Estructuras disipativas en catálisis heterogénea: estudios experimentales y simulaciones para la oxidación de amoníaco sobre superficies monocristalinas de Pt

Abstract

Various heterogeneous catalytic reactions, when performed on a flux reactor (i.e. on open systems) under UHV conditions display complex spatiotemporal patterns. Specifically, TPR, Work Function (WF) measurements (both global via Kelvin probe and local via Photoemission Electron Microscopy-PEEM), and low energy electron di raction (LEED) experiments during the NO + NH3 reaction on Pt(100) surface show oscillations in reaction rates, surface structure of the catalyst and spatio temporal patterns (named "dissipative structures" after I. Prigogine), if pressures are maintained in the range of 1 × 10−6 mbar, and temperatures between 300 − 750K. In order to explain these results, the starting point should be the answer of some important questions, like the following. Which are the elementary steps in the reaction mechanism? Which one of these steps are the relevant ones? Which role plays the surface in the global process? Within the answers we can find lies the initial steps to build models that explain from first principles the observed dynamical behavior. Experimental methods in surface science allow the examination of significantly different spatial scales, ranging from the macroscopic limit to nearly atomic resolution. Bearing this in mind, we should consider that there is no unique tool (both from experiments and modeling) capable of a complete description of the observed phenomena. Therefore, we will have to formulate and discuss models to simulate behavior in different spatial scales. We begin with mesoscopic lengths (≈ μm), then we move to macroscopic scales (≈ 1×102μm), and finally we go down to microscopic lengths (≈ 1 × 10−2μm). We used for the mesoscopic scale a realistic mechanism of the reaction of NO + NH3 on Pt(100), the Lombardo, Fink and Imbihl (LFI) mechanism. Within the "mean-field" approximation we reproduce with our simulations the temporal evolution of intermediate species coverages, and succesfully eliminate adiabatically a subset of the total variables used to describe the system (we reproduced the dynamical behaviour using only three from the seven original variables). The following step was to move to the macroscopic range, and investigate the spatial distribution of adsorbates on the surface. We used for this end the information gathered from the "mean-field" simulations and coupled the cells defined in the microscopic approximation via diffusion, both Fickean and non Fickean (i.e. we considered lateral interactions between adsorbates). We successfully reproduced some features observed in PEEM experiments concerning the spatiotemporal evolution of the surface as temperature was varied (e.g., reaction fronts and homogeneous phase transitions). Finally, to cover also the microscopic scale, we used discrete models where every adsorbate is explicitly identified on a surface adsorption site, and we perform Monte Carlo simulations of the spatiotemporal evolution of the system through Markovian processes.We test the hypothesis considered in our Monte Carlo reduced model by using realistic values for the energies of surface processes involved. Main results obtained were the finding of kinetic oscillations in some parameter window, and transitions from coupled (homogeneous oscillations) to non-coupled (non-homogeneous oscillations) regimes as temperature was increased in a constant reactants pressure atmosphere. In addition, we have also found poisoned regimes and characterized surface evolution in the parameter space. Regarding the experimental part of this thesis work, we analized a reaction system related to the one used in the simulations above detailed, NH3 + O2 over Pt(100). Even though replacing NO with O2 as oxidizing agent is a significant change, there are a number of similarities maintained, e.g., Oads on the surface, intermediate NH3 dissociation species, and adsorbed NO (which we can speculate to be involved in some kind of surface restructuring process). In particular, we study the fefect of different coverages of preabsorbed oxygen regarding ammonia dissociation on the surface, the activity of Pt surface (especially the possibility of surface restructuring caused by adsorbates); and spatiotemporal evolution during reaction. We characterized the behavior of this reaction system, but one of the most stricking results obtained was the observation, for the first time to our knowledge, of surface phase transitions in NH3 + O2 over Pt(100).