Understanding Capacity Fade in Organic Redox-Flow Batteries by Combining Spectroscopy with Statistical Inference Techniques

Sanat Vibhas Modak,Wanggang Shen,Dylan Herrera,Bryan R. Goldsmith,Xun Huan,David G. Kwabi
DOI: https://doi.org/10.1149/ma2023-013760mtgabs
2023-08-28
ECS Meeting Abstracts
Abstract:Aqueous-soluble organic and organometallic redox-active molecules show considerable promise as redox-flow battery (RFB) reactants because their costs at scale are projected to be low. However, most aqueous organic RFB undergo rapid degradation resulting in high rates of capacity fade. Considerable research has been devoted towards discovering chemically stable charge carriers for RFBs. However, the decomposition mechanisms of underperforming chemistries during flow cell cycling are often poorly understood, and in some instances, competing or incompatible hypotheses about the relationship between reactant decay and capacity fade have often been proposed. In this talk, we will demonstrate that combining spectroscopic and statistical inference techniques can be used to shed light on relationship between RFB reactant degradation and capacity fade. Specifically, we employ in operando ultraviolet-visible spectrophotometry along with Bayesian inference, Bayesian model selection, and multivariate curve regression in order to quantify the relative probabilities of different decay mechanisms given the data at hand, and to estimate rate constants for specific reactions within the most probable decay scheme. In principle, rate constants thus obtained can be fed into a zero-dimensional electrochemical flow cell model, recently developed by our group, that can simulate capacity fade given a reactant decay mechanism, associated rates, and the cycling protocol. This work has the potential to shed light on the relationship between capacity fade and reactant decay in various organic RFB chemistries, and to enable prediction of the lifetimes of commercial-scale RFBs.
What problem does this paper attempt to address?