SOLID OXIDE FUEL CELLS

Pathway for electrochemical O2 incorporation

Identifying the rate-determining step (RDS) for oxygen incorporation into mixed ionic and electronic conducting electrodes is very challenging, particularly since the local composition changes during the reaction. Now, a generally applicable method for identifying the RDS is presented, with the example of a Pr0.1Ce0.9O2–x electrode.

Electrochemically driven reactions in which both ions and electrons cross the interface between phases, involve a large variety of technologically important processes, including metal electroplating, lithium intercalation in batteries or gas incorporation in sensors and fuel cells. These mixed ion and electron transfer (MIET) reactions are far more complex than reactions that depend solely on electron transfer. One of the most extensively studied and commonly used MIET reactions is electrochemical O2 incorporation (ionization to O2–). This occurs at the solid–gas interface in oxygen sensors (there is at least one in every automobile), oxygen permeation membranes, solid oxide fuel cells and electrolysers. Over the last two decades, mixed ionic/electronic conductors containing a large concentration of oxygen vacancies, for example, doped ceria and transition metal perovskite oxides, have become prime candidates to serve as electrode materials supporting oxygen incorporation reactions (OIRs):

$${\mathrm{O}}_2\left( {{\mathrm{gas}}} \right) + 4e^ - \to 2{\mathrm{O}}^{2 - }\left( {{\mathrm{electrode}}\;{\mathrm{interior}}} \right)$$

The presence of mobile oxygen vacancies in a solid material with electronic conductivity enables OIRs to take place over the entire solid–gas interface, favourably distinguishing these materials from traditional electrodes that require an active triple-phase boundary: gas, ionic and electronic conductors. OIRs involve a series of steps: adsorption of an oxygen molecule, its dissociation, electron transfer, incorporation of the ions into the surface crystal lattice and finally, diffusion of these ions into the bulk of the electrode (Fig. 1a). It is difficult to identify the RDS for such a lengthy chain of reactions, particularly because the electrode undergoes local compositional changes during the OIR process. Among the large arsenal of techniques employed in the study of OIRs1,2,3,4,5, measuring current–voltage curves provides the most direct information concerning mass and charge transfer across the interface6,7. Interpretation of these data is challenging because voltage application alters the activity of oxygen ions (\(a_{{\mathrm{O}}_2}\)) in the electrode interior and, consequently, also at its surface. Simply put, concentrations of intermediate species participating in OIRs have been very difficult to track experimentally.

Fig. 1: Schematics of charge and mass transfer in the Pr0.1Ce0.9O2–x-based experimental fuel cell.
figure1

Adapted with permission from ref. 8, Springer Nature Ltd.

a, Schematic of the oxygen incorporation reaction at the Pr0.1Ce0.9O2–x (PCO) surface. An adsorbed oxygen molecule must first dissociate into atomic oxygen, which upon charge transfer, may be incorporated into vacant oxygen sites (\({\mathrm{V}}_{\mathrm{O}}^{ \cdot \cdot }\)). Electrons supplied by the metallic current collector replenish the surface oxygen vacancy concentration by reducing Pr4+ (\({\mathrm{Pr}}_{{\mathrm{Ce}}}^{\mathrm{x}}\)) to Pr3+ (\({\mathrm{Pr}}_{{\mathrm{Ce}}}^\prime\)). b, Schematic of the experimental setup. The cell is based on a yttria-stabilized zirconia (YSZ) solid-oxide electrolyte, with dense PCO as the working electrode (WE) and (La0.6Sr0.4)(Co0.2Fe0.8)O3 (LSCF) as the counter electrode (CE). Patterned Pt is embedded in the PCO thin film for efficient current collection.

Now, reporting in Nature Catalysis, a team of scientists from Stanford, MIT and Lawrence Berkeley National Laboratory have used Pr-doped ceria (Pr0.1Ce0.9O2–x, 0 ≤ x ≤ 0.05), a promising solid oxide fuel cell cathode material with well-characterized point defect chemistry, to demonstrate a combined experimental and analytical method for direct determination of the RDS, as well as the reaction order for each step of the OIR8 (Fig. 1a,b). The authors measured current density–overpotential curves at 450 °C or 600 °C while controlling oxygen gas partial pressure \(\left( {P_{{\mathrm{O}}_2}} \right)\) and oxygen activity in the electrode interior (\(a_{{\mathrm{O}}_2}\)). They unambiguously identified the reaction pathway based on two sets of data: one was acquired at constant \(P_{{\mathrm{O}}_2}\) with varying \(a_{{\mathrm{O}}_2}\), and the second with constant \(a_{{\mathrm{O}}_2}\) under varying \(P_{{\mathrm{O}}_2}\). The key advantage of this approach is the ability to determine oxygen activity in the electrode interior as a function of applied voltage, without the necessity of postulating that the surface is electrically neutral, and without assuming the nature of the adsorbed species and pathways.

Of the four ionic species at the surface — Pr3+, Pr4+, oxygen vacancies and oxygen ions — the concentrations of Pr3+ and Pr4+ cations were measured directly at the Pr M4,5-edge with 100 mTorr O2 ambient pressure. Measurements were taken with operando X-ray absorption spectroscopy (Fig. 1b) as a function of applied overpotential. X-ray photoelectron spectroscopy of the O 1s photoelectron binding energy was measured as a function of applied voltage, providing the authors with insight into the surface potential and an independent determination of the reaction orders with respect to \(P_{{\mathrm{O}}_2}\) and \(a_{{\mathrm{O}}_2}\) (ref. 5). Previously, researchers would have needed to treat these parameters together9,10.

Distinguishing reaction orders with respect to \(P_{{\mathrm{O}}_2}\) and \(a_{{\mathrm{O}}_2}\) is crucial for reaction pathway reconstruction. The reaction order with respect to \(P_{{\mathrm{O}}_2}\) is ~1 if the RDS involves adsorbed oxygen molecules and ~0.5 if the RDS involves adsorbed oxygen atoms. The reaction order with respect to \(a_{{\mathrm{O}}_2}\) depends on the number of participating ionic and electronic lattice defects. In this case, the overall OIR can be represented in terms of stoichiometric coefficients describing possible reaction pathways. In general, there are four stoichiometric coefficients, leading to 108 possible pathways for OIR, reflecting the number of ions and electrons that participate before, during, and after the RDS. Numerical fitting of these stoichiometric coefficients to the experimentally measured reaction orders revealed that only one RDS can explain the experimental data: dissociation of neutral molecular oxygen into neutral oxygen atoms. The most surprising implication of this finding is that electrons are not involved before or during the RDS. Therefore, neither lowering of the electron transfer barrier nor increasing the oxygen vacancy concentration would be expected to improve surface catalytic activity. Instead, the authors encourage focusing on decreasing the barrier height for O2 dissociation and for subsequent oxygen incorporation into vacancies. Considering that the kinetics of oxygen incorporation is a bottleneck in a variety of technologies, this recommendation could not come too soon.

Finally, we note that the approach presented for studying MIET8 is not limited to oxygen incorporation reactions, but is also applicable when identifying the RDS for a variety of reactions occurring on mixed conducting electrodes. These reactions may involve other ions, such as Li+, Na+, OH or H+. However, one should keep in mind that application of the proposed method requires that the concentration of the active species at the electrode surface is measured independently of the activity of external reactants. We may expect that the next application of this powerful method will be in the fields of polymer fuel cells and lithium batteries.

References

  1. 1.

    Kilner, J. A., DeSouza, R. A. & Fullarton, I. C. Solid State Ion. 86–88, 703–709 (1996).

    Article  Google Scholar 

  2. 2.

    Gopal, C. B. & Haile, S. M. J. Mater. Chem. A 2, 2405–2417 (2014).

    CAS  Article  Google Scholar 

  3. 3.

    Baumann, F. S. et al. J. Electrochem. Soc. 154, B931–B941 (2007).

    CAS  Article  Google Scholar 

  4. 4.

    Guan, Z. X., Chen, D. & Chueh, W. C. Phys. Chem. Chem. Phys. 19, 23414–23424 (2017).

    CAS  Article  Google Scholar 

  5. 5.

    Schmid, A., Rupp, G. M. & Fleig, J. Phys. Chem. Chem. Phys. 20, 12016–12026 (2018).

    CAS  Article  Google Scholar 

  6. 6.

    Mizusaki, J., Amano, K., Yamauchi, S. & Fueki, K. Solid State Ion. 22, 313–322 (1987).

    CAS  Article  Google Scholar 

  7. 7.

    Kawada, T. et al. in Solid Oxide Fuel Cells (SOFC VI) (eds Singhal, S. C. & Dokiya, M.) 396–403 (The Electrochemical Society, 1999).

  8. 8.

    Chen, D. et al. Nat. Catal. https://doi.org/10.1038/s41929-019-0401-9 (2020).

  9. 9.

    Merkle, R. & Maier, J. Phys. Chem. Chem. Phys. 4, 4140–4148 (2002).

    CAS  Article  Google Scholar 

  10. 10.

    De Souza, R. A. Phys. Chem. Chem. Phys. 8, 890–897 (2006).

    Article  Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Igor Lubomirsky.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Wachtel, E., Lubomirsky, I. Pathway for electrochemical O2 incorporation. Nat Catal 3, 94–95 (2020). https://doi.org/10.1038/s41929-020-0426-0

Download citation

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing