Dry Reforming of Methane with a Ni/Al2O3‐YSZ Catalyst: The Role of the Catalyst Preparation Protocol

Abstract

AbstractMany studies of Ni based ceramic supporting reforming catalysts are found in the literature. A synthesis of the reported results shows that their efficiency and durability are significantly affected by their fabrication protocol. This research has been aimed at evaluating how the conditions of 1) the ceramic support preparation and 2) the Ni deposition, through an impregnation‐drying‐calcination‐reduction protocol, affect the catalytic activity and the catalyst deactivation over time during methane dry reforming. The catalyst support used in this study was obtained by the mixing and pressing of alumina and YSZ (Yttria Stabilized Zirconia) powders, then calcining the mixtures at high temperature to form pellets of limited porosity (specific surface of 1.5‐10 m2/g), without inducing change to the crystalline phases. The results show that the surface density of the nickel particles, the catalyst activity, and its life span are highly dependent upon the catalyst preparation protocol. The initial nitrate solution concentration, the duration of the impregnation and the specific surface of the ceramic support have, all of them, a considerable influence on the size range of the deposited nickel particles. The surface density, the amount and the size of the latter highly affect the catalytic activity. It has been also shown that an increase in the ratio CH4/CO2 is detrimental to the catalytic activity of the tested formulations; a small excess of methane is enough to initiate the deactivation process of the catalyst very quickly for all of the composition tested in this study. A phenomenological deactivation kinetics model has been built and optimized. Although there are differences in deactivation rates among the different formulations tested, the model shows that the deactivation rate is highly dependent upon the reaction rate constant and that zero‐ and first‐order kinetics give statistically the same prediction error; the latter is always lower or equal to the experimental error.