Importance of Heat Transfer During Carbon Monoxide Oxidation

Abstract

Interest in Carbon Monoxide Oxidation dates back to 1922 by Irving Langmuir himself. Since, a wide body of work has surfaced. Commercially available catalysts work in low humidity and high CO concentrations, such as Moleculite and Carulite at temperatures above 25 °C. Aurolite, a more recent gold based catalyst, is active all the way down to -40 °C. Above room temperature and 50% relative humidity, RH, there are few options. A Pt-CeO2 on SiO2 catalyst has been developed to fill the gap above 50% RH. Experimental results are obtained and modeled for a wide variety of working conditions with special attention paid to heat effects. Oxidation of CO is widely known to display negative order kinetics with respect to CO. It is also highly exothermic with an activation energy of 283 kJ/mol. This combination of factors amplifies sensitivity to temperature and concentration changes. It also muddies kinetic data with high CO concentrations or conversions. Literature typically operates around 1% CO which corresponds to a 100 °C adiabatic heat rise in a packed bed. Experiments were performed at low CO concentrations, below 2500 ppm, and will be presented and modeled using the Pt-CeO2 on SiO2 catalyst in excess oxygen. The large humidity range, from 0% to 90%, demands water adsorption on the SiO2 based packed bed also be accounted for. This has strong heat effects of its own, which are typically the domain of water/silica based adsorption chillers. The demands of the system necessitated a heterogeneous surface coverage based model which accurately models the oxidation of CO in a wide range of face velocities, CO concentrations, temperatures and humidity levels. The effect of axial conductivity, keff, on the reaction system is also shown to significantly alter reactor performance. keff is modified by using four different carrier gases: He, N2, Ar and Kr. This changed keff from .51 W/(m*K) to .091 W/(m*K). The results show an increase in the temperature at the front of the packed bed, and thus better CO conversion. The model is employed to offer insights into the reactor’s behavior. Finally, a comparison is made using Microfibrous media, MFM. The MFM holds the catalyst in a copper sinter locked mesh which allows for excellent heat conduction. The MFM entrapped catalyst has a keff of .54 W/(m*K) in flowing N2 and exhibits an increase in catalytic performance and longevity relative to a packed bed.