Understanding the surface structure of a catalyst under a reaction environment is challenging, yet necessary. Now, a combination of in situ methods reveals the reversible formation of a surface alloy as the active phase for core–shell Ni–Au nanoparticles during CO2 hydrogenation, which could not be detected by ex situ methods.
Since the direct observation of Cu/ZnO nanoparticles (NPs) reshaping under reactive gas environments using environmental transmission electron microscopy (ETEM) in 2002, a number of works have shown that supported metal NPs can undergo structural changes during reaction and evolve with respect to their size, morphology, composition, or surface chemical states1,2,3,4,5. This poses serious challenges for conventional catalyst design strategies that rely exclusively on static characterizations. In fact, it is known that the general strategies to increase the number of active sites or to tune the electronic properties of a catalyst are based on engineering its structure outside of the reaction environment in which it is supposed to work6. Obviously, it becomes challenging if these well-defined structures change during actual reactions.
As for core–shell NPs, a very popular type of heterogeneous catalyst, it is accepted that their catalytic properties originate from the synergy between the core and the shell, which imparts specific electronic and geometric features. However, recent pioneering works on in situ X-ray photoelectron spectroscopy revealed important changes in the near-surface structure of working bimetallic catalysts2,5,7. Therefore, directly visualizing the localized atomic coordination/arrangement via in situ microscopies is important for an in-depth mechanistic understanding of catalytic processes on core–shell NPs.
Reporting in Nature Catalysis, Wei Liu, Yi Gao, Meng Gu, Bing Yang and colleagues have now made a crucial step to address this challenge by designing a core–shell Ni–Au bimetallic nanocatalyst model system and subsequently investigating its atomic and electronic structures via multiple in situ microscopy and spectroscopy methods8 (Fig. 1). Based on former discoveries focusing on in situ microscopy study of bimetallic catalysts, which covered the topics of thermally driven reconstruction dynamics and dissociation promotion of surface adsorbates9, Ni–Au bimetallic core–shell NPs with an ultrathin Au shell of 2–3 atomic layers were designed. Using ETEM, the atomic structure of the Ni–Au bimetallic catalyst during the CO2 hydrogenation reaction can be well resolved, and a remarkable loss of core–shell configuration can be observed in the range of 450–600 °C (Fig. 1a,b). Interestingly, the characteristic core–shell configuration can be observed by ex situ methods before — but also after — the reaction, pointing to a reversible phenomenon. During ETEM observation, it is challenging to extract reliable structural information about the nanocatalyst surface with atomic precision since the imaging focus fluctuates during the reaction, introducing remarkable artefacts that resemble core–shell structures. However, the image analysis script developed in this work can filter over 3,000 images to obtain identical focusing conditions and proves to be an efficient solution for tracking real-time changes to the Ni–Au surface on an atomic scale throughout the reaction. The real active surface catalysing CO2 hydrogenation is therefore determined to be an alloy of Ni and Au by ETEM, and further verified by in situ synchrotron X-ray spectroscopy and infrared spectroscopy. Theoretical simulations indicate that reaction-induced alloying on the catalyst surface arises from the significant decrease in the Ni segregation energy going from the bulk to the surface, which is driven by the adsorption of CO molecules. The CO-induced Ni surface segregation benefits the stability of the alloyed surface and is additionally responsible for the highly selective production of CO (>95%) (Fig. 1c).
The catalytic performance of materials is highly related to their surface atomic and electronic structures. In most cases, however, it is assumed that the structure of a working catalyst retains most of its initial state features during reactions, especially for cases where structures of the spent and fresh catalyst are identical. Limiting the observation of a catalyst to its structural features before and after the reaction may therefore result in overlooking important aspects of the process. Because of harsh environment changes during the reaction, the structure — especially the surface structure of the catalyst — may be changed by surface reconstruction, molecule absorption, mass transfer, or phase transformation phenomena, among others. The difficulties associated with in situ characterization methods at atomic level under realistic reaction conditions restrict our understanding of the dynamic nature of catalytic surfaces, which in turn may hinder the design and synthesis of highly efficient and low-cost catalysts. The discovery of the transformation occurring for the Ni–Au bimetallic catalyst not only provides deep insights into the reaction mechanism of this specific catalyst, but also advances the scope of in situ characterization methodologies for the structure of materials during reactions. Similar phenomena can also be expected for other bimetallic NP systems, as well as transition metal–semiconductor heterostructures. Both the loss of core–shell configuration for core–shell structured NPs/heterostructures, or the formation of core–shell configuration for alloyed bimetallic NPs, are possible under appropriate reaction conditions, driven by dynamic atomic diffusion. For example, in the case of Au/Ni2P core–shell composites, it has recently been found that Au yolks can diffuse and be stabilized within the matrix of Ni2P shells — in the form of Au single atoms and tiny clusters — leading to a highly active catalyst for the oxygen evolution reaction10. The combination of multiple state-of-the-art in situ techniques shown in this work can also be inspirational for the investigation of nucleation processes, crystal growth, structure evolution of crystals and deactivation processes of catalytic crystals. In the future, such in situ methodologies may not be limited to solid–gas reactions, but could shed light on the dynamics of liquid–solid systems too.
Hansen, P. L. et al. Science 295, 2053–2055 (2002).
Tao, F. et al. Science 322, 932–934 (2008).
Xin, H. L. et al. Nano Lett. 14, 3203–3207 (2014).
Fan, H. S. et al. Nano Energy 48, 1–9 (2018).
Wu, C. H. et al. Nat. Catal. 2, 78–85 (2019).
Tao, F. & Crozier, P. A. Chem. Rev. 116, 3487–3539 (2016).
Teschner, D. et al. Science 320, 86–89 (2008).
Zhang, X. et al. Nat. Catal. https://doi.org/10.1038/s41929-020-0440-2 (2020).
Liu, W., Sun, K. & Wang, R. M. Nanoscale 5, 5067–5072 (2013).
Cai, C., Han, S. B., Wang, Q. & Gu, M. ACS Nano 13, 8865–8871 (2019).
About this article
Advanced Materials Interfaces (2020)
Atomic-scaled surface engineering Ni-Pt nanoalloys towards enhanced catalytic efficiency for methanol oxidation reaction
Nano Research (2020)