Speaker
Description
Coking is unavoidable in various catalytic reactions, e.g. non-oxidative propane dehydrogenation (PDH), which enforces the need for removal of carbon deposits in regeneration processes.$^{1}$ This is typically achieved by air oxidation at temperatures between 600 - 700 °C.$^{2}$ In the case of Pt-based catalysts, oxidative treatments at such high temperatures lead to irreversible changes to the Pt particles that alters morphology and dispersion, and finally to deactivation of the catalyst.$^{2}$ In this contribution, we report on the feasibility of coke removal from supported Pt catalyst coked under PDH at temperatures as low as 310 °C in a nitric acid vapor (NO$_{x}$) containing atmosphere. $^{3}$ Samples of Pt catalyst at three states: fresh, used and used regenerated in NO$_{x}$ were analyzed using Raman spectroscopy, DRIFT spectroscopy, temperature programmed oxidation (TPO), transmission electron microscopy (TEM) as well as microcalorimetry. The tests under PDH reaction of HNO3 vapor treatment catalyst show an activity reaching to >95% of the activity of the fresh catalyst. In contrast, a used catalyst shows <55% and a conventionally air regenerated used catalyst shows <75%. The TPO experiments reveal that a HNO3 vapor treatment efficiently removed a high portion (ca. 90%) of the PDH coke, and the small amount of residual coke requires milder conditions for removal. Raman spectroscopy confirmed these findings and additionally showed that the residual coke provides a strong disordered character and thus is notably different from the ordered graphite-like original PDH coke. TEM imaging of NO$_{x}$ regenerated used catalyst did not hint at presence of any trace of ordered carbon near the Pt particles or the support. DRIFTS experiments using CO probe molecules revealed that the freed Pt surface after NO$_{x}$ regeneration is comparable to that of fresh catalyst and no evidence for Pt sintering could be found. CO adsorption microcalorimetry further confirmed these results by revealing a quantitative similar number of Pt adsorption sites for both fresh and NO$_{x}$ regenerated used catalyst. However, both catalyst states display differences in the CO differential heat profile due to a minor transition from sub-nanometer/atomic-Pt to nm-size Pt or changes in the Pt-support interaction. Finally, the minor loss of about 5% propylene yield after NO$_{x}$ regeneration can be explained by a small amount of residual disordered graphitic carbon still present while Pt clusters are left almost unchanged. Oppositely, oxygen regeneration unselectively removes carbon whereas with NO$_{x}$ regeneration interstitial carbon remains to major extent in the bulk having a beneficial effect on the propylene production compared to conventional regeneration.
| Abstract Number (department-wise) | AC 4.3 |
|---|---|
| Department | AC (Schlögl) |
