The catalytic reduction of NOx in lean burn vehicular exhaust is a critical technology for minimizing the effect of mobile sources of ground-level ozone. NOx storage and reduction (NSR) is an interesting variant on three-way catalysis, involves alternating NOx trapping of 1-2 minute duration followed by an intense reduction of several seconds duration. The rational design of the “lean NOx trap” requires a deep understanding of the catalytic chemistry and kinetics, and the coupling of the chemistry with the mass and heat transport processes. The objective of this study is to develop a predictive LNT model with detailed catalytic microkinetics spanning the storage and reductions steps.

The approach taken in this study is to build the microkinetic model for the various subsets of reaction chemistries. A complementary experimental study in our effort provided both the steady-state and cyclic operation using H2 as the model reductant and Pt/Ba/alumina as the washcoat. Particular attention is placed on the formation and reaction of ammonia, an important byproduct during NO reduction by H2. Data for the following chemistries was used to build the microkinetic subsets:

Reactants Products

• H2 + NO N2O, N2, NH3, H2O

• H2 + O2 H2O

• NO + O2 NO2

• NO + NH3 N2O, N2, H2O

• NH3 + O2 N2O, N2, NO, H2O

This involves defining a set of elementary or quasi-elementary steps guided by the measured product distribution and published mechanistic/kinetics studies. This introduces a number of kinetic parameters (pre-exponential and activation energies). Use of published kinetic parameters and of thermodynamic constraints reduces the number of parameters to be estimated. Some tuning of energy barriers and pre-exponential factors is necessary to capture the main and even subtle trends in the steady-state data.

The microkinetic model is combined with the short monolith flow model to simulate the conversions and selectivities from experiments. The predicted trends are in excellent qualitative and reasonable quantitative agreement with the steady-state data. Several key activity and selectivity trends during H2/NO, NH3/NO and NH3/O2 reactions are captured, including the inhibitory effect of H2 on NH3 reduction of NO.

The simulation of the cyclic operation requires the addition of a NOx storage model as well as reduction steps that occur at the interface of the Pt and Ba storage component. The modeling provides useful guidance to identify rate limiting steps and primary reaction pathways. In particular, the traveling fronts of H¬2 and NH3 during the regeneration step are simulated. These and other effects will be discussed.