1. Computational studies of adsorption, diffusion and reactions of atoms and molecules on metal oxide surfaces.

Focus of this project is on microscopic mechanisms of oxidation of carbon monoxide on Cu₂O surfaces and nanoclusters and oxidation of ammonia (NH₃) on RuO₂(110) surface. Our calculations result in adsorption energies, as well as, diffusion and reaction activation barriers. These quantities provide important information about reaction energetics and pathways and determine the rate limiting reaction steps. The valence charge density and densities of electronic states obtained from such calculations tells us about character of chemical bonding between reactants and surface at various stages of the reaction and help to rationalize the revealed catalytic effects. By introducing the  chemical potentials into the free energy of the system we can consider  the oxide surface in equilibrium with gaseous reactants and calculate from first principles the surface structure (composition and geometry) as a function of temperature and pressure of oxygen. Finally, a set of activation barriers for elementary catalytic processes obtained from first principles calculations will be used in kinetic Monte Carlo simulations which result in a consistent description of surface reaction including the temperature and pressure dependence of reaction rate.

2. Rationalization of the promotion and poison effects of co-adsorbed atoms on reactivity of metal catalyst surfaces.

One part of the project is the first principles study of the effect of adsorption of carbon and sulfur (poisons) or alkali metals (promoters) on the electronic structure of flat and stepped metal catalyst (Pd, Cu, Ni) surfaces. Another part is a computational study of the energetics of the adsorption, diffusion and reactions of CO and NH₃ on metal catalyst surfaces with and without co-adsorbed promoters or poisons. Analysis of the electronic structure of the systems in initial, transition, final and some intermediate states will help us to understand the mechanisms of influence of co-adsorbed poisoning or promoting atoms on reaction activation barriers.    

3. Adsorption and reactivity of NO on RuO₂(110) surface.

 We report here results of density functional theory (DFT) calculations in the generalized gradient approximation (GGA) using the Perdew-Wang functional for exchange-correlation energies on the energetics of adsorption, oxidation and dissociation of NO on stoichiometric RuO₂(110) surface and compare them to experimental data.  We find that NO adsorbs on top of the undercoordinated Ru with a tilted axis and the adsorption energy changes from 1.65 eV for half monolayer coverage to 1.43 eV for a full monolayer coverage. Unlike CO, NO oxidation was not observed in the surface and we find that it can be explained by the difference in their electronic structure, i.e, different locations of antinbonding states. For NO dissociation, the relevant calculated energy barrier is 2.3 eV and a stable final state is not found suggesting no dissociation on the surface.

 Some recent results:

• Based on first principles calculations, we find alkali adsorbates to induce a dramatic enhancement of the electronic polarizability of metal surfaces extending it several angstrom into the vacuum {Phys. Rev. Lett. Accepted for publication}. We show that this effect is responsible for the experimentally observed softening of vibrational modes of molecules co-adsorbed with alkalis on metal surfaces. Our results also suggest that the enhanced polarizability can be a driving force for well-known promotion effect in heterogeneous catalysis (increase of surface reactivity upon alkali adsorption).        

• Unidirectional adsorbate motion on a high-symmetry surface: ``Walking'' molecules can stay the course {Phys. Rev. Lett. 95, 166101 (2005) see also Physics Today 58, No.12, p. 9 (2005), Physics News Update (AIP) No751 #2 (2005). This is also in the list of the Top Physics Stories for 2005, Physics News Update (AIP) No757 (2005)}. STM experiment performed in  University of California, Riverside reveals unusual diffusive motion of an organic molecule along a straight line on the Cu(111) surface, and first principles calculations of potential energy surface for diffusion of the molecule explain the mechanism of this phenomenon.

• Adsorption, diffusion and dissociation of NH₃ on steps and terraces of Ni and Pd surfaces J. Chem. Phys. 123, 204716 (2005). Based on the first principles calculations we show that the activation barrier for NH₃ dissociation on metal surfaces is much lower for the molecule adsorbed on monoatomic steps than on terrace. In contrast to the Ni surface case, for Pd surface the activation barrier is found to be higher than disorption energy even if the molecule is adsorbed at the step. This explains the experimental finding, which shows that NH₃ decomposition rate is much higher on Ni than on Pd.

• As coverage of  carbon atoms adsorbed on the Ni(001) surface exceeds 1/3 of monolayer (ML), so-called “clock reconstruction” of the surface is observed in experiments. The results of our calculations of the geometric structures of the system with the coverage of 1/4 and 1/2 ML are in a very good quantitative agreement with the experimental findings. Furthermore, we obtain the reconstruction pathway and energetics. Analysis of the calculated valence charge density distribution in the system allows us to explain the mechanism of the reconstruction. {Phys. Rev. 72,155423 (2005)}); also selected for the October 31, 2005 issue of Virtual Journal of Nanoscale Science & Technology (AIP, APS) http://www.vjnano.org/.                    

• Effect of adsorption of C and S on geometric and electronic structure of stepped Pd surfaces –       poison reactivity {Phys. Rev. B 70, 155410 (2004)}; We have performed first principles calculation of C and S adsorption energies for various possible adsorption sites on Pd(211) surface to reveal the preferred adsorption geometries.   We find that hybridization between C p-states and neighboring Pd d-states leads to the formation of strong  covalent C-Pd bonds, while S-Pd bonding has predominantly ionic character. The depletion of the electronic density of states of the Pd surface atoms at the Fermi level suggests a poisoning effect of C on catalytic activity of the surface.

• Our calculations help to understand unusual reflection anisotropy spectra obtained for nanostructured copper surfaces by our colleagues experimentalists from Switzerland {Phys. Rev. Lett. 90, 177402 (2003)}.   

• Our collaborators, experimentalists from Germany, show that oxygen works as surfactant promoting 2D Ni film growth on the Cu(001) surface. During the growth O atoms are found to float up and stay on the film surface that is critical for a surfactant. Our calculations reproduce this effect and suggest a mechanism for that floating {Surf. Sci. 531, 53 (2003)}.   

• Studying so-called “missing row reconstruction” of the Cu(100) surface upon oxygen adsorption we have found that the long-range electrostatic interaction between the surface and adsorbates is the main driving force for this phenomenon.  {Phys. Rev. Lett. 89, 116101 (2002)} .