Fig. 4. Thermograms of coked zeolites under N2, along with identification of evolved gases from IR spectroscopy: (a) sample A and (b) sample B.
| seemed better correlated with the increase of “hard coke”, and that of micropores with the increase of “soft coke”.
3.2. Regeneration of the catalysts by ozonation For sample A, very high carbon conversion up to 80% could be achieved in the investigated conditions (see Fig. 6). For sample B, performance was more modest (35% C removal), mainly due to the fact that the particles were coated by a PMMA rich layer, as found by TG/IR, that limits the access of ozone to the zeolite. Therefore parametric study was performed with sample A only. 3.2.1. Effect of operating parameters Carbon removal was found to increase almost proportionally with the inlet concentration of ozone in the investigated range, as seen in Fig. 5. As expected, conversion was improved by increasing time on stream, but it seemed to reach a plateau after 2 h (Fig. 6a). The effect of temperature, depicted in Fig. 6b, is more complex as an optimum was found at about 100 ◦C regardless time on stream, due to some competition effect between the chemical and mass transport kinetics involved, as it will be explained below. Measurements of ozone concentration at the reactor outlet showed that less than 10% of the ozone was thermally decomposed at 100 ◦C (in an empty tube), while over fresh ZSM-5 extrudate, ozone decomposition could reach more than 90% along the string reactor, as seen in Fig. 7. Outlet ozone concentration was also analyzed on the beginning of the decoking reaction (0–30 min): lower |
ozone decomposition was observed with the coked catalyst below 100 ◦C due to its lower activity.
Optimal conditions were then further investigated by varying operating parameters at about 100 ◦C for the same amount of ozone provided to the sample. Similar results were obtained when increasing the gas flow rate from 12.7 to 45.4 l/h (and decreasing time on stream accordingly from 2 h to 33 min): about 40% C removal (for an inlet O3 concentration of 19 g/m3). In comparison, conversion was 24.7% for 12.7 l/h and 30 min TOS (other conditions being equal). Increasing the gas flow rate reduced the ozone concentration gradient inside the reactor, which compensated for the diminution of TOS. Furthermore, carbon removal was decreased when using a higher gas flow rate with lower ozone content (at same total ozone input): 55.8% for 12.7 l/h and 48.2 g/m3 (over 30 min) against 32.9% for 41.3 l/h and 13.7 g/m3. This confirms that the decoking rate increases with ozone concentration. 3.2.2. Carbon profiles in the reactor and in the particles Axial coke conversion profiles were measured at about 100 ◦C after 1 h of reaction by dividing the reactor into 5 zones (containing 3–4 pellets each) and analyzing separately their residual carbon content (which was compared to that of the corresponding reference half pellets). As usual, each separate lot of particles was crushed and homogenized prior to flash combustion analysis. Fig. 8 shows that carbon conversion varied slightly along the reactor for sample A, from 59% to 50%, and from 26% to 14% for sample B. Surprisingly, the axial gradient of ozone concentration appears thus to |
