The effects of the strong metal-support interaction on Pd/ZnO and its catalytic behavior is investigated in an article by Patrick Kast et al. They reduced samples of differing loadings under various conditions of reducing temperature and atmosphere and used them to catalyze CO-oxidation. These were characterized using a variety of techniques.
The strong metal-support interaction (SMSI) is an important concept when dealing with catalysis. Using a very simplistic model, only the catalyst and the reactants are regarded as participating in the catalytic reaction. The support material, by contrast, merely provides an electrical connection between the catalyst and the external load, and therefore needs to be electrically conductive, while being stable in the operating environment, but does not participate in the catalytic reaction, as such. As it turns out, this view is a bit naïve, or at least it does not always apply. Sometimes, the support material can interact with the catalyst metal adsorbed to its surface so as to modify the metal’s electronic properties, and thus its catalytic properties – this is the strong metal-support interaction. A classic example of SMSI is the interaction between Pt and TiO2: the titanium oxide interacts with the platinum so as to reduce its ability to bind with H2. SMSI can also induce the migration of the catalyst and support materials, resulting in the formation of alloys, over-layers, etc.
The authors of the paper used Pd/ZnO as their catalyst/support structure, and investigated the effects of SMSI on CO-oxidation. The catalyst was synthesized via a co-precipitation technique, as it was thought that this would allow for more uniform SMSI, not favoring specific sites, than with an impregnation technique. Given as SMSI depends upon the reducibility of the catalyst support, temperature-programmed reductions of the samples were performed, in which 5% hydrogen-helium, and 2% Co-He gas mixtures were applied, and the results compared. Structural changes during these reductive heat treatments were monitored using XRD and TEM. The authors found that, with the 5 wt% sample, there was detectible PdZn alloy formation. They concluded that an amorphous, partially reduced ZnO compound forms at the surface of the sample during reduction.
Samples were also subject to reduction at various temperatures, followed by a series of catalytic CO-conversion cycles. When reduced at 423 K, the catalyst was found to be less active for the first CO-conversion cycle than with the following cycles. After reduction at 523 K, the first conversion cycle exhibited higher catalytic activity than following cycles, however, when repeated at 523 K, the catalytic behavior was replicated almost exactly. With reduction at 673 K, there was less pronounced difference between the first conversion cycle and following cycles. This behavior of reduced activity on repeated cycles for those samples reduced at higher temperatures led the authors to conclude that the activation that occurred during reduction was being counteracted by other processes that were reducing the active surface area. Corroborating this, using HAADF-STEM, the authors were able to identify a ZnOx over-layer which had formed on the Pd particles, which explains the reduction in catalytic activity. From XRD and kinetic studies, it was concluded that, at 623 K and higher, the Pd should be completely alloyed. The authors therefore proposed that the 673 K reduction resulted in the formation of PdZn particles on the ZnO support, with a ZnOx over-layer on top of those particles.
Thus we see that in the case of this Pd/ZnO system, the SMSI both helped and hindered its catalytic activity – being responsible for alloying which activated the catalyst, but also the migration that resulted in the formation of an active site-limiting ZnOx over-layer. This work with its monitoring of catalytic activity after reductions at various temperatures, points to the prospect of using SMSI under controlled conditions to tailor the structure and behavior of catalyst/support materials.
Source: P. Kast et al. Catal. Today, 260, 21 (2016).