Novel Approaches for Improving Solubility of Pyridinium Anolytes in Non-Aqueous Redox Flow Batteries

Abstract

Organic redox flow batteries (ORFBs) offer significant promise as a means of efficient grid-scale energy storage to compliment the intermittent power supply of renewable energy sources such as solar and wind power. ORFBs efficiently store energy in the form of redox-active organic molecules (ROMs) by passing them over two polarized electrodes to generate an oxidized (catholyte) and a reduced (anolyte) species, which are held in separate tanks until energy is needed. Structural diversity of anolytes and catholytes for ORFBs has expanded dramatically over the past decade as researchers have raced to discover or design ROMs suitable for use in grid-scale energy storage devices; however, a lack of correlations between ROM structure and a desired chemical or physical property has led to a “guess and check” approach that can overlook possibly transformative electrolyte structures. Recent work has demonstrated the ability to employ multivariate regression analysis to correlate computationally derived structural descriptors of ROMs to their corresponding stability. Based on this, we considered the possibility of employing a similar approach to identify quantitative structure-property relationships (QSPRs) for critical physical properties of ORFB electrolytes, such as solubility.Electrolyte solubility is an essential design characteristic in ORFBs as it directly impacts the corresponding battery’s energy storage capacity. Nevertheless, the fundamental question, “What makes an electrolyte soluble?” (especially in organic solvents) still lacks a concise answer despite decades of progress. As part of a collaboration with the Guarr group at the Michigan State University Bioeconomy Institute, our group has identified a novel class of pyridinium anolytes with exceptionally low redox potentials and highly persistent charged states. We will discuss our group’s recent progress to combine experimental and computational techniques to develop QSPR models to accurately predict solubility of pyridinium anolytes. The resulting models have provided new insights into the role that self-interaction between anolyte species in solution plays in dictating their corresponding solubility. Using these models, we have successfully expanded the solubility range for a class of pyridinium derivatives by >200% in acetonitrile.