Characterizing catalysts under working conditions is crucial to understand and to optimize their behavior and performance. However, when Raman spectroscopy is used, attention has to be paid to laser-induced artefacts. While laser irradiation is often claimed to lead to a temperature gradient between the integral catalyst bed and the sampling point, neither the circumstances when such effect appears, nor if it systematically occurs or not, are really explored in details. The present paper shows that the sensitivity of a catalyst to laser-induced heating largely depends on the gas composition under which the analysis is done, in particular that it depends whether the catalyst has adsorbed reactant molecules or not. These aspects are here addressed via the Raman in situ exploration of H3PW12O40. This heteropolyacid is a widely used acid catalyst due to its very high Brönsted acidity, approaching the superacid region. In particular, we have investigated the impact of laser irradiation in the Raman monitoring of solid H3PW12O40 at work under a flow of methanol in nitrogen at 50°C. When 1 single spectrum of H3PW12O40 was measured after 3h of exposure to methanol, the characteristic CH vibration bands of adsorbed methanol appeared. However, when spectra were measured continuously throughout the experiment, the same CH vibration bands were observed only during the first hour, then they disappeared and the characteristic bands of polyaromatic molecules appeared. Under continuous laser irradiation, adsorbed methanol was thus converted into polyaromatic coke as resulting from a laser-induced heating. However, the spectra collected under pure nitrogen show that the laser does not heat the catalyst in the absence of methanol. UV-Vis revealed the reason of the laser-induced heating in the presence of methanol, and the subsequent formation of coke. Actually the catalyst gets reduced by the adsorbed methanol, what darkens the catalyst bed. Such a darkening renders the catalyst sensitive to laser-induced heating, which in turn leads to the formation of coke. Under continuous laser irradiation, methanol thus auto-initiated its own catalytic conversion, finally leading to the deposition of coke. Such artefact must be avoided if one wants to study the true behavior of the catalyst at work. This paper shows that, for reducible samples analyzed in the presence of reductive molecules, this is only possible by shining the laser intermittently and not continuously. More generally, it actually shows that the adequate way to irradiate a catalyst (continuous vs intermittent) in an in situ/operando Raman analysis depends on the gas flow composition.
Keywords: Coke; Heteropolyacid; Laser-induced heating; Methanol; Operando Raman spectroscopy; Polyaromatics.
Copyright © 2016 Elsevier B.V. All rights reserved.