Understanding and controlling the stability and reactivity of noble metal nanoparticles for CO oxidation

Abstract

Heterogeneous catalysts comprising of noble metal nanoparticles (such as Pt, Pd and Rh) supported on oxides, play a key role in a wide range of important chemical transformations including automotive exhaust control. However, due to the increasing demand for these catalysts and scarce resources of noble metals, there is a pressing need to reduce the consumption of noble metals in catalysts. Sintering of nanoparticles can cause catalyst deactivation, hence stabilizing nanoparticles can be a solution to reduce the loss of noble metals. Moreover, by improving the catalytic activity of noble metal nanoparticles, the amount of noble metals in catalysts can be reduced while still maintaining the required catalytic performance. Therefore, this thesis focuses on some novel nanostructures of noble metal catalysts that can potentially lead to enhanced stability and improved activity, in order to use noble metals more efficiently. CO oxidation was chosen as the model reaction in this study, because of its importance in automotive exhaust control. Interactions between metal nanoparticles and the support can have big influences on particle stability, as it was demonstrated in this thesis that weak particle-support interactions would lead to nanoparticle sintering under reaction conditions and hence destroy the dedicatedly designed nanostructures. In addition, the particle-support interactions may also bring some emergent functionalities that can affect the catalytic activity. Hence, different approaches were attempted to enhance the particle-support interactions, and their effects on the stability and activity of the catalysts were investigated. In the first approach, noble metal nanoparticles (Pd) were enclosed into porous organic cages (POCs, a class of hollow, cage-like macromolecules). The POCs were able to confine the nanoparticles, which resulted in a uniform particle size distribution. However, the limited accessibility of active sites in POCs and the thermal decomposition of the POC support (~300 °C) largely restricted the activity of the resulted catalysts hence their practical applications. The alternative approach was to partially embed (socket) noble metal nanoparticles in perovskite oxides via redox exsolution method. This has been previously demonstrated to produce highly stable transition metal nanoparticles. For the first time this thesis investigates the in situ formation of the socketed particles while at the same time providing valuable mechanistic insight for designing more efficient exsolved materials. Experiments have been conducted in situ in a latest generation environmental transmission electron microscope (ETEM) which allowed for the direct observation of the socket formation, metal particle nucleation and growth. The socket was found to form simultaneously with the particle growth due to the rise of perovskite lattice around particles. The particle growth seemed to be limited by the availability of exsolvable ions near the perovskite surface, which highlighted the importance to reduce the perovskite grain size when attempting to exsolve from dilute compositions. All the above mechanistic insight was employed to design materials that can exsolve from dilute substitution of noble metals thus potentially allowing for more efficient use of noble metals. Parameters such as substitution levels and reduction time and temperature were used to control exsolved particle characteristics and relate them to the catalytic activity. By comparing with the state-of-the-art Rh catalyst, the exsolved Rh catalyst with the same nominal metal loading exhibited similar activity despite that only a part of Rh in the bulk of perovskite exsolved on the surface. This indicates that the activity of exsolved catalysts can be enhanced probably due to the emergent functionalities that arise from the strained particles, which could potentially reduce the amount of noble metals in catalysts if the extent of exsolution can be increased. This thesis highlights the following design principles for noble metal catalysts. The stability of metal nanoparticles on the support must be high enough to maintain the designed nanostructures. That means that we need to have stronger particle-support interactions. Attempting to do this by full encapsulation was successful but compromised activity. Therefore, a partial embedding via the exsolution method results in a combined stabilizing effect and increased activity due to strain. Ultimately, this appeared to be the most promising method for designing efficient noble metal catalysts.