Computational Design of Graphene–Nanoparticle Catalysts
Nanoclusters possess electronic properties that are inherently different from their bulk counterparts due to quantum effects that are strongly manifested at the nanoscale. Therefore, the catalytic properties of nanoclusters cannot be understood by a simple extrapolation of our understanding of catalysis on bulk metal surfaces. The presence of supports further modifies the electronic properties, morphology, and catalytic activity of a nanoparticle due to interplay between the electronic properties of the support and the nanoparticle. The objective of this research was to develop a systematic computational approach for designing and evaluating the activity and selectivity of supported catalyst nanoclusters, in the specific context of graphene-supported Pt and Pt-Ru nanoparticles for methanol decomposition. While the enhanced activity of nanoclusters has long been exploited in Pt–carbon black catalysts, recent experiments have shown that Pt–graphene catalysts can far outperform their predecessors, in particular, for methanol oxidation—a reaction of crucial importance for methanol fuel cells. However, the fundamental mechanisms by which graphene enhances the activity of nanoparticles for methanol oxidation as yet remain to be understood. Over the course of this research, we developed a comprehensive computational program toward modeling and understanding the activity of Pt nanoparticles on graphene supports. We developed genetic algorithms to identify low-energy isomers of Pt clusters on both pristine and defective (experimentally realistic) graphene supports. From structure devolve other properties such as reactivity and selectivity, which were probed in the context of CO oxidation and methanol decomposition. Motivated by the need for accurate structure prediction, we developed a new tight-binding parameterization for Pt-Ru alloys as well as empirical potential parameters for the Pt-Ru-C system to model support–cluster interactions. In conjunction with genetic algorithms, these tight-binding/empirical potential parameters allowed for efficient prediction of low energy structures that were then analyzed in greater detail with density functional theory (DFT). In particular, DFT modeling shows that strong cluster–support interactions—especially at point defects in graphene supports—are beneficial in rendering the Pt nanoclusters more tolerant to CO poisoning. Via DFT-informed microkinetic modeling, defective graphene-supported graphene Pt clusters were shown to be as active as Pt(111) surfaces for methanol decomposition further confirming that Pt-graphene nanocomposites can indeed efficiently catalyze this reaction with significantly lower Pt loading than traditional Pt catalysts. Interestingly, the active pathways on nanoclusters were altered relative to the bulk (111) surface indicating the potential for control over reaction intermediates, mediated by a combination of electronic and structural features of supported clusters. For Pt-Ru alloy clusters, our DFT-informed genetic algorithms consistently predict the formation of low-energy Ru-core/Pt-shell nanoparticles, thereby maximizing the use of expensive Pt for catalyzing surface reactions. Our studies open new avenues for computationally guided design of supported nanoclusters catalysts and, specifically, provide a sound theoretical basis for rational design of nanoparticle–graphene composites for fuel cell electrodes.
- Research Organization:
- University of Massachusetts Amherst
- Sponsoring Organization:
- USDOE Office of Science (SC), Basic Energy Sciences (BES). Chemical Sciences, Geosciences, and Biosciences Division
- DOE Contract Number:
- SC0010610
- OSTI ID:
- 1575127
- Report Number(s):
- DOE-UMass-SC0010610
- Country of Publication:
- United States
- Language:
- English
Similar Records
Palladium Monolayer and Palladium Alloy Electrocatalysts for Oxygen Reduction
Adsorption of CO on Low-Energy, Low-Symmetry Pt Nanoparticles: Energy Decomposition Analysis and Prediction via Machine-Learning Models