Support and promoter effects in automotive exhaust catalysis

Abstract

Automotive catalysis being a mature technology, it can only be improved by the introduction of new breakthroughs. The ideas generating these technical advances in material science can be found thanks to the synthesis and study of model systems with controlled geometries, compositions, interactions and dimensions in well-known environments. This approach limits the amount of possible parameters influencing the experimental results and facilitates their interpretation. The aim of this PhD thesis was to enlarge insight into the fundamental issue of metal-support interactions for Pt- and Rh-based exhaust gas emission catalysts. This has been done by synthesising model catalysts made of well-defined Pt and Rh nanoparticles dispersed on various oxide supports, differing in their chemical composition and pore architecture. The composition effect was mainly examined with a series of zeolite Y supports containing different alkali or alkaline earth metal ions and further confirmed with promoted macroporous SiO2. The impact of the support porosity was investigated thanks to all silica supports varying by their pore size. The electronic properties of the supported Pt and Rh nanoparticles were then studied by Infrared (IR) spectroscopy in combination with CO or NO as probe molecule and, for Pt, further confirmed by X-ray Absorption Fine Structure spectroscopy (XAFS), a direct measurement technique of the electron density. Finally, the impact of the support composition on the catalytic activity of Pt and Rh was evaluated for two simple reactions relevant to the automotive exhaust gas converters: the CO oxidation in excess of oxygen and the NO reduction by CO at stoichiometry. Combining all results, it has been possible to derive relationships between the chemical composition of the support material, the electronic density of the supported metal nanoparticles and their catalytic activity. For all catalyst materials, the IR spectra showed clear differences, the probe molecules were adsorbed on the Pt or Rh surface in linear or bridge positions. It was found that the relative intensity of the bridge bonded probe molecule as referred to the total absorbance of the metal-bonded CO or NO molecules was decreasing for increasing Lewis acidity of the promoting cation, as expressed by the Kamlet-Taft parameter ?. As the CO IR linear-to-bridge band intensity ratio increased, so did the AXAFS peak intensity, while the promoting cation was more acidic and the Pt particles became electron poor. This confirmed that AXAFS is a novel spectroscopic method to directly probe the electronic properties of supported Pt nanoparticles. In general, the pore size effect was small while the support composition had a big impact. Finally, it was possible to correlate the above observations with the activity of the Pt or Rh catalysts for the CO oxidation reaction in excess of oxygen, i.e. the T50% for CO oxidation decreases with electron density on the precious metal nanoparticles. Concerning the stoichiometric NO reduction by CO on Rh, no clear trend could be extracted, probably due to other important factors affecting the Rh catalytic activity, such as its oxidation state and its particle size.