 23 de marzo de 2023

Departamento:
  1. Polímeros y Materiales Avanzados: Física, Química y Teconología

Tipo: Tesis

Teseo: 805457 DIALNET lock_openADDI editor

Resumen

Carbon monoxide (CO) is a neurotoxic gas emitted for instance in combustion reaction. Therefore it hasbeen sought for air treatment solution, where CO oxidation is a straight forward choice. In ultra highvacuum conditions the ruthenium has been found to be very inactive for CO oxidation. Experimentally ithas been shown the opening of a new reaction path for CO oxidation on ruthenium surfaces by means offemtosecond laser irradiation. Accurate simulations of the photo-reaction dynamics are required to give aproper characterization of this kind of experiments. This thesis is dedicated to the study of the photoinduceddesorption and oxidation of CO molecules, coadsorbed with oxygen (O) adatoms on Ru(0001)with different surface coverages. We began with the characterization of three (O, CO) mixed surfacecoverages on Ru(0001). We first found the adsorption configuration of minimum energy for each surfacecoverage, then we computed the desorption potential of a CO molecule, and found the minimum energypath to CO oxidation on all three surface coverages. Then we ran ab-initio molecular dynamics withelectronic friction simulations, and we have been able to show the complexity of the reaction path tooxidize the CO molecule, and explain its low probability of occurrence. Next, we showed the importanceof surface deformations on the desorption and oxidation probabilities of CO, and on the adsorbatemotion. Then, we have shown in detail and characterized the different mechanisms of CO oxidation.Finally we created a potential energy surface based on neural networks and showed that it is a verypromising tool to solve the problem of the computational cost of ab-initio molecular dynamicssimulations.