hide
Free keywords:
-
Abstract:
Temperature programmed reduction/oxidation (TPR/TPO) and X-ray photoelectron spectroscopy (XPS) are two different fundamental techniques to acquire information on the oxidation state of metal oxide catalysts. In TPR experiments the consumption of hydrogen by a catalyst sample is measured as a function of the catalyst temperature. TPR works under ambient pressure, and the amount of hydrogen consumed can be correlated to the decrease of the oxidation state of the analyzed sample. In contrast to TPR, XPS analysis is performed under ultra-high vacuum (UHV) conditions where the surface oxidation state can be obtained directly from the analysis of the shift of the core level binding energy for the particular metal atoms and by the deconvolution of the correspondent energy peaks using reference values for the binding energies at different oxidation states.
Analyzing the large amount of literature data for supported vanadium oxide catalysts, in case of TPR, there is a certain confusion which oxidation state of vanadium should be assumed as the initial or final oxidation state. Some authors proposed V(V) as the initial oxidation state, but they partially observed H/V rations significantly less than 2 [1,2,3]. In contrast, in [4] there is assumed a reduction of V(V) to V(IV) for γ-Al2O3 and TiO2 supported catalysts and of V(V) to V(III) for SiO2 supported catalysts. A reduction only to V(IV) was also proposed in [5]. In [6,7] H/V ratios partially larger than 2 were reported, what would indicate the presence of V in oxidation state smaller than +3. The only consensus is that vanadate species are more reactive with hydrogen than bulk-phase V2O5.
For XPS, similar discrepancies are reported, and in addition, the question of the stability of the sample composition under UHV conditions was not clarified up to now. Our latest results [8] show that the (usual) assumption of the VOx sample stability in UHV is generally saying not fulfilled. Using a multi-chamber UHV system (equipped with a PHOIBOS 150 hemispherical analyzer, SPECS), which allows a fast sample transfer (the first XPS spectrum can be obtained after less than 5 min only after the start of the evacuation of the load lock) we proved that a significant reduction of the sample can occur within first 15-20 min of exposition to UHV. Thus, some previous XPS results should be evaluated critically and there is still a need of new systematic investigations.
The scope of this contribution is to perform such a systematic study comparing the results of TPR and XPS studies. For this purpose, several series of VOx catalysts supported on γ-Al2O3, SiO2 and TiO2 with loadings up to 15 % V were prepared by impregnation of the supports with suspensions of VO(acac)2 in acetone or V2O5 in ammonia, water and acetone, followed by subsequent calcination at 650 °C under air for 8 h. Beside TPR/TPO and XPS the calcinated catalysts were characterized by XRD, BET, AAS and thermal analysis techniques to obtain additional information about the textural properties and the nature of the vanadyl species on the catalyst surface.
Our results indicate, that the dispersion of V increases in the order TiO2 < SiO2 < Al2O3. The acetylacetonate route of catalyst preparation can yield a higher V dispersion than the inorganic route. XPS data show, that the oxidation state of the calcinated samples increases with the increased presence of the bulk V2O5 phase, meaning below monolayer coverage, a significant part of vanadate is V(IV). Further, time resolved XPS measurements demonstrate that under vacuum conditions bulk-like V2O5 is more reducible than the dispersed vanadate species.
In contrast, at ambient pressure (during the TPR experiments), vanadate species are better reducible than V2O5. Summarizing the results, four different reduction peaks (attributable to isolated, dimeric and polymeric vanadate and bulk-like V2O5) can be clearly distinguished. In TPO experiments different types of curves for vanadite and bulk-like V2O3 were observed. Performing TPR-TPO cycling experiments, it can be observed, that the distribution between the different V phases changes during the redox cycles, for low V loadings a new low temperature reduction peak appears (Fig. 1). The H/V ratio increases with increasing V loading and with decreased V dispersion. By subtraction of the H/V ratio form the XPS results it can be concluded, that the final oxidation state of V after TPR experiments is +3 independent from the nature of V species, vanadate or V2O5 (Fig. 2). In the TPR experiments, the deficit in hydrogen consumption for low loaded catalysts is clearly located on the oxidized side and should be attributed to the significant presence of V(IV) in the “fresh” samples and not to an incomplete reduction of V(V). In contrast to V2O5, vanadate species do not undergo the full redox cycle between V(V) and V(III), a significant vanadate part remains at +4 after re-oxidation (TPO).
The financial support of the DFG (Forschergruppe 447, “Membranunterstützte Reaktionsführung” is acknowledged.
REFERENCES
[1] J, Keränen, A. Auroux, S. Ek, L. Niinistö, Appl. Catal. A 228 (2002), 213
[2] J. M. Kanervo, M. Elina Harlin, A. O. I. Krause, M. A. Banares, Catal. Today, 78 (2003), 171
[3] P. Concepción, M. T. Navarro, T. Blasco, J. M. López Nieto, B. Panzacchi, F. Rey, Catal. Today 96 (2004), 179
[4] M. L. Ferreira, M. Volpe, J. Mol. Catal. 184 (2002), 349
[5] K. V. R. Chary, G. Kishan, C. P. Kumar, G. V. Sagar, Appl. Catal. A, 246 (2003), 335
[6] E. P. Reddy, R. S. Varma, J. Catal. 221 (2004), 93
[7] A. Gervasini, P. Carniti, J. Keränen, L. Niinistö, A. Auroux, J. Catal. 96 (2004), 187
[8] Y. Suchorski, L. Rihko-Struckmann, F. Klose, Y. Ye, M. Alandjiyska, K. Sundmacher, H. Weiss, Evolution of oxidation states in vanadium-based catalysts under conventional XPS conditions, Appl. Surf. Sci., in press, available online at www.sciencedirect.com
