Experimental and Kinetic Modeling of Degreened and Aged Three-way Catalysts: Aging Impact on Oxygen Storage Capacity and Catalyst Performance

The aging impact on oxygen storage capacity (OSC) and catalyst performance was investigated on one degreened and one aged (hydrothermally aged at 955 °C for 50 h) commercial three-way catalyst (TWC) by experiments and modeling. The difference of OSC between the degreened and aged TWCs was dependent on catalyst temperature. The largest difference was found at 600 °C, at which the amount of OSC decreased by 45.5%. Catalyst performance was evaluated through lightoff tests at two simulated engine exhaust conditions (lean and rich) on a micro-reactor. The aging impact on the catalyst performance was different under lean and rich environments and investigated separately. At the lean condition, oxidation of CO and C₃H₆ was significantly suppressed while oxidation of C₃H₈ was relatively less degraded. At the rich condition, the inhibition effect was more pronounced on the aged TWC and inhibiting hydrocarbon species from C₃H₆ partial oxidation can survive at temperatures up to 450 °C. However, NO reduction activity declined less compared to CO and C₃H₆ oxidation. More NH₃ formed at low temperature and N₂O formation was suppressed on the aged TWC.

A generic TWC model including a dual-site oxygen storage sub-model and PGM kinetics was developed to predict the aging impact on dynamic OSC and catalyst performance. The PGM kinetics include oxidation of H₂, CO, and hydrocarbons as well as water-gas shift (WGS) and hydrocarbon steam reforming. NO reduction kinetics including N₂O and NH₃ formation and decomposition were also considered. The TWC models were calibrated on the degreened and aged TWCs separately based on experimental data. With the dual-site OSC model and calibrated kinetics, the dynamic OSC and lightoff performance on the fresh and aged TWCs were successfully predicted. The resulting changes of the OSC as well as lightoff performance due to aging were quantified and discussed with the help of the TWC models.