Classical dynamics of gas-surface scatteringfundamentals and applications

  1. RODRIGUEZ FERNANDEZ, ALBERTO
Zuzendaria:
  1. Laurent Bonnet Zuzendaria
  2. Ricardo Díez Muiño Zuzendaria
  3. Joseba Iñaki Juaristi Oliden Zuzendaria

Defentsa unibertsitatea: Universidad del País Vasco - Euskal Herriko Unibertsitatea

Fecha de defensa: 2021(e)ko otsaila-(a)k 05

Epaimahaia:
  1. Peter Saalfrank Presidentea
  2. Maite Alducin Ochoa Idazkaria
  3. Cristina Díaz Blanco Kidea
  4. Thierry Stoecklin Kidea
  5. Geert-Jan Kroes Kidea
Saila:
  1. Polimero eta Material Aurreratuak: Fisika, Kimika eta Teknologia

Mota: Tesia

Teseo: 154009 DIALNET lock_openADDI editor

Laburpena

This thesis manuscript is devoted to the theoretical study of several reactive andnon-reactive processes that take place at the gas-solid interface. Two classical trajec-tory methods, different and complementary, were used to simulate the dynamics ofthese processes. The first one relies on large sets of classical paths obtained by nu-merically solving Hamilton equations on a previously constructed potential energysurface (PES). Classical paths are then assigned statistical weights based on twosemiclassical corrections: Gaussian binning and the adiabaticity correction. Thisapproach, in a quantum spirit, was applied to the scattering of H2 on a Pd(111) sur-face. First, the study focused on collisions where H2 is initially in the rovibrationalground state. Then, rotationally excited states were considered. On this occasion,a variation of the adiabaticity correction based on firmer semiclassical grounds wasintroduced. In both cases, the predictions of the sticking and state-resolved reflec-tion probabilities were found to be in remarkably good agreement with those ob-tained through exact quantum time-dependent calculations, contrary to standardquasi-classical trajectory predictions. The classical approach in a quantum spiritcould thus be very useful for future studies.The second method used in this work, known as Ab-Initio Molecular Dynamics(AIMD), calculates the inter-nuclear forces from density functional theory and usesthem to classically move the nuclei. Contrary to the previous approach, AIMD doesnot require the very demanding construction of a PES. The price to pay, however,is that the numerical cost of each trajectory is much higher than with the previ-ous method. AIMD allowed us to study the dissociation process of H2 on W(110)surfaces. The functional we use includes a van der Waals term which provokesan increase of the far distance attraction that is compensated by a stronger repul-sion at short distances. The combination of both effects appreciably decreases thevalue of the dissociation probability, bringing it closer to the experimental resultwhen a clean surface is used. When oxygen atoms are previously adsorbed onthe surface, the dissociation probability drops considerably. This effect increaseswith the amount of oxygen on the surface. An ordered phase of O adsorbates onthe W surface is used to explain the nonexistent sticking probability for coverages¿ > 0.35 ML observed experimentally. We show that the oxygen atoms push the H2molecules away from the narrow bottlenecks that open the paths to dissociation inthe absence of oxygen atoms. This effectively eliminates any chance of dissociationin the surface for high coverages. At lower coverages, our calculations demonstratethat the dissociation dynamics resemble those in the clean surface just in very spe-cific surface regions.