Towards Systematic Understandings and Designs of Electrolyte/Electrode Interfaces By Computational Physics Approaches

Abstract

Reducing the amount of platinum group metals (PGMs) in catalyst layers is one of major issues in developments of polymer electrolyte fuel cells (PEFCs). Obviously, the mass activity of oxygen reduction reaction (ORR) needs to be enhanced to reduce the PGM-loading. Specific surface area of PGMs also needs to be large enough to decrease the local oxygen transport loss near the catalyst surface [1]. In addition, the catalysts need to be stable enough to ensure long-term operations of PEFCs [2]. To satisfy all these requirements, extensive studies have been carried out on explorations of optimal composition, size and shape of PGM nanoparticles [3] as well as desired controls of catalyst and ionomer distributions [4,5], and remarkable progresses have been achieved in past decades. In those studies, atomistic simulations have made significant contributions on revealing mechanisms underlying experimental observations and designing new materials. Electronic structure calculations adopting density functional theory (DFT), for example, have been utilized as a standard tool to predict the catalytic activity since the first successful reproduction of the volcano plot of ORR [6]. Molecular dynamics (MD) simulations using semi-empirical force fields have been successfully utilized to predict structures and transport properties of ionomer, too [7-9]. In this presentation, we will show results of our applications of DFT and MD simulations to cathode catalysts and ionomer thin films on catalyst surfaces [8,10]. After discussing the power of the conventional methods as well as their limitations, we will discuss recent progresses in state-of-the-art machine-learning methodologies [11-14] as a promising theoretical tool to examine complex finite-temperature properties of electrode/electrolyte interfaces.[1] A. Kongkanand et al. J. Phys. Chem. Lett. 7, 1127 (2016).[2] P. J. Ferreira et al. J. Electrochem. Soc. 152, A2256 (2005).[3] L. Pan et al. Current Opinion in Electrochem., 17, 61 (2019).[4] S. Lister and G. McLean J. Power Sources 130, 61 (2004).[5] S. Ott et al. Nature Mateir. 19, 77 (2019).[6] Nørskov et al. J. Phys. Chem. B 108, 17886 (2004).[7] S. S. Jang et al. J. Phys. Chem. B 108, 3149 (2004).[8] R. Jinnouchi et al. Electrochim. Acta 188, 767 (2016).[9] Y. Kurihara et al. J. Electrochem. Soc. 164, F628 (2017).[10] R. Jinnouchi et al. Phys. Chem. Chem. Phys. 13, 21070 (2011).[11] R. Jinnouchi and R. Asahi J. Phys. Chem. Lett. 8, 4279 (2017).[12] R. Jinnouchi et al. Phys. Rev. Lett. 122, 225701 (2019).[13] R. Jinnouchi et al. Phys. Rev. B 100, 014105 (2019).[14] R. Jinnouchi et al. Phys. Rev. B 101 060201 (2020).