In-Situ Probe Technique Applied to Partial Oxidation of Methane and Ethanol On Rh/Al2O3 Coated Monoliths

Catalytic partial oxidation (CPOX) of hydrocarbon fuels to hydrogen consists of a complex coupling of heterogeneous and homogeneous reactions as well as mass and heat transport phenomena. A profound understanding of these processes is crucial for the design and optimization of CPOX reformers. To gain a deeper insight, in-situ sampling techniques are employed [1-3], which yield spatially-resolved concentration and temperature profiles. However, the influence of the probe on the measured data has to be evaluated [4]. Herein, we present experimentally-obtained axial profiles for CPOX of ethanol on a Rh/g-Al₂O₃ coated honeycomb catalyst for various C/O ratios. In addition, different channels of the monolith were examined for CPOX of ethanol and methane. The influence of the probe on the measured data was evaluated by CFD simulations for CPOX of methane.

For the experiments, a monolithic honeycomb (600 channels per square inch (cpsi), Ø 19 mm, L = 10 mm) coated with Rh on a g-Al₂O₃ washcoat was used. Uncoated monoliths (600 cpsi, Ø = 19 mm, L = 10 mm) were placed as heat shields in front of (front heat shield, FHS) and behind the catalyst (back heat shield, BHS), with a gap of 5 mm between the FHS and the catalyst. With a capillary (outer diameter = 170 mm), a constant gas sample was sucked from the channel at the tip of the capillary. The gas composition was analyzed by means of FT-IR and MS. The temperature profiles were obtained by a thermocouple (gas phase temperature) and an optical fiber connected to a pyrometer (surface temperature).

The computational fluid dynamics (CFD) code FLUENT was applied for the calculations. The computational domain (Figure 1, right) included nine channels of the monolith. Nine different positions of the probe tip inside the catalytic channel were examined. To solve the residence time of the fluid t downstream z = 0, an additional transport equation was implemented into the code. The solid was not included in the simulations; the measured wall temperature was used as a boundary condition.

Typical profiles along a catalytic channel for the concentrations of the reactants and main products are shown for CPOX of ethanol (C/O = 0.75) in Figure 2 (lines with symbols). Two different catalytic zones are observed in the catalytic channel: an oxy-reforming zone with total oxidation and reforming reactions as dominating reactions, and a reforming zone, where all oxygen has been consumed, with only reforming reactions as prevalent reactions. Water (Figure 2) is formed in total oxidation at the catalyst entrance and is partially consumed along the catalytic channel in steam reforming (SR). The surface temperature (Figure 2, solid line) reflects the dominating reactions with a hot spot on the first millimeter inside the channel caused by exothermic total oxidation and a following decrease in temperature due to endothermic reforming reactions, e.g. SR. For ethanol CPOX, pre-catalytic gas-phase reactions, mainly oxidative dehydrogenation of ethanol to water and acetaldehyde (not shown), are observed. The gas-phase temperature (Figure 2, dashed line) increases in front of the catalyst due to heat radiation from the front of the catalyst, which favors the gas-phase reactions.

 Figure 1 shows the used computational domain (right) and the calculated distribution of H₂ (left, upper half) and the residence time (left, lower half) for CPOX of methane in a yz-plane starting upstream of the catalyst and ending at 6 mm inside the catalytic channel. Comparing the residence time in a reference channel without the probe (channel_ref) with the channel with the probe (channel_probe), a longer residence time is observed for channel_probe. Also, a higher hydrogen concentration is observed in the channel_probe than in the channel_ref.

A comparison of the experimentally measured profiles for methane CPOX (symbols) with calculations for a channel with capillary (channel_probe, dashed lines) and a channel without capillary (channel_ref, solid lines) is shown in Figure 3.

A deviation between channel_ref and channel_probe is clearly observable. This is due to the boundary layer, which forms at the outer wall of the capillary. The suction rate does not influence the simulated results, as the sampled gas volume is small compared to the volume flux in the channel. The influence of the probe is dependent on the position of the probe tip in the catalytic channel. The further downstream the probe tip is positioned, the bigger the deviation between channel_probe and channel_ref. The measured data are more similar to the calculated values for a channel with capillary than for one without_.

By utilizing the in-situ sampling technique, a deeper insight into the reaction network is gained. The spatially-resolved concentration and temperature profiles reveal the dominant reaction pathways and pre-catalytic reactions like oxidative dehydrogenation in the case of ethanol. However, the influence of the probe on the measured data has to be taken into account. 3D simulations are necessary to evaluate the influence of the probe.

[1]          R. Horn, K.A. Williams, N.J. Degenstein and L.D. Schmidt, Journal of Catalysis, 242 (2006) 92.

[2]          A. Donazzi, D. Livio, M. Maestri, A. Beretta, G. Groppi, E. Tronconi and P. Forzatti, Angewandte Chemie International Edition, 50 (2011) 3943.

[3]          D. Livio, C. Diehm, A. Donazzi, A. Beretta and O. Deutschmann, in preparation.

[4]          M. Hettel, C. Diehm, B. Torkashvand and O. Deutschmann, Catalysis Today, submitted.