Heterogeneous Catalysis Understanding the Structural Deactivation of Ruthenium Catalysts on an Atomic Scale under both Oxidizing and Reducing Conditions** Jens Aßmann, Daniela Crihan, Marcus Knapp, Edvin Lundgren, Elke Löffler, Martin Muhler,* Vijay Narkhede, Herbert Over,* Michael Schmid, AriP. Seitsonen, and Peter Varga The science and technology of catalysis are of central practical importance. About 80% of all industrial chemicals are manufactured by utilizing (heterogeneous) catalysis. Besides activity and selectivity, catalyst deactivation during use is a key issue in practical catalysis. “The importance of under- standing and being able to predict loss of activity during catalyst usage must not be under-estimated” [1] since replace- ment of a catalyst means high operational costs. Industrially used catalysts are, however, far too complex to allow for a microscopic understanding of why a catalyst deactivates. This knowledge calls rather for the use of model catalysts (such as single-crystalline surfaces) and their investigations under well-controlled ultrahigh vacuum conditions. The trade off for this so-called surface-science approach [2] is the introduc- tion of a pressure and a materials gap by which catalytic properties determined under well defined conditions may not be extrapolated to those at realistic reaction conditions. [3] For a ruthenium-based catalyst, activity loss was reported for the CO oxidation reaction. In particular, under oxidizing reaction conditions the activity of supported ruthenium catalysts declines substantially. [4, 5] This finding has been quite puzzling as recent investigations clearly indicate that RuO 2 is much more active than ruthenium in the oxidation of CO. [6] Since the pressure and materials gap for the CO oxidation over ruthenium are considered to be bridged [7] we can utilize the surface-science approach to clarify the micro- scopic processes determining the structural deactivation of ruthenium-based catalysts and how this atomic-scale knowl- edge is used to optimize the performance of practical ruthenium catalysts. We concentrate herein mainly on polycrystalline RuO 2 powder which is calcined at 573 K, resulting in a specific surface area of 0.9 m 2 g À1 . Complementary data of supported ruthenium catalysts are provided in the Supporting Informa- tion. The mean diameter of the particles in the RuO 2 powder is about 1 mm. Therefore RuO 2 powder represents a natural link between single–crystal ruthenium model catalysts and ruthenium catalysts supported on SiO 2 or MgO with an active surface area of 10 m 2 g À1 . [8] Applied partial pressures of CO and O 2 are in the range of 5–35 mbar. Figure 1 a displays the conversion of CO over oxidized RuO 2 polycrystalline powder as a function of time on stream, using a stepwise temperature variation and a CO/O 2 feed ratio of 1:2. During the first temperature cycle each temper- ature jump is accompanied by a rapid increase of the CO conversion followed by a transient decrease of the CO conversion whose steady state is not reached within 1 h. This deactivation process occurs faster at higher temperatures, whereas the extent of deactivation increases with higher concentrations of O 2 (see Supporting Information). For all investigated CO/O 2 feed stocks it is found that the conversion Figure 1. Conversion of CO over a) oxidized polycrystalline RuO 2 powder and b) pre-reduced polycrystalline RuO 2 powder as a function of time on stream and temperature. The applied cyclic temperature program (363 K–456 K) is shown in the gray trace (right-hand axis). The total flow rate was 50 mL min À1 with a CO/O 2 feed ratio of 1:2 (1.8 % CO/3.6 % O 2 ). The transient decrease of the CO conversion is indicated as a broken curve in the first cycle of Figure 1 a. [*] D. Crihan, M. Knapp, Prof. Dr. H. Over Physikalisch-Chemisches Institut Justus-Liebig-Universität Heinrich-Buff-Ring 58, 35392 Giessen (Germany) Fax: (+ 49) 641-99-34559 E-mail: herbert.over@phys.chemie.uni-giessen.de Dr. J. Aßmann, Dr. E. Löffler, Prof.Dr. M. Muhler, V. Narkhede Lehrstuhl für Technische Chemie Ruhr-Universität Bochum 44780 Bochum (Germany) Fax: (+ 49) 234-32-14115 E-mail: muhler@techem.rub.de Dr. E. Lundgren Department of Synchrotron Radiation Research University of Lund, Sölvegatan 14, 22362 Lund (Sweden) Prof. Dr. M. Schmid, Prof. Dr. P. Varga Institut für Allgemeine Physik, TU Wien Wiedener Hauptstrasse 8–10, 1040 Wien (Austria) Dr. A. P. Seitsonen Physikalisch-Chemisches Institut, Universität Zürich Winterthurerstrasse 190, 8057 Zürich (Switzerland) [**] H.O. and M.M. gratefully acknowledge financial support by Deutsche Forschungsgemeinschaft (DFG, SPP 1091). We acknowl- edge the Leibniz-Rechenzentrum in Munich for providing us with massive parallel computing time. The work in Vienna was supported by the “Fonds zur Förderung der wissenschaftlichen Forschung”. E.L. thanks for financial support by the Swedish Research Council. Partial financial support is acknowledged from the European Union under contract no. NMP3-CT-2003-505670 (NANO2). Supporting information for this article is available on the WWW under http://www.angewandte.org or from the author. Angewandte Chemie 939 Angew. Chem. 2005, 117, 939 –942 DOI: 10.1002/ange.200461805  2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim