278518 Kinetic and Reactor Modeling of the NH3-SCR Converter: Fe- and Cu-Exchanged Zeolites As Single or Dual Layers
Kinetic and Reactor Modeling of the NH3-SCR Converter:
Fe- and Cu-exchanged Zeolites as Single or Dual Layers
Pranit S. Metkar, Michael P. Harold and Vemuri Balakotaiah
Department of Chemical & Biomolecular Engineering, University of Houston, Houston, TX-77204
The lean selective catalytic reduction of NOx is a critical challenge in diesel and lean burn gasoline vehicles. The development of predictive kinetic models is essential for identifying and optimizing improved catalysts. For example, we recently showed that the combination of Fe- and Cu-exchanged zeolites in the form of either sequential monoliths or dual layers can significantly increase the temperature window for which high NOx conversion is achieved . In the current study we describe the development of global kinetic models for NH3-SCR on both Fe- and Cu-exchanged zeolite monolith catalysts (Fe-ZSM-5 and Cu-SSZ-13 (chabazite)). To this end, the general SCR kinetic model accounts for the following reactions: NH3 adsorption, NH3 oxidation, NO oxidation, standard SCR (NO + NH3 + O2), fast SCR (NO + NH3 + NO2), NO2 SCR (NH3 + NO2), ammonium nitrate formation and its decomposition to N2O, N2O decomposition and N2O reduction by NH3. We incorporate the kinetic models into a monolith reactor model to simulate the dual-layer catalysts and in so doing help to interpret the performance of SCR monoliths over a wide range of conditions. The two-dimensional, two-phase reactor model includes the contribution of both external and internal mass transport processes so that conditions are determined when diffusion limitations are present.
Results to date are described as follows. The Cu-chabazite catalyst showed a higher NH3 storage capacity and higher activities for NH3 oxidation and standard SCR reactions compared to Fe-ZSM-5. The NOx reduction activity on the Fe-ZSM-5 catalyst was found to be strongly dependent on the feed NO2. For the Cu-chabazite catalyst, the NOx conversions were less sensitive to the feed fraction of NO2. In the presence of excess NO2, both N2O and ammonium nitrate byproducts are formed on both the catalysts. The Fe-ZSM-5 catalyst showed a higher selectivity towards these byproducts compared to Cu-chabazite. For different feed conditions (NO2/NOx = 0-1), Cu-chabazite was found to be more a more active NOx reduction catalyst at lower temperatures (≤ 350 oC) while Fe-ZSM-5 was more active at higher temperatures (≥ 400 oC). The model accurately predicts the steady state NOx/NH3 conversions and selectivity for different products formed during these reactions.
Finally, a systematic study of various SCR reactions was carried out on a combined system of Fe- and Cu-zeolite monolithic catalysts to determine if a high NOx conversion could be sustained over a wider temperature range than with individual Fe- and Cu-zeolite catalysts. Amongst various configurations of these combined catalyst systems, a dual-layer catalyst with a thin Fe-zeolite layer on top of a thick Cu-zeolite layer and a sequential arrangement of short Fe-ZSM-5 brick followed by longer Cu-chabazite brick resulted in achieving a very high NOx removal efficiency over a broad temperature range of practical interest. The model could successfully predict all the experimental data with combined Fe- and Cu-zeolite catalysts. A typical comparison of experimental and modeling results is shown in Figure 1. The model captures the main trends in the data and can be used to quantify the optimal Fe/Cu ratio to maximize the NOx conversion.
Figure 1: Steady state NOx conversions obtained during the standard SCR reaction studied on Cu-chabazite, Fe-ZSM-5, and combined system of series arrangements of Fe-ZSM-5 (in front) followed by Cu-chabazite. a) Experiments b) model predictions.
1. Metkar, P., V. Balakotaiah, and M.P. Harold, “Selective Catalytic Reduction of NOx on Combined Fe- and Cu-Zeolite Monolithic Catalysts: Sequential and Dual Layer Configurations,” Applied Catalysis B: Environmental, 111– 112, 67– 80 (2012).