Lean burn operation such as encountered in diesel engines provides higher fuel efficiency compared to stoichiometric fuel-air operation as encountered in most gasoline vehicles. However, the unconverted oxygen in lean burn vehicle exhaust prevents the use of conventional three-way catalysts to effectively reduce NOx to N2. During the past ten years there has been an increased interest in the development of exhaust after-treatment systems for lean NOx reduction. NOx Storage and Reduction (NSR) has been shown to be an effective technology by providing high NOx conversion during transient conditions. However, NSR requires periodic operation to sequentially store and reduce NOx. The objective of this study is to quantify the conversions and selectivities of NOx reduction by H2 over a wide range of operating conditions.

Cycling experiments were carried out with different catalysts (Pt/Ba, Pt, and Ba), loadings and dispersions over a large temperature range (50oC-500oC). Individual sets of experiments were run with varied reductant (H2), lean cycle time, and rich cycle. The results show that the Pt/Al2O3 catalyst is able to obtain similar time-average NOx conversion (30-60%) below 150oC as the Pt/BaO/Al2O3 catalyst where NOx storage on alumina occurs. Further, the time-average selectivity to NH3 and N2 was found to be strongly dependent on the amount of NOx stored (lean time). The NOx conversion maximum occurs in the 300 to 350 oC temperature window, similar to the temperature window for maximum NOx storage. Excessively rich regeneration conditions produce significant amounts of NH3 over the range temperatures studied. The amount of NH3 formed decreases with increasing temperature due to the surface coverage of H2 decreasing. Instantaneous effluent composition data reveal that N2O is produced at the onset of the rich pulse while NH3 is only substantially measured at the end of the rich cycle. The NOx conversion and NH3 formation and consumption are complex and dependent on the amount and type of stored NOx, stoichiometric number of the pulse (SN,P=(2[O2]+[NO]+2[NO2])/[H2]), H2/NO, temperature, and mass transfer. The NOx storage capacity determines the maximum attainable conversion, while the amount of reductant and rich pulse timing will determine the percent of stored NOx that will be reduced.

Results will also be discussed in which a systematic comparison is made between a Pt/Ba catalyst series having a fixed Pt loading with varied dispersion. The results will be interpreted with a phenomenological chemical reaction network within the framework of a monolithic reactor. The implications on effective NOx reduction over a range of operating conditions will be assessed.