Investigation of thermal effects on heterogeneous exothermic reactions and their impact on kinetics studies

• Published in: Chemical engineering journal (Lausanne, Switzerland : 1996) Vol. 377; p. 120179
• Main Authors: Biasin, A., Fabro, J., Michelon, N., Glisenti, A., Canu, P.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01-12-2019
• Subjects: CO oxidation; Dynamic modelling; Monolith catalyst; Thermal hysteresis; Transient experiments; Monolith catalyst; Dynamic modelling; CO oxidation; Transient experiments; Thermal hysteresis
• ISSN: 1385-8947; 1873-3212
• DOI: 10.1016/j.cej.2018.10.116

•TPO may easily induce a thermal wave in the catalyst.•Dispersions in small-scale monolith diverge significantly from adiabatic assumption.•Hysteresis changes when the position of the thermocouple that controls heating is moved.•Hysteresis in X(T) appears very different if IN, OUT or internal T is considered.
Experiments and modelling were performed to investigate CO oxidation over a Pd-Rh monolith. We focused on thermal effects and hysteresis, to validate by modelling a thermal explanation of the results. Different feed composition (0.07–4% vol. CO) and heating rates (0.5–5 °C/min) have been used to reproduce both ignition and extinction stages, up to 300 °C, thus measuring the catalyst activity under transient conditions.
The heating rate plays a marginal role in producing hysteresis, whereas the reactants concentration appears the real cause, because of its effect on the rate of heat production. A significant increase of the monolith temperature compared to the inlet gas is measured after ignition. The local overheating of the catalyst surface explains the hysteresis observed. When the reactor thermal control is based on the internal temperature, instead of the inlet one, the hysteresis appears dramatically different. The choice of the temperature used to control the oven, and report the activity results, may induce very misleading indications, including inexistent multiple steady-states.
A model accounting for the thermal dynamics of the solid predicts the observed hysteresis, even with simple rate equations. That supports the thermal explanation for the direct hysteresis. The estimated activation energy is quite reasonable and compare well with literature. Preexponential factors accommodate for the weakening of the adiabatic channel critical assumption, less and less realistic as the reaction heat increases. The overall conclusion is that the local temperature can vary widely, in time and space, and any kinetic study not accounting for a precise knowledge of that will inevitably produce poorly representative parameter estimates. The limitation can be overcome with spatially-resolved measurements.