Table 2
Elemental analysis from EDX spectroscopy and carbon content from flash combustion.
| Fresh zeolite extrudate | Sample A | Sample B | (Partially) regenerated sample A (95 ◦C, 2 h) |
| Elemental analysis from EDX (wt.%/at.%) | . | . | . | . |
| C | – | 3.7/6.0 | 9.2/14.6 | a |
| O | 50.1/63.6 | 51.7/62.8 | 0.11 | 0.19 |
| Al | 0.8 | 339 | 0.13 | 0.17 |
| Si | 0.6 | 326 | 0.13 | 0.16 |
| Carbon content from flash combustion (wt.%) | – | 361 | 0.13 | 0.18 |
a Below the limit of detection.
|
similar pore size distribution as the fresh catalyst after 2 h of ozonation at 95 ◦C, while at higher regeneration temperature (140 ◦C) only its mesoporous volume was fully restored. 3.1.2. Elemental analysis The elemental profile along the cross sectional area of the particles was analyzed by SEM/EDX for both coked samples (Fig. 3): the carbonaceous deposit was found well distributed across the extrudate, with the exception of a narrow zone of higher concentration at the very outer surface (depth < 50 m). The mean carbon contents estimated from EDX spectroscopy were in accordance with those determined from flash combustion of the materials (Table 2), with a much higher value for coked sample B. The Si/Al atomic ratio of about 2.5 also agreed well with the theoretical value (accounting for the presence of alumina binder) and kept unchanged after regeneration (no dealumination). 3.1.3. Thermogravimetry The thermograms of coke samples (realized under N2), as well as the identification of evolved gases from IR spectroscopy, are given in Fig. 4. The amount of physisorbed water decreased logically from fresh zeolite (9.5 wt.%) to sample A (5.8%) and sample B (3.2%). Then three distinct stages were observed for coked samples after the 120 ◦C plateau. A first weight loss of less than 1% was observed from 120 ◦C to 250–300 ◦C corresponding mainly to the release of water: it might be due to coordinated water, as it was also observed on the fresh zeolite. CO2 emission then started to increase with a maximum intensity observed between 400 ◦C and 600 ◦C, and was |
. detected up to the end. CO release came later, mainly after 600 ◦C. Some compounds were only detected for sample B, namely MMA (with a narrow peak around 405 ◦C), methanol (around the same temperature), methane (385–665 ◦C) and ethylene (425–585 ◦C). This difference might be attributed to a PMMA rich layer coating the top particles. It should be also noticed that the corresponding rate of weight loss (from 300 ◦C to 450 ◦C) was much higher than that of other stages. Some complementary TG analyses were also carried out under air. They showed that the sample weight decreases up to 700 ◦C, with a total weight loss after the 120 ◦C plateau of 6.5% and 17.3% for samples A and B, respectively (for a respective carbon content of 4.1% and 11.5%). The coke deposit is usually qualified as “soft coke” or “hard coke”, depending to its ability to be volatilized/decomposed under nitrogen and oxygen. Following the work of Wang and Manos [28], the amount of coke precursors or “soft coke” can be roughly estimated from the weight loss observed between 250–300 ◦C and 600 ◦C under N2, resulting in 1.5% and 9.5% for samples A and B respectively. The amount of “hard coke” is obtained from the difference between the total weight loss under air and the weight loss up to 600 ◦C under N2: 4.2% and 7.3% for samples A and B respectively. According to the conclusions of Elordi et al. on the cracking of polyethylene [24], “soft coke” would be mainly deposed outside the zeolite crystals, while “hard coke” would be formed in the interior of crystalline channels. However the location of the carbonaceous deposits depends also on the hydrocarbon used as the coke-forming agent [20]. Here the reduction of mesopores |
Fig. 3. Elemental carbon profile a on a cross-section of extrudate A.
