*3.3. MOC Performance in Regeneration*

The regeneration, recovering the catalyst efficiency for methane oxidation, already indicates SO2 is most probably released from the catalyst during the regeneration and the SO2 measurement confirms this. In each case, during the 5 min regeneration a release of SO2 was detected with the FTIR downstream of the catalyst. The total SO2 amounts released over the 5 min regeneration periods are collected to Table 3. The lowest SO2 amounts were released in regeneration of MOC, which was tested without any additional SO2 in the exhaust. In the second case ("MOC + SO2") where more SO2 was present in the exhaust, also more SO2 was found to be released from the MOC during the regenerations indicating that more SO2 is collected to MOC also. If we look at for example the third regenerations (Table 3) we see that in the case of "MOC-no SO2", 0.33 g SO2 was released while in the case of "MOC + SO2" 0.54 g SO2 was released. In the case of "trap + MOC" the SO2 released during the regeneration was 1.52 g (see Table 2), indicating roughly that the "trap" is contributing to the release with SO2 amount of 0.98 g. This means that the trap is collecting at least the same amount of SO2. In addition, this SO2 release indicates that the trap itself is regenerating at the same time as the MOC. Also, since the methane oxidation efficiency is recovered (Figures 3 and 4), no significant amount of the SO2 released form trap during the regeneration period is expected to be collected in MOC but only flowing through the MOC.

**Table 3.** The SO2 amounts (g) released during regenerations. \* Note. 4.reg was done after weekend. In case of Italic labelled values, engine-driving mode was not exactly as intended, see text for more details.


When the extra SO2 was added to exhaust, the total SO2 available during 24 h of driving was 4.10 g. Therefore, in the case of "MOC + SO2" the released SO2 amount was 13% of the amount available, while in the case of "trap + MOC" the released amount was 37% of the amount available. Since the "trap + MOC" released significantly more SO2 compared to MOC only, the "trap + MOC" must also be collecting more SO2. This is what we saw in the catalyst performance and ageing tests, since the methane conversion of "trap + MOC" was higher than in the case of MOC only indicating the poisoning effect of SO2 was less in the case of "trap + MOC".

The results in Table 3 also show that in the case of "trap + MOC" when regeneration is done after a longer time period (4th regeneration) some more SO2 is released during the regeneration. One should note that in this case (of "trap + MOC") the engine-driving mode was changed between third and fourth regenerations, meaning that the methane slip increased to near 2000 ppm level in ageing mode (discussed above). The actual regenerations were repeated as similarly as manually possible and now significant changes were observed in the regeneration mode engine out emission levels.

Regarding the regenerations conducted during the first tests (the case of "MOC-no SO2") the adjustment to stoichiometric mode was not always realized similarly. In the cases of the 5th and the 6th regenerations, changing the driving mode to stoichiometric did not happen as smoothly as for the other cases (therefore we marked those in italic to the table).

During the regeneration, the exhaust gas composition changed significantly (Table 1). However, the MOC also worked during the regeneration. Although the engine out CO level increased to above 9000 ppm in the short regeneration phase (Table 1), the level

measured downstream of MOC increased only after approximately 2 min from the start of the regeneration (i.e., turning the engine to stoichiometric driving mode) and did not reach to higher than 3000 ppm (Figure 5A). This also had strong influence on the exhaust temperature since oxidizing high amount of CO contributes to the temperature increase. This was also observed by temperature measurement downstream the MOC showing an increase of 50–60 degrees during the regeneration (Figure 5A). Furthermore, no H2 was observed downstream the MOC, confirming that also the H2 (engine out concentration 0.6%, Table 1) was oxidized during the regeneration, influencing exhaust temperature (measured downstream the MOC) as well.

**Figure 5.** Exhaust temperature downstream the MOC. SO2 and CO downstream the MOC during the regeneration phase. (**A**) the case of "trap + MOC" for 1st and 3rd regenerations (**B**) the case of "trap + MOC" for 4th regeneration.

Interestingly, the regeneration of "trap + MOC" done after longer time period (68–140 h) resulted in different behavior than all the other regenerations. As discussed above, in this case the methane slip from the engine increased during 68–140 h to a level of 2000 ppm. The temperature increase during regeneration was 20–30 degrees higher in this case (Figure 5B) compared to other regenerations (Figure 5A). The exhaust temperature downstream the catalyst was a few degrees higher in the normal driving mode ("ageing") as well (see Figure 5A,B prior to regeneration start). Since the methane conversion over the MOC was similar for both methane slip levels (normal 1000 ppm versus this higher 2000 ppm), the MOC was oxidizing more methane in this higher slip case and this methane oxidation increase can also result in temperature increase.

The CO level downstream the MOC in this regeneration of Figure 5B (i.e., max. 1000 ppm) was significantly lower than in other regenerations although the engine out CO level was similar in all cases. This more effective CO oxidation can be due to temperature increase while this CO oxidation also influences to the temperature increase itself. In addition, a higher SO2 release was observed (Figure 5B). The temperature increase might be one reason for this while also the fact that after a longer period without regenerations also more SO2 is most probably collected to the trap and therefore more SO2 can be released during the regeneration as well. Also, the SO2 level at the end of regeneration is on higher level in the regeneration of Figure 5B being approx. 190 ppm while in the case of Figure 5A the SO2 level at the end of regeneration is approx. 50 ppm. The higher SO2 level at the end of regeneration might indicate that if the regeneration mode is to be continued for a longer time period also more SO2 could be released. However, on the other hand, especially in the case of Figure 5A i.e., daily regenerations, most of the SO2 is released in the first half of the regeneration period. This indicates that shorter regeneration times could be relevant.
