Katherine E. MacArthur‡
*a,
Liliana P. L. Gonçalves‡bcd,
Juliana P. S. Sousab,
O. Salomé G. P. Soarescd,
Hans Kungle,
Eva Jodate,
André Karl
e,
Marc Heggena,
Rafal E. Dunin-Borkowski
a,
Shibabrata Basak
*e,
Rüdiger-A. Eichel
efg,
Yury V. Kolen'ko
b and
M. Fernando R. Pereira
*cd
aErnst Ruska-Centre for Microscopy and Spectroscopy with Electrons, Forschungszentrum Jülich, 52425 Jülich, Germany. E-mail: k.macarthur@fz-juelich.de
bInternational Iberian Nanotechnology Laboratory (INL), Braga 4715-330, Portugal
cLSRE-LCM – Laboratory of Separation and Reaction Engineering – Laboratory of Catalysis and Materials, Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal. E-mail: fpereira@fe.up.pt
dALiCE – Associate Laboratory in Chemical Engineering, Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal
eInstitute of Energy Technologies – Fundamental Electrochemistry (IET-1), Forschungszentrum Jülich GmbH, Jülich, Germany. E-mail: s.basak@f-juelich.de
fInstitute of Physical Chemistry, RWTH Aachen University, Aachen, Germany
gFaculty of Mechanical Engineering, RWTH Aachen University, Aachen, Germany
First published on 16th July 2025
CO2 methanation offers a pathway to produce a carbon-neutral methane fuel. Although a number of research efforts have been conducted on this topic, a greater understanding of the mechanism of the reaction, which is still under debate, is needed. Here, using in situ transmission electron microscopy, we provide direct insights into the dynamics of a metallic nickel catalyst supported on activated carbon during CO2 methanation. The keys to the high performance of the catalyst are the in situ formation and dynamic behavior of a Ni@NiO core@shell nanostructure. Based on the detailed electron microscopy investigation, the mechanism of such nanostructure formation during methanation is proposed. Our studies revealed that the deactivation of the catalyst is not due to the accumulation of carbon coke over nickel nanoparticles, but an increase in the size of the nickel nanoparticles that is responsible for the deactivation of the catalyst over time.
Keywords: CO2 valorization; Hydrogenation; in situ TEM; Carbon catalysts; Core–shell nanoparticles; Microstructure.
Notably, the mechanism of CO2 methanation is still under debate. The proposed mechanisms are largely governed by the catalytic system and the reaction conditions, and the methanation could proceed through (i) the associative pathway, in which CO2 is adsorbed associatively with a former adsorbed H atom, forming formate as an intermediate, and (ii) the dissociative pathway, where CO2 is dissociated in CO, which is the main intermediate of the reaction.4 The mechanism behind nickel-catalyzed CO2 methanation involves the dissociation of H2 over the Ni nanoparticles, while CO2 adsorption usually occurs on the supporting material or nickel–support interface with dissociation and subsequent adsorption of the resultant carbonyl on Ni nanoparticles.21
A comprehensive understanding of the behavior of the supported Ni catalyst under reaction conditions, coupled with a thorough elucidation of the associated reaction mechanism, is imperative for facilitating the rational design of high-performance catalyst materials. Hence, in situ techniques are particularly important and have been extensively used to elucidate the mechanism of CO2 methanation over various supported catalytic systems. For instance, in situ infrared (IR) spectroscopy has been used for underpinning the reaction intermediates,20,22–25 while X-ray photoelectron spectroscopy (XPS) and near edge X-ray absorption fine structure (NEXAFS) have been employed to elucidate the chemical nature of the real catalysts under methanation conditions.26,27 At the same time, the fine microstructural modifications, that the supported Ni catalysts undergo during CO2 methanation, have been studied to a less extent. For example, deactivation of the Ni catalysts is often attributed to the sintering of the Ni nanoparticles and/or coke building up on the active phase, which has been investigated by ex situ transmission electron microscopy (TEM).4,28 Importantly, in situ observation of fine microstructural modifications should assist in understanding and further developing active, selective and stable methanation catalysts. Hence, in this study, in situ TEM is applied to elucidate the behavior of the Ni catalyst supported on the reduced activated carbon during CO2 methanation, affording insight into the presently lacking understanding of the catalyst’s real microstructure and its dynamics.
The understanding and atomic level information provided by aberration-corrected TEM is unparalleled, particularly for the study of catalyst nanoparticles. The development of a specialized microelectromechanical system (MEMS)-based sample holder allows us to contain gas near the sample at more than 1 bar pressure with precise control over temperature, as well as gas flow and gas compositions.
In this study, we have utilized a MEMS-based nanoreactor in which two Si/SiNx chips are positioned on top of each other and sealed with an O-ring, allowing for a maximum gas pressure of up to 2 bar. A 5 μm spacer on the bottom chip guides the gas mixture from the gas supply system towards the microheater. The system can be heated up to 1000 °C with small power consumption (in the range of mW), thus offering a low drift, while the 30 nm thin electron transparent SiNx windows, that are present in these chips, allow transmission of electron beams, thus affording structural characterization of the samples under realistic reaction conditions.
To exactly replicate the reaction conditions, the drop-cast Ni/ACR catalyst was first reduced under a H2 environment (4% H2 with N2 as the carrier gas at 1 bar pressure) at 400 °C, and then imaged under the reactive H2 and CO2 gases providing methanation conditions. It is found that once a temperature of 450 °C is reached under the reactive gases, the Ni nanoparticles began to dynamically change shape whilst still retaining crystallinity. Fig. 2(a–d) shows the dynamics of such a nanoparticle within the thick ACR supporting material. The arrows highlight the dynamic reduction-growth of the nanoparticle that simulates a ‘breathing’ behaviour. We believe that this ‘breathing’ behavior is the macroscopic manifestation of a dynamic equilibrium at the nanoscale. It likely involves the continuous, competitive oxidation of the nickel surface by CO2 to form NiO, and the simultaneous reduction of this oxide shell by H2. The observed shape changes reflect the local fluctuations in the rates of these opposing reactions. The corresponding breathing process can be better portrayed in Movie S1 in the ESI.† It should be pointed out that the reduction performed at 400 °C during the in situ experiments followed from our previous work on the same Ni on the ACR sample, where during the temperature programmed reduction (TPR) experiment, it was verified that even the Ni species that interact more strongly with the carbon were reduced at ca. 380 °C.3
To determine whether the dynamic behavior is impacted by the H2 pre-treatment, Fig. 3 compares in situ results where the sample was initially reduced in an H2 atmosphere to another set of experiments where the H2 pre-treatment was not used. Overall, like the nanoparticle shown in Fig. 2, Fig. 3 shows the dynamic behaviour of Ni nanoparticles towards the edge of the ACR supporting material under the dynamic conditions. The red arrows highlight a few of these changes in two particles, as examples. Interestingly, the presence of an oxide shell is clearly visible in both cases. The green dashed lines in Fig. 3a and j highlight the oxide shell. The high-resolution image in Fig. S1† shows the polycrystalline nature of the oxide layer outside the nanoparticles. The apparent absence of such an oxide layer for the particle shown in Fig. 2 is most likely due to the difficulty to resolve such a layer in the presence of the thick ACR supporting material. Notably, it is clear that the particles towards the edge of the ACR support changed more dynamically and, in some cases, appear almost molten, rolling around within an oxide shell. This is likely due to improved mass transport, allowing for more efficient diffusion of the CO2 and H2 reactant gases to the nanoparticle surface compared to particles embedded deeper within the porous support. The red arrows in Fig. 3a–i indicate one such nanoparticle. This behaviour was seen regardless of whether or not the reducing H2 pre-treatment was applied, as observed in Fig. 3a–i (the sample without H2 reduction) vs. Fig. 3j–n (similar sample, under similar conditions, but with H2 reduction pre-treatment). Primarily, we noticed that the formation of the oxide shell was not a mechanism of the deactivation of the particles. In fact, the most dynamic behaviour was seen for particles exhibiting an oxide shell, thereby suggesting a permeable structure in which the reactive gases can penetrate to reach metallic Ni active sites. Thus, we demonstrated that the observed NiO shell after CO2 methanation for 90 h (Fig. 1) is not a result of the non-reduced catalyst but rather formed under reaction conditions, as observed during the in situ TEM studies.
Using in situ XPS, Heine and co-workers29 reported the formation of NiO even at room temperature in CO2 gas alone, which requires the transfer of two electrons from metallic Ni to the CO2 molecule, forming Ni2+ due to the oxidation of Ni0 and CO due to the reduction of CO2. However, after the introduction of H2 into the system mimicking methanation conditions, NiO species were found to be reduced again to metallic Ni0. In another interesting study that uses in situ XPS and in situ NEXAFS spectroscopies, Giorgianni and co-workers26 reported a similar behavior, wherein the metallic Ni surface was found to oxidize rapidly upon exposure to CO2. Under methanation conditions (i.e., CO2 + H2), while in situ XPS revealed that Ni is predominantly in a reduced state, in situ NEXAFS indicated that the shell of Ni nanoparticles is prevalently in a Ni2+ state due to incomplete reduction. Interestingly, Mutz and co-workers30 showed that, in the presence of CO2, Ni starts to oxidize after the removal of H2 from the gas flow. When H2 is again introduced into the stream, the reduced fraction of metallic Ni was found to increase to 94%, yet having a 6% fraction of oxidized Ni. In our study, we would not expect the Ni nanoparticles to oxidize, since they are under an atmosphere rich in H2; however, CO2 should be responsible for the oxidation of the surface of the Ni nanoparticles, thus forming the NiO shell according to the following reaction: Ni + CO2 = NiO + CO. The Ni0 species can be oxidized by CO2 to form Ni2+ species, activating the CO2 as well, to form CO.31 Importantly, an experiment where only CO2 gas was used (Fig. S2†) shows that the presence of both H2 and CO2 is required for the oxide-shell formation (i.e., only surface oxidation), since in the presence of CO2 alone, no oxide shell can be seen. Therefore, the synergetic relationship between CO2 and H2 should be responsible for the formation of the Ni@NiO core–shell structure during CO2 methanation, as observed herein.
The formation of this Ni@NiO core@shell structure under reaction conditions might be one of the reasons for the good performance of this catalyst. On the one hand, the NiO formation in close contact with the supporting carbon material, where CO2 is adsorbed, leads to the activation of this molecule,25,31 while the Ni0 in the core is responsible for H2 dissociation, since the dynamic nature of the oxide shell allows for the diffusion of the H2 to the metallic Ni0 core of the nanoparticles. This way, both reactants can be in close contact, which facilitates the reaction. On the other hand, NiO might present oxygen vacancies, which are known to induce the dissociation and activation of CO2 on the surface of the catalyst.32 It has been reported before that the rate-determining step for the CO2 methanation reaction should be related to the availability of adjacent adsorbed H to hydrogenate adsorbed CO;33 thus, this proximity of CO2 activation and CO formation on the NiO shell and H2 dissociation in the Ni core might explain the good performance on the catalyst.3
After examining the dynamic behaviour of the particles under a reactive methanation environment, we then explored the mechanisms of deactivation of the catalyst. The main deactivation mechanism was observed to be related to a particle size increase. This can be observed in Movie S2,† obtained at 450 °C in the presence of both H2 and CO2. The larger particles demonstrated no dynamic changes, while the smaller particles demonstrated to be very active. Upon cooling, the particle structures stabilised to fill the oxide shell again (this can be seen in Fig. S3†), yielding a structure that matches the original ex situ characterisation (Fig. 1b). Finally, post in situ energy dispersive X-ray spectroscopy analysis in scanning TEM mode (STEM-EDX) was performed to confirm the presence of a 2–3 nm thick NiO shell (Fig. 4(a–f)). The heating and gas profile during these in situ experiments can be found in Fig. S4.†
Particle size distribution was not thoroughly studied in this work; however, since methanation is a known structure-sensitive reaction, the catalyst particle size is an important aspect. Thus, further experiments should be carried out to address this topic.
Our experimental in situ TEM findings have demonstrated that the nickel oxide shell observed on the catalytic Ni nanoparticles supported on reduced activated carbon in our previous study3 is in fact formed under reaction conditions of CO2 methanation, and it is expected to be an integral part of the reaction process. Fig. 5 presents a summary of the findings of this work. We believe that the Ni oxide shell structure, in situ formed during methanation, exhibits a permeable nature, allowing for the penetration of the reaction CO2 and H2 gases to the active metallic Ni0 core, otherwise dynamic changes under reaction conditions would not continue to be observed in the nanoparticles demonstrating a shell. We hypothesize that the Ni oxide shell acts as an enhancer of the CO2 adsorption/activation, while the Ni0 in the core is responsible for H2 dissociation, which leads to both reactants being in close proximity, facilitating the CH4 formation.35,36 Notably, we observed in situ that the main deactivation mechanism was related to a particle size.
Footnotes |
† Electronic supplementary information (ESI) available: Detailed experimental procedures, together with the additional TEM data and a video. See DOI: https://doi.org/10.1039/d5im00033e |
‡ These authors contributed equally. |
This journal is © Institute of Process Engineering of CAS 2025 |