Publication date: 10th April 2024
Overpotentials in porous electrodes lead to undesirable behavior during battery cycling and understanding the origins of these overpotentials is crucial. In Li-ion batteries, a significant fraction of the electrode overpotentials in layered cathodes arise from low-frequency sources. These low-frequency overpotential sources and the impedances they lead to are traditionally analyzed in the time domain using galvanostatic intermittent titration technique (GITT) or the frequency domain using electrochemical impedance spectroscopy (EIS). Currently in the standard porous electrode model, solid-state diffusion of Li in the cathode material is the only mechanism that is considered to lead to the electrochemical behavior seen at low frequencies and longer timescales. However, there are various experimental inconsistencies that emerge when attributing voltage relaxation behavior and low-frequency impedance behavior solely to solid-state Li diffusion. Thus, it is necessary to further investigate and understand the true nature of the impedances that arise in these timescale and frequency regimes.
In this work, we investigate the origins of the voltage relaxation behavior and low-frequency impedance in porous electrodes. Using a 3-electrode coin cell system, we vary the kinetics of the electrode/electrolyte interface by systematically changing the salt concentration of the electrolyte used. We perform GITT and EIS measurements and find that the relaxation behavior and low-frequency impedance changes as the kinetics of the electrode/electrolyte interface changes. These results contradict the experimental dependencies that are expected to arise from a system that is only diffusion controlled. Instead of a solid-state diffusion limitation, we propose an alternative mechanism that could also contribute to the observed GITT and EIS behavior: a distribution of reaction constants. We demonstrate the validity of this hypothesis with a multi-particle simulation. Furthermore, we characterize the hierarchical morphology of porous electrodes through particle physisorption and nano-computed tomography and investigate how their complex structures could lead to inherently broad distributions of electrochemical reaction times.
We would like to thank the Toyota Research Institute for funding this work. Additional thanks to Peter Csernica, Emma Kaeli, and Emma C. Choy for discussion.