Mapping heterogeneous ageing inside operational hydrogen fuel cells

19-08-2021

Imaging catalyst degradation with diffraction tomography inside large, functional fuel cells reveals high spatial heterogeneity linked to chemical gradients in pH, hydration and temperature inside the cell. While absent in model systems, these effects are critical for understanding device performance.

  • Share

Hydrogen fuel cells, devices which convert the energy stored in hydrogen gas to electricity, require large quantities of expensive platinum nanoparticle catalysts to electrochemically reduce O2 to water. The durability of this catalyst is currently a limiting factor in consumer automotive applications and the emerging hydrogen economy.

Recent work from beamline ID31 has contributed to a rising consensus that heterogeneous chemical microenvironments are both ubiquitous inside electrochemical systems, and play a significant role in determining the performance and durability of devices [1]. X-ray diffraction and small-angle scattering computed tomography (XRD-CT/SAXS-CT) inside a full-size, operational fuel cell show that electrocatalyst degradation phenomena such as corrosion, Ostwald ripening and nanoparticle aggregation are tightly correlated with the distribution of liquid water inside the device. This water distribution depends on the operating conditions of the cell and the larger-scale engineering geometry of the device, which define the gradients of chemical and physical quantities affecting the performance and degradation far more than previously anticipated. This heterogeneous environment leads to large differences in ageing throughout the electrode, seen in Figure 1.

 

drnec_Fig1.jpg

Fig. 1: XRD-CT maps of Pt nanoparticle electrocatalyst in a fuel-cell cathode showing low particle size heterogeneity before ageing (a), which becomes much more polydisperse after ageing (b). c) Higher-resolution maps of the aged sample show lined patterns produced by chemical gradients inside the cell. Bright-field transmission electron microscopy of the Pt catalyst before ageing (d) and after ageing (e) confirm the quantitative analysis via XRD-CT.


Furthermore, it was demonstrated that Pt catalyst degradation inside a full hydrogen fuel cell is massively accelerated versus the conventional liquid cells typically used to evaluate catalyst performance/stability in electrochemical laboratories. X-ray diffraction was able to determine both the crystallite size and quantity of Pt catalyst remaining over an accelerated ageing test of 10,000 voltage cycles, simulating the lifespan of a fuel-cell vehicle (Figure 2). The size of the Pt nanoparticles aged under conventional lab conditions (room temperature, ultrahigh purity water, etc.) increased from an initial 4.5 nm diameter to 6 nm. The same catalyst particles inside the electrode of the working device (gas-phase, 80°C, 10x lower pH) grew to ~11 nm, yielding dramatically lower active surface areas.

 

drnec_Fig2.jpg

Fig. 2: The hotter, more acidic conditions inside fuel cell devices accelerate nanoparticle catalyst ageing phenomena (corrosion, aggregation, ripening) versus conventional electrochemical testing environments. The Pt degradation is approximately five-fold faster inside a real device, as seen by XRD and electron microscopy (top). The XRD also quantitatively tracks the corrosion over time, allowing for a detailed understanding of the catalyst life cycle (bottom).


This study demonstrates that the X-ray scattering tomography techniques developed at the ESRF can be scaled up to image industrially sized fuel cell and battery samples, 5-10x larger than previously feasible. These advances leverage improvements in the brilliance provided by ESRF-EBS, but also from X-ray-compatible electrochemical sample environments [2], and advanced data-analysis strategies [3] developed in collaboration with international partners.

The data help to explain why next-generation fuel cell electrocatalysts reported in the literature have repeatedly failed to reach commercial viability, despite excellent performance in laboratory testing. The heterogeneous environment imprinted by the macroscopic design of the cell, together with the harsher conditions found inside the practical device, can have detrimental effects not seen under idealised conditions. Therefore, a more holistic understanding obtained through multimodal operando characterisation, as shown in this work, is critical in order to understand the chemistry inside complex electrochemical devices. This will guide the optimisation of each component of the cell, as well as the interaction between the components, allowing incorporation of new-generation, high-performance materials in industrial devices and bringing the hydrogen economy closer to reality.

 

Principal publication and authors
Imaging Heterogeneous Electrocatalyst Stability and Decoupling Degradation Mechanisms in Operating Hydrogen Fuel Cells, I. Martens (a,b), A. Vamvakeros (a,c,d), N. Martinez (e), R. Chattot (a), J. Pusa (a), M.V. Blanco (a), E.A. Fisher (b), T. Asset (f), S. Escribano (e), F. Micoud (e), T. Starr (g), A. Coelho (h), V. Honkimaki (a), D. Bizzotto (b), D.P. Wilkinson (i), S.D.M. Jacques (d), F.  Maillard (f), L. Dubau (f), S. Lyonnard (j), A. Morin (e), J. Drnec (a), ACS Energy Lett. 6, 2742-2749 (2021); https://doi.org/10.1021/acsenergylett.1c00718
(a) ESRF
(b) Advanced Materials and Process Engineering Laboratory, University of British Columbia, Vancouver (Canada)
(c) University College London, London (UK)
(d) Finden Limited, Abingdon (UK)
(e) Université Grenoble Alpes, CEA, Grenoble (France)
(f) Université Grenoble Alpes, Université Savoie Mont Blanc, CNRS, Grenoble INP, LEPMI, Grenoble (France)
(g) Independent researcher, St. Johann im Tirol (Austria)
(h) Coelho Software, Brisbane (Australia)
(i) Department of Chemical and Biological Engineering, University of British Columbia, Vancouver (Canada)
(j) Université Grenoble Alpes, CEA, CNRS, IRIG, SyMMES, Grenoble (France)

References
[1] I. Martens et al., J. Phys. Energy 3, 031003 (2021).
[2] I. Martens et al., J. Power Sources 437, 226906 (2019).
[3] A. Vamvakeros et al., J. Appl. Cryst. 53(6), 1531 (2020).