The NOx storage-reduction (NSR) process, first commercialized for diesel applications by Toyota, is a promising way to reduce nitrogen oxides (i.e. NO and NO₂) emission from diesel and lean burn gasoline engines. The NSR process is operated alternately under lean and rich conditions, where NOx is stored on the catalyst under lean conditions and subsequently converted to nitrogen under rich conditions. The goal of this research is to develop a kinetic model that can help understand the NOx storage mechanism on NSR catalytic systems and predict their performance.

Experiments for NO oxidation and NOx storage were carried out in a plug flow reactor (PFR) with a Pt/BaO/Al₂O₃ monolith catalyst. This plug flow, monolithic reactor was modeled by using a transient, 1-dimensional, two phase approximation. External and internal diffusion within the washcoat were also incorporated into the model. The kinetic part of the model includes the rate expression for NO oxidation on Pt sites and the irreversible adsorption of NO₂ and NO on BaO and Al₂O₃ sites, respectively. The resulting differential equations were solved by the finite element method using FEMLAB^®. Model predictions of the outlet gas concentrations were fit to experimental data by coupling the FEMLAB^® model with a nonlinear least-squares optimization function in MATLAB^®. The NO oxidation model, based on the kinetics measured from a separate study in our laboratory, was shown to agree well with the experimental observation on a Pt/Al₂O₃ catalyst within a wide temperature and concentration range (240-320^oC, 100-500 ppm NO, 3-25% O₂, 25-300 ppm NO₂). The breakthrough curves of total NOx under different conditions have an asymmetric shape, indicating the existence of different time scale processes. Two possible models were proposed to explain the adsorption of NOx on NSR catalyst: 1) A 2-sites in parallel model, which includes a fast adsorption site corresponding to BaO close to Pt and a slow adsorption site accounting for isolated BaO, and 2) A 2-sites in series model, which involves a "surface" BaO site that can uptake NOx quickly and a "bulk" BaO site accounting for the bulk diffusion of barium nitrate from the surface to the bulk. Both of these models were found to be capable of explaining the NOx breakthrough curves at different temperatures from 150^oC to 300^oC. These models were further validated by comparing model predictions of the NOx breakthrough curves with experimental data under different inlet NO₂ concentrations for a given temperature.

Both the 2-sites in parallel and 2-sites in series model, coupled with the adsorption kinetics assuming an NO₂ disproportionation mechanism (i.e., release of one molecule of NO for the consumption of three molecules of NO₂), agree well with the experimental breakthrough curves of NO₂ and NO when feeding NO₂ only to the Pt/BaO/Al₂O₃ catalyst. Again, the validation of these two models was tested by predicting experimental NO₂ and NO breakthrough curves under different inlet NO₂ concentrations.