This presented paper deals with NOx trap sulphur poisoning and its regeneration. Sulphur poisoning has been studied with different SO2 gas concentrations under laboratory and engine test bench conditions. The sulphur poisoning studies have shown that the different NOx-traps available in the market have different behaviours toward SO2 poisoning and are all very sensitive to it. The results outline a non linear relationship of the NOx trap sulphur poisoning as a function of SO2 concentration. For instance, engine bench tests show that with a 50 and a 110ppm sulphur containing gasoline, a decrease of 50% in the NOx-trap storage capacity is respectively observed after 20 and 15 hours. With a gasoline containing 20ppm of sulphur, the same deactivation level is observed after 90 hours. A simulation carried out from these experiments and a study dealing with the European customers driving profile suggests that with a gasoline containing very low amount of sulphur (for example: <10ppm), the NOx-trap regeneration could be executed naturally in most cases. These results have to be confirmed on long duration durability tests.
Concerning the sulphur regeneration, the results show that the NOx-trap initial activity recovery can take place without the complete desorption of SO2 initially adsorbed by the NOx-trap.
In other respects, it is noticed that regeneration time does depend not only on conditions such as temperature, lambda or NOx-trap formulation but also on the NOx-trap ageing. For example, while between 2 and 5 min are required to purge the sulphur at 700°C with lambda=0.97 for a fresh catalyst, it takes 15 min for an aged one.
A parametric study concerning the regeneration has shown that a good strategy to decrease sulphur regeneration time could be to begin the regeneration at high temperature with low overrichness and continue it by lowering the temperature but increasing the richness.