Hydrogen Production by Methanol Steam Reforming on Copper Boosted by Zinc–Assisted Water Activation**

For use of polymer electrolyte membrane fuel cells (PEMFC) in mobile power applications, an efficient source of COdepleted hydrogen is needed. To avoid technical and safety problems of hydrogen handling, storage, and transport, methanol can be used as practical and abundant energy carrier for on-board H2 generation, as it has the advantage of a high energy density. Hydrogen generation from methanol can be performed by catalytic methanol steam reforming (MSR): CH3OH+H2O!CO2+ 3H2. Methanol conversion must be carried out with very high CO2/H2 selectivity to avoid CO poisoning of the fuel-cell anode. A number of promising selective MSR catalysts are already available. Apart from advanced copper-based catalysts, special attention is presently paid to the highly MSR-selective reduced state of Pd/ZnO, containing a particularily stable intermetallic PdZn (1:1) active phase. Therefore, we recently studied related “inverse” near-surface PdZn intermetallic phases, showing that three-dimensional PdZn active site ensembles are equally important for selective dehydrogenation of methanol (thus avoiding CO) and for efficient water activation. For the less costly Cu/ZnO catalysts, originally designed for methanol synthesis, improvements towards a technical MSR application regarding sintering stability, pyrophoricity, and selectivity are still required. Empirical development of Cu/ZnO catalyst preparation and activation has aimed in a particularily large Cu–ZnO contact. Nevertheless, it is very difficult to derive an unambiguous causality for the role of the contact on technical catalysts. It is known that zinc leads to an improvement in the desired properties, but a clear assignment of a predominant promotional effect (both from the theoretical and experimental side) is still missing. In the Cu/ZnO literature, seemingly incompatible model interpretations can be found, involving the “metallic copper model”, the “special site model”, the “morphology model”, the “spillover model”, and last but not least the “Cu-Zn alloy model”. Consequently, the Cu-ZnO(H) contact most likely constitutes a combination of promotional effects. The central aim of our study is to highlight the aspect of zinc-promoted water activation. This is achieved by using an ultrahigh-vacuum (UHV) “inverse” model catalyst approach, which, in contrast to investigations on real catalyst systems, allows the zinc segregation behavior and the changes in redox chemistry of both copper and zinc to be better followed. This provides a solid basis for directional promotion of microkinetic steps leading to enhanced CO2 selectivity. A series of CuZn near-surface alloy and bulk brass samples were tested. The near-surface alloy preparations on Cu foil involved variable thermal Zn evaporation and thermal annealing conditions. Commercial bulk brass foil samples with 10 to 37 weight% Zn content were cleaned by usual UHV procedures. The respective bimetallic pre-reaction state was characterized by depth-resolved X-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy (AES) analysis of the alloy composition. Quantitative kinetic MSR reaction studies were then performed in a UHV-compatible all-glass reaction cell operating between 10 10 and 1000 mbar. Finally, simultaneous “in situ” analysis using product detection by mass spectrometry and ambient-pressure XPS (APXPS) was performed at the HZB/BESSY II synchrotron (for details see the experimental section in the Supporting Information). On this basis, we are able to show that the CO2 selectivity and activity in MSR strongly scales with the available Cu(Zn)/Zn(ox) interface, which forms in situ by partial oxidative Zn segregation in the MSR gas phase on a suitable near-surface alloy “pre-catalyst”. It should be emphasized that the exact degree of initial Zn doping is crucial to establish a maximum Cu(Zn)/Zn(ox) interface and thus activity. Water activation and total oxidation toward CO2 is inhibited on bulk and surface-clean, thermally annealed, structurally equilibrated copper foil, which is a particularily unreactive state of Cu (for a brief discussion of potential Cu activators other than Zn, see the Supporting Information, S5). Near-surface alloys that were too Zn-rich and all investigated bulk brass samples also showed complete deactivation by [*] Dr. C. Rameshan, Mag. W. Stadlmayr, Dr. S. Penner, Mag. H. Lorenz, Prof. N. Memmel, Prof. B. Klçtzer Institute of Physical Chemistry, University of Innsbruck Innrain 52A, 6020 Innsbruck (Austria) E-mail: [email protected]