Fuel cells based on the crystalline proton-conducting electrolyte CsH₂PO₄ (CDP) typically operate between 230 °C and 260 °C¹.  In this intermediate temperature regime, so-called solid acid fuel cell (SAFC) anodes are tolerant to high concentrations of impurities², are amenable to internal reforming of multiple fuels³, and may possibly omit platinum entirely⁴.  These properties make SAFCs excellent candidates for stationary fuel cell systems operating on fuels such as methanol or natural gas. 

            A longstanding barrier to the adoption of SAFC technology has been the low activity and high precious metal content of the cathode.   In the state-of-the-art cathode architecture, a macroporous framework of the CDP electrolyte is decorated with a percolating network of platinum nanoparticles that acts as both the oxygen reduction catalyst and the sole electronic conductor in the electrode.    Typical SAFC cathode Pt loadings are greater than 1.75 mg/cm², an amount circumscribed by the minimum quantity needed to obtain sufficient electrical conductivity in the electrode⁵. 

            Despite the success of the current SAFC cathode design, only limited work has been done to date to understand the structure-property relationships governing its performance.   Here we address this problem using an approach coupling ex situmicroscopic characterization and mathematical modeling with experimental input from electrochemical cell testing. 

            Porous SAFC cathodes were sectioned using FIB/SEM to render a 3D reconstruction of a representative electrode sub-volume. A machine-learning algorithm was used to automatically segment image stacks to address segmentation problems due to gray level gradients and overlapping gray level peaks in the image histogram.  Numerical analysis was performed to extract electrode physical properties (i.e. porosity, tortuosity). 

            Electrode physical parameters from the 3D electrode reconstruction were used as inputs to a 1D macrohomogenous model similar to those applied previously to PEM fuel cells⁶.  The model was used to self-consistently fit SAFC polarization curves, yielding, among other quantities, an electrochemically active area for the electrode, which has been notoriously difficult to measure in the SAFC system.   We comment on these and other findings, as well as the utility of the model in addressing the SAFC system given some of the categorical distinctions between it and the PEM system. 

Acknowledgments

This work is supported by ARPA-E via cooperative agreement DE-AR0000499.

References

¹ S.M. Haile, C.R.I. Chisholm, Kenji Sasaki, D.A. Boysen, and T. Uda, Faraday Discuss. 134, 17 (2007).

² C. R. I. Chisholm, Dane A. Boysen, A. B. Papandrew, Strahinja K Zecevic, Sukyal Cha, Kenji A Sasaki, Áron Varga, Konstantinos P. Giapis, and S.M. Haile, Electrochem. Soc. Interface 18, 53–59 (2009).

³ T. Uda, D.A. Boysen, C. R. I. Chisholm, and S.M. Haile, Electrochem. Solid-State Lett. 9, A261 (2006).

⁴ Alexander B. Papandrew, Robert W. Atkinson III, Raymond R. Unocic, and Thomas a. Zawodzinski, J. Mater. Chem. A 3, 3984–3987 (2015).

⁵ Alexander B. Papandrew, Calum R.I. Chisholm, Ramez A Elgammal, Mustafa M. Özer, and Strahinja K Zecevic, Chem. Mater. 23, 1659–1667 (2011).

⁶ T.E. Springer, T.A. Zawodzinski, and S Gottesfeld, J. Electrochem. Soc. 138, 2334 (1991).