Table 3
Acid properties of zeolite extrudates by pyridine adsorption method.
| Pyridine desorption temperature (◦C) | Fresh zeolite extrudate
|
Sample A
|
(Partially) regenerated sample A (95 ◦C, 2 h)
|
|||
| Brønsted sites (µmol/g) | Lewis sites (µmol/g) | Brønsted sites (µmol/g) | Lewis sites (µmol/g) | Brønsted sites (µmol/g) | Lewis sites (µmol/g) | |
|
150 |
599 |
219 |
374 |
136 |
562 |
202 |
| 250 | 529 | 151 | 325 | 104 | 498 | 128 |
| 350 | 419 | 124 | 273 | 94 | 396 | 107 |
Fig. 11. PMMA degradation yields: distribution of gas, liquid and solid products.
| As expected, the longer the regeneration, the higher the recovery of total acid sites, up to 71% of that of fresh catalyst after 4 h at 95 ◦C (Fig. 10a). In more details, TPD spectra deconvolution showed that time on stream mainly affected weak and medium acid sites, while the quantity of strong acid sites reached a quasi-plateau after 60 min. This level off of acidity restoration complied with the trend shown in Fig. 6a for carbon conversion. Thus the refractory coke fraction mainly blocked stronger acid sites, which are also known to promote oligomerization and condensation reactions [30].
The increase of the regeneration temperature from 80 to 100 ◦C (Fig. 10b) improved the recovery of zeolite acidity. On the other hand, a further increase of the outlet temperature from 100 to 140 ◦C decreased the amount of all acid sites. Again this result matched well with the carbon removal performance. From pyridine adsorption results, Table 3, it can be concluded that sites able to retain pyridine at 350 ◦C in fresh catalyst were mainly of Brønsted type and they should represent main fraction of medium and strong acid sites. The amount of total acid sites measured by this method was however lower than that from NH3– TPD. The difference between both techniques might be explained by a lower accessibility of pyridine to the microporous material due to its larger molecular size [31]. Moreover the pyridine method measured the remaining acid sites after desorption at 150 ◦C, while for NH3-TPD desorption started at a lower temperature (around 50 ◦C) and therefore weaker acid sites were accounted for by the integration of the desorbed NH3 amount. The pyridine results also suggest that both Brønsted and Lewis acid sites (estimated from desorption at 150 ◦C) were in similar amounts on the partially regenerated catalyst at 95 ◦C and on the fresh one (less than 10% difference). Moreover, the amount of total acid sites on sample A (prior to regeneration) would be about 60% of that on fresh ZSM-5, Lewis and Brønsted sites being equally affected. This too optimistic result with respect to NH3-TPD might be here explained by a partial desorption of coke molecules from acid sites caused by pyridine, as previously reported [27]. |
Nevertheless the two techniques agree in that the 2 h ozonation at 95 ◦C restored about 30% of original acid sites.
3.3.2. PMMA degradation results Fig. 11 compares the distribution of products from PMMA cracking obtained when using the different zeolite catalysts: fresh, coked (samples A and B), and after regeneration (varying TOS, as well as outlet temperature). After drying, the weight of solid – including catalyst – was in the range 0.39–0.43 g for all experiments: solid products accounting for less than 0.1% of the initial PMMA weight were not accounted for in the balance. The conversion of PMMA was thus considered total. The light oil was mainly composed of MMA, while heavier liquid products were C6–C16 (consisting in aliphatic, alicyclic as well as aromatic compounds, most of them being oxygenated: dimethyl esters, carboxylic acids). As expected from the acidity and carbon removal results, increasing the regeneration duration improved the zeolite selectivity towards light hydrocarbons. After 4 h ozonation at 95 ◦C, the regenerated zeolite gave almost the same (gas + light liquid) yield as the fresh one, yet with a smaller contribution of the gaseous compounds (especially for sample B). As previously shown, a temperature of 95 ◦C was better than 140 ◦C to restore the catalyst activity, but again main change was found in the gas yield. Therefore the differences in medium and strong acid sites observed between fresh and best regenerated zeolites affected only the ultimate stages of the cracking process, what should not be detrimental for the continuous process. 4. Conclusion Regeneration of coked ZSM-5 extudates of 3 mm diameter was successfully achieved using an ozone-enriched oxygen stream at about 100 ◦C: up to 80% of the coke was removed, restoring specific surface area and 70% of acid sites. No dealumination was observed. |
