Dual reverse spill-over: Microkinetic simulations of the CO oxidation on Pd nanocatalysts C.J. Harding * , S. Kunz, V. Habibpour, U. Heiz Lehrstuhl für Physikalische Chemie, Technische Universität München, Lichtenbergstr. 4, D-85748 Garching, Germany article info Article history: Received 21 April 2008 In ﬁnal form 3 July 2008 Available online 9 July 2008 abstract Simulations of cluster-based catalytic oxidation of CO were performed using a dual reverse spill-over microkinetic model (DRSO-MK simulations) in order to improve the description and understand the underlying physics of the experimental turn-over data. The description of mass-selected Pd 13 TOF proﬁles as a function of mole fraction was reproduced to an excellent level at a range of temperatures. The DRSO- MK model was extended to produce predictions of reactants which are both mobile on the support, illus- trating the predictive power of the model and the importance of support interactions in the accurate modelling of cluster-based nanocatalysts. Ó 2008 Elsevier B.V. All rights reserved. The ﬁeld of model catalysis has diversiﬁed from the use of sin- gle crystals [1,2] as a catalytic medium to supported nanoparticles and clusters [3–6]. The potential to optimise the turn-over of reac- tants while reducing overheads in the production of bulk chemicals provides industrial applications in this area of research. The soft- landing of metallic clusters on metal-oxide substrates has shown highly tuneable properties, turn-over frequencies and remarkably low temperature catalysis [3,6,7]. To understand the physics be- hind the processes experiments can be performed and augmented with kinetic simulations which include intrinsic support effects of two reactants. Non-intrinsic support effects such as the role of ﬁlms in dimensionality cross-over [8,9] or charge transfer [10,11] are typically not measured by such methods. Recently intrinsic support effects have been added to a kinetic model, but the reverse spill-over of only one reactant was considered. The simultaneous dual reverse spill-over (of the two reactants) [12] in microkinetic simulations (herein referred to as DRSO-MK simulations) show interesting results and furthermore offers an added dimension to the predictive power of the model. Microkinetic simulations of oxidation reactions have already shown themselves to provide an excellent theoretical model of the CO oxidation reactions on nanoparticles and cluster-based nan- ocatalysts [13,14]. The catalytic material itself consists of support material namely a thin metal-oxide ﬁlm such as MgO [15]. After growing a ﬁlm to a thickness of 10 ML, clusters can be deposited onto the ﬁlm to produce the model catalyst. Mass selection of the clusters allows the tuning of the catalytic material affecting the turn-over frequency [3,16,17]. Experimental measurements of such catalysts form the basis of a comparison between the simula- tions and the physical situation. They are realised through a high level of control over the reactants. The use of two pulsed molecular beams, each containing a reactant, allows the mole fraction of the reactants to be accurately manipulated [3,16]. Coupled with the ability to independently vary the reaction temperature, coverage and size of clusters, a detailed investigation into turn-over fre- quency proﬁles shows remarkable effects such as the near room temperature catalysis of reactions ordinarily not observed for other catalytic systems. For a given temperature and coverage, varying the mole fraction of the reactants allows a turn-over frequency reaction proﬁle for the catalyst to be determined [3]. The variation of these proﬁles with temperature then provides the basis for kinetic information to be extracted. This was achieved by the comparison of microki- netic simulations to the experimentally measured data. Based on Langmuir laws, microkinetic simulations have shown to reproduce experimental data based on larger nanoparticles [14,18] as well as smaller clusters [13]. Although Langmuir laws have been shown for Pd single crystals [1,2] it has not been clearly shown if the same mechanism applies to small size-selected clusters. Theoretical cal- culations have indeed shown that the Mars-van Krevelen mecha- nism involving a Pd nano-oxide could also offer a route to the catalytic oxidation of CO [19]. However as the Langmuir laws have been shown to be suitable for nanoparticles this mechanism is con- sequently adopted for smaller Pd clusters. For the smaller cluster- based catalysts, important experimental observations were only reproduced when the kinetic formulations were augmented by an additional calculation; that of reverse spill-over using the cap- ture-zone model [20–22]. It has been illustrated that for small clusters (Pd 13 ), due to the ratio of active centres to support material, the theoretical descrip- tion is greatly enhanced by the addition of the migration of reac- tant molecules from the support to the reactive centres (reverse spill-over) using the capture-zone model. An increase in support 0009-2614/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2008.07.017 * Corresponding author. Fax: +49 89 289 13389. E-mail address: christopher.harding@mytum.de (C.J. Harding). Chemical Physics Letters 461 (2008) 235–237 Contents lists available at ScienceDirect Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett