Pablo Beato, Annette Trunschke, Robert Schlögl

1. synthesis reaction synthesis precursor-synthesis sol-gel, precipitation spin coating fundamental understanding powder thin film time T T in situ bulk- analytics in situ surface- analytics Mo-V-W-oxide catalysis catalyst catalyst Si/SiO2 at 298 K - after reaction after Ref. Mo5O14 before at 673 K after Si (NH4)6(Mo7O24) , V2O5/oxalic ac.,(NH4)6(W12O41) in H2O spray drying Mo: 68%, V: 23%, W: 9% (xrf of powder) grating powder + gelatin gel Si[100] powder thin film spin-coating 10 mm thin film slit Si[100] 350°C/air A. Müller et al., Current Opinion in Solid State and Materials Science4 (1999) 141-153 10 mm 300°C/air powder pellet Ø = 5 mm (5-15 mg) ccd catalyst model catalyst confocal hole Notch filter Laser (623 nm) cover plate middle plate side view base plate heater sticks (up to ~800°C) gas out to PTR-MS gas in In situ Raman studies on Mo-based mixed oxides during the partial oxidation of propene- an attempt to bridge the “materials gap” - Department of Inorganic Chemistry, Fritz Haber Institute of the MPG, Faradayweg 4-6; DE - 14195 Berlin, Germany, http://www.fhi-berlin.mpg.de Introduction The partial oxidation of light alka(e)nes e.g. propa(e)ne to higher valuable chemicals e.g. acrolein and/or acrylic acid is technically performed over multi metal oxides (MMO). Although much effort has been made to understand the working mechanism of MMO’s, the complexity of such systems has prevented up to now unequivocal proofs for the existing theories. It is obvious that in situ analytical methods are indispensable to get such proofs. The main problem is that most of the research on MMO’s has only been conducted via bulkanalytical techniques (XRD, EXAFS, IR, Raman, etc.) because charging effects and surface roughness of real powder catalysts make the reliable analysis by surface science methods impossible. However, a combination of surface sensitive and bulk sensitive methods are necessary to fully characterize and understand such complex systems. Our ultimate goal is therefore to prepare adequate models ranging from pure molybdenum oxides to complex mixed metal oxides which should preserve as much as possible of the chemical and structural complexity of real catalysts but at the same time stay accessible to both surface science and bulk analytic techniques. A promising strategy to meet these requirements is to deposit a thin film or nanoparticles of the metal oxide on an inert, conducting substrate, such as silicon. Our preliminary results are presented here. before Pablo Beato, Annette Trunschke, Robert Schlögl before Sumary & Conclusions & Outlook We believe that the concept to develop complex model systems as thin films in order to mimic real mixed oxide catalysts is promising. The prepared Mo0.68V0.23W0.09Ox thin film was able to directly convert propene to acrylic acid and was further activated during time on stream. This enables us to study such complex reactions on flat conducting catalysts by surface sensitive methods as could be in situ XPS. The presented micro-reactor can give valuable information on the reaction kinetics and at the same time offers the possibility to apply different types of optical in situ spectroscopies. The combination of in situ Raman spectroscopy and ex situ HRTEM gave a detailed view of the structural dynamics occurring during reaction. The possibility to directly compare thin film data and powder data has the great advantage to obtain information about bulk and surface structure under the same conditions. The observed structural changes during reaction clearly indicate, that the active phase of the investigated Mo0.68V0.23W0.09Ox catalysts is actually a very dynamic one and not to describe as a stiff crystal structure. For the catalyst to be active in the selective oxidation of propene there seems to be no need for a crystalline Mo5O14-type structure since the presence of small clusters containing the main structural motifs is sufficient. However, activity and selectivity seem to be enhanced when an ordered “superstructure” is present. More information about the composition of the catalytically active surface species will hopefully be gained in the near future by using the presented model for in situ XPS investigations. after