Refine
Document Type
Language
- English (5)
Keywords
- catalytic oxidation (1)
- pattern (1)
- stochastic (1)
Institute
The rate of CO oxidation on Ir(111) surfaces exhibits bistability at T = 500 K in a range of the CO fraction Y in the CO + O reactant gas flux. Quadrupol mass spectrometer measurements of the CO2 rates as a function of the noise strength imposed on Y are well reproduced by parameter-free modeling. We present Photoelectron emission microscopy (PEEM) measurements and 2D calculations of the spatio-temporal patterns of CO and O rich domains. The role of combined multiplicative and additive noise on Y for CO and O domain wall motion and island nucleation-growth-coalescence processes is analysed. For small noise amplitudes few islands nucleate and grow up to some 100 µ m in diameter, before they merge with another island. With increasing noise amplitudes more islands nucleate. With large noise amplitudes bursts to and switching between the branches is observed. The domain wall velocity is found to be independent of the noise strength and island size.
The CO oxidation reaction on Iridium(111) surfaces shows bistability in a limited range of the CO fraction of the reactant gas flux Y and a wide range of temperatures T. The two branches are characterized by their reactivity for CO2 formation. The upper rate (high CO2 formation) rate is related to high oxygen coverage on the surface, the lower rate (little CO2 formation) to high CO coverage. Quadrupol mass spectroscopy and PEEM (photoelectron emission microscopy) was employed to study the influence of a noisy reactant gas flux composition on the spatio-temporal pattern development in the CO oxidation reaction on flat Ir(111) and stepped Ir(977) surfaces. PEEM shows nucleation and growth of few oxygen resp. CO islands at small noise amplitudes. Anisotropic diffusion of CO parallel and normal to the steps causes elliptic shapes of large islands. The long axes of the ellipses are aligned along the steps. At increased noise amplitudes the density of islands becomes larger. 2D modeling of the phenomena based on reaction-diffusion differential equations reproduces the experimental findings quite nicely.