Elsevier

Journal of Catalysis

Volume 404, December 2021, Pages 518-528
Journal of Catalysis

How the surface state of nickel/gadolinium-doped ceria cathodes influences the electrochemical performance in direct CO2 electrolysis

https://doi.org/10.1016/j.jcat.2021.10.027Get rights and content

Highlights

  • Electro-reduction of NiO to Ni is feasible upon polarization in CO2.

  • Metallic Ni is a prerequisite for CO2 electrolysis.

  • CO2 electoreduction on Ni/GDC occurs primarily at the TPB.

  • Formation of a nanometer-thick surface Ni layer over NiO is adequate to maintain CO2 electrolysis.

  • i-V characteristics of Ni/GDC electrodes in direct CO2 electrolysis are improved as compared to CO2/H2 fuel conditions.

Abstract

The effect of the surface state on the electrochemical performance of nickel gadolinium-doped ceria (NiGDC) cermet electrodes in direct CO2 electrolysis was studied by operando near ambient pressure X-ray photoelectron spectroscopy combined with on-line gas phase and electrical measurements. The CO2 electrolysis was limited at overpotentials below the carbon deposition threshold to avoid irreversible cathode deactivation. The results revealed the dynamic evolution of the NiGDC electrode surface and disclose the side reactions associated to electrode activation in CO2 electrolysis. Comparison of reduced and oxidized electrodes shows that metallic Ni is a prerequisite for CO2 electrolysis, at least at low potentials, suggesting that CO2 electoreduction occurs primarily at the three phase boundaries between gas, metallic nickel and partially reduced ceria. We also provide evidences of in situ reduction of NiO upon polarization in CO2, implying that addition of reductive gases to CO2 is not indispensable to maintain the cermet electrode in the reduced state. Inspired by this observation, we use a conventional button cell setup to demonstrate improved i-V characteristics of NiGDC electrodes in direct CO2 electrolysis as compared to CO2/H2 fuel conditions and we rationalize this behavior based on NAP-XPS results.

Introduction

CO2 is the most common environmental pollutant and is considered largely responsible for global warming through the “greenhouse effect”. In recent years, there is increasing interest to consider CO2 as resource rather than waste, and transform it catalytically to value-added chemicals and fuels [1], [2]. In order to convert CO2 to higher-value products, it should be first transformed to a more chemically active molecule, for example carbon monoxide. Electrolysis of CO2 to CO at elevated temperature using solid oxide electrolysis cells (SOEC) appears as a very promising solution for large-scale energy storage [3], [4], [5], [6], [7], [8], since it offers higher energy efficiency as compared to low temperature conversion methods. In addition, simultaneous electrolysis (co-electrolysis) of steam and CO2 is an alternative route to directly produce synthesis gas (i.e. CO and H2) from non-fossil fuel sources [9], [10].

Despite the great efforts devoted to the design and preparation of new electrocatalysts with high activity and stability, composites of ceramic and metal (cermets) that have been previously optimized for solid oxide fuel cells (SOFC), remain the state-of-the-art (SoA) cathode electrodes for SOEC as well [9], [11]. The two principal SoA cermet cathode electrodes are nickel composites with yttria stabilized zirconia (NiYSZ) or gadolinium doped ceria (NiGDC). Ni oxidation and carbon formation of Ni-based cermet cathodes during CO2 electrolysis are related to cells and stacks performance degradation [3], [9], [12]. To prevent Ni oxidation a reductive gas, such as H2 or CO, is typically mixed with CO2 [6], [10]. However, recent studies demonstrated a stable operation in direct CO2 electrolysis [13], [14], [15]. Please note that the term “direct” [16], or sometimes also called “pure” [13], or “without safe gas” [15], [17] CO2 electrolysis, refers to feed conditions where CO2 is not mixed with a reductant gas.

Carbon formation on Ni and ceria containing electrodes is attributed mainly to the Boudouard reaction (2CO → CO2 + C), while direct electrochemical carbon deposition reactions are also possible [18]. In this respect, C. Graves and coworkers demonstrated that cells with GDC cathodes can operate steadily in CO2/CO mixtures, while in similar conditions their Ni/YSZ equivalents are rapidly deactivated due to carbon deposition [5].

Although significant effort has been devoted to understand the carbon deposition mechanism during CO2 electrolysis over Ni-based cathodes [5], [12], [18], [19], [20], [21], [22], [23], [24], [25], the effect of the electrode surface state in the electrocatalytic performance is much less studied. Typically, literature considers that the active electrode area of NiYSZ is the convergence of Ni, YSZ and gaseous phase, the so-called three-phase boundaries (TPBs), while for NiGDC, the mixed ionic electronic conduction of GDC allows the reaction to take place at the two-phase boundary (2PBs) between GDC and the gas phase [3], [11]. Cells composed of porous doped-ceria fuel electrodes, i.e. without nickel, show considerable CO2 electrolysis activity and stability [5], while the primary role of 2PBs on GDC, as compared to TPB, was also proposed as well for H2O electrolysis [26] and H2 electroxidation [27] reactions. The aforementioned observations put into question the essential role of nickel in electrochemical reactions over NiGDC, which is somehow surprising considering that SoA NiGDC electrodes are composed of 65 wt% nickel.

We have previously studied the role of nickel surface state in the electrocatalytic performance of NiGDC electrodes during H2O electrolysis [26], [28], showing that partially oxidized nickel does not lead to deactivation, but on the contrary, can be even beneficial for the cell performance. The aim of the present work is to examine how nickel and ceria oxidation states affect the CO2 electrolysis performance and, in particular, to find evidences about the role of nickel in the reaction. To do so, we combine operando near ambient-pressure X-ray photoelectron spectroscopy (NAP-XPS) with online gas phase and electrical measurements over cells with porous NiGDC cathodes. It is important to underline that, contrary to the majority of CO2 electrolysis NAP-XPS studies which concentrate to the carbon deposition mechanism [5], [20], [23], [24], [29], [30], [31], [32], here we employed operational conditions bellow the carbon deposition threshold, in order to focus on the surface oxidation state and lift complications related to modifications due to irreversible carbon deposition.

Section snippets

Fabrication of cells for NAP-XPS experiments

The cell was composed of a 40 µm thick NiO/Gd0.1Ce0.9O2 (65/35 wt% ratio) cathode (fuel electrode) with mass loading of approx. 25 mg/cm2, deposited by screen printing method directly on a 300 µm thick 8YSZ electrolyte (Kerafol GmbH). A Pt layer on the reverse side, deposited by magnetron sputtering, was acting as the anode (O2 electrode). The samples were sintered in air at 1250 °C for 5 h. Due to limitations related to the NAP-XPS sample holder dimensions, fragments of button cells with

Interaction of CO2 with NiGDC without polarization

Initially, we follow the capacity of CO2 atmosphere to modify the surface oxidation state and composition of NiGDC under the specific conditions of the NAP-XPS experiment. To examine this, the NiGDC was pre-treated in 0.1 mbar H2, O2 or H2O and subsequently the gas atmosphere switched to 0.1 mbar CO2 at 700 °C. Since the effect of H2, O2 or H2O pre-treatment on the NiGDC surface state has been discussed in detail previously [35], here we focus on CO2. Fig. 1a shows Ni 2p and Ce 3d NAP-XPS

Conclusions

Summarizing, operando NAP-XPS was applied to examine NiGDC cermet cathodes during CO2 electrolysis at potentials below the threshold of carbon deposition. It was shown that the functional NiGDC surface consists of metallic nickel and partially reduced ceria. Activation of the CO2 electrolysis reaction is correlated to electro-reduction of NiO to Ni, which is feasible at relatively low CO2 electrolysis potentials. Partially reduced nickel particles are organized in a core-shell structure, with

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

DC would like to thank the China Scholarship Council (CSC) for supporting his studies at ICPEES. Y.T. Law help during BESSY synchrotron measurements is greatly appreciated. The authors would like to thank the electronic structure group and especially Dr A. Knop-Gericke, Dr M. Haevecker and Dr D. Teschner, of the Fritz-Haber Institute der MPG and Helmholtz Zentrum, Berlin Germany for providing access at ISISS beamline. The authors would also like to thank Mr. P. I. Giotakos, MSc. researcher at

References (63)

  • B. Bozzini et al.

    An in situ near-ambient pressure X-ray photoelectron spectroscopy study of CO2 reduction at Cu in a SOE cell

    J. Electroanal. Chem.

    (2017)
  • B. Bozzini et al.

    An in situ near-ambient pressure X-ray Photoelectron Spectroscopy study of Mn polarised anodically in a cell with solid oxide electrolyte

    Electrochim. Acta.

    (2015)
  • S. Hu et al.

    Cathode activation process and CO2 electroreduction mechanism on LnFeO3-δ (Ln=La, Pr and Gd) perovskite cathodes

    J. Power Sources.

    (2021)
  • S. Turczyniak et al.

    Effect of the surface state on the catalytic performance of a Co/CeO2 ethanol steam-reforming catalyst

    J. Catal.

    (2016)
  • J.J. Yeh et al.

    Atomic subshell photoionization cross sections and asymmetry parameters: 1 <Z< 103

    At. Data Nucl. Data Tables.

    (1985)
  • M.A. Henderson et al.

    Redox properties of water on the oxidized and reduced surfaces of CeO2(1 1 1)

    Surf. Sci.

    (2003)
  • M. Canavar et al.

    Effects of mesh and interconnector design on solid oxide fuel cell performance

    Int. J. Hydrogen Energy.

    (2015)
  • T. Staudt et al.

    Ceria reoxidation by CO2: A model study

    J. Catal.

    (2010)
  • A. Nenning et al.

    Excellent kinetics of single-phase Gd-doped ceria fuel electrodes in solid oxide cells

    Mater. Adv.

    (2021)
  • P.I. Giotakos et al.

    Physical modeling of the electrochemical impedance spectra for the O2 reduction reaction in HTPEM fuel cells’ cathodic electrochemical interface

    Electrochim. Acta.

    (2018)
  • A.A. Kulikovsky

    The regimes of catalyst layer operation in a fuel cell

    Electrochim. Acta.

    (2010)
  • C. Dueso et al.

    Reduction and oxidation kinetics of nickel-based oxygen-carriers for chemical-looping combustion and chemical-looping reforming

    Chem. Eng. J.

    (2012)
  • S. Zafeiratos et al.

    The interaction of oxygen with ultrathin Ni deposits on yttria-stabilized ZrO2(1 0 0)

    Surf. Sci.

    (2001)
  • S. Gunduz et al.

    Advances in High-Temperature Electrocatalytic Reduction of CO2 and H2O

    Adv. Catal.

    (2018)
  • F.S. Torknik et al.

    Effect of microstructure refinement on performance of Ni/Ce0.8Gd0.2O1.9 anodes for low temperature solid oxide fuel cell

    Ceram. Int.

    (2014)
  • E. Lay-Grindler et al.

    Degradation study by 3D reconstruction of a nickel-yttria stabilized zirconia cathode after high temperature steam electrolysis operation

    J. Power Sources.

    (2014)
  • E. Ioannidou et al.

    Au-doped Ni/GDC as an Improved Cathode Electrocatalyst for H2O Electrolysis in SOECs

    Appl. Catal. B Environ.

    (2018)
  • C. Neofytidis et al.

    Affecting the H2O electrolysis process in SOECs through modification of NiO/GDC; experimental case of Au-Mo-Ni synergy

    J. Catal.

    (2019)
  • G. Centi et al.

    Catalysis for CO2 conversion: A key technology for rapid introduction of renewable energy in the value chain of chemical industries

    Energy Environ. Sci.

    (2013)
  • Y.Y. Birdja et al.

    Advances and challenges in understanding the electrocatalytic conversion of carbon dioxide to fuels

    Nat. Energy.

    (2019)
  • Y. Song et al.

    Bao, High-Temperature CO2 Electrolysis in Solid Oxide Electrolysis Cells: Developments, Challenges, and Prospects

    Adv. Mater.

    (2019)
  • Cited by (10)

    • Ni-doped CeO<inf>2</inf> nanoparticles to promote and restore the performance of Ni/YSZ cathodes for CO<inf>2</inf> electroreduction

      2023, Applied Surface Science
      Citation Excerpt :

      The fuel electrodes were initially reduced in H2 and then the gas switched to CO2 and left to equilibrate for about 5 min. During this period the open circuit voltage (O.C.) of the two electrodes gradually decreased due to NiO formation [35]. Just before polarization the O.C. for NiYSZ and NiCe@NiYSZ cells was −200 mV and − 733 mV respectively.

    • Synthesis of Ni-doped ceria nanoparticles and their unusual surface reduction in hydrogen

      2022, Materials Today Chemistry
      Citation Excerpt :

      Ceria (CeO2) is a material commonly employed in various catalytic processes, including three-way-catalysis [1,2], NOx reduction [3], solid oxide cells [4], CO oxidation [5], water-gas-shift reaction [6], CO2 valorization [7,8] and many more [9].

    • Ceria nanoparticles as promoters of CO<inf>2</inf> electroreduction on Ni/YSZ: An efficient preparation strategy and insights into the catalytic promotion mechanism

      2022, Nano Energy
      Citation Excerpt :

      More importantly, nickel migration over ceria nanoparticles can increase the TPB and contrary to the common view, can improve the electrochemical performance. A similar Ni0-Ce3+ surface arrangement was previously found in the case of Ni-GDC cermet electrodes under CO2 and steam electrolysis reaction conditions [45,73], indicating that this is a common feature of the Ni-ceria cermet electrodes. The promotion effect of ceria on Ni/YSZ has been previously attributed to the increase of the active sites population, assuming that (reduced) ceria can activate adsorbed CO2, while, due to its mixed ionic-electronic conductivity, it allows exchange of charge carriers with the bulk [24,30,78].

    View all citing articles on Scopus
    View full text