No abstract

For redox active organic molecules (ROMs) used in grid-scale energy storage applications, such as redox flow batteries, solubility is an essential physicochemical property. Specifically, solubility is directly proportional to the volumetric energy density of an energy storage device and thus affects its corresponding spatial footprint. Recently pyridiniums have been introduced as a class of ROMs with high persistence in multiple redox states at low potentials. Unfortunately, solubility of pyridinium salts in non-aqueous media remains low (generally less than 1 M), and relatively few practical molecular design strategies exist for generalized improvement of ROM solubility. Herein, we convey the extent to which discrete, attractive interactions between C-H groups and the p-electrons of an aromatic ring (C-H···pi interactions) can describe the solubility of N-substituted pyridinium salts in a non-aqueous solvent (acetonitrile). We find a direct correlation between the extent of crystalline C-H···pi interactions for each pyridinium salt and its solubility in acetonitrile (R2 = 0.93, solubility range = 0.3 – 2.1 M). The presence of C-H···pi interactions reveals how large disparities in solubility between (e.g.) N-(p-tolyl)-4-phenyl-2,6-dimethylpyridinium (0.32 ± 0.03 M) and N-(p-tolyl)-4-(p-tolyl)-2,6-dimethylpyridinium (1.06 ± 0.03 M) tetrafluoroborate may arise despite differing in structure by only three atoms. The correlation presented in this work highlights a surprising consequence of disrupting strong electrostatic interactions with weak dispersion interactions, showing how minimal structural change can have dramatic effects on ROM solubility.

Redox flow batteries (RFBs) are a strong candidate for grid-scale energy storage applications. Recent pursuits for chemical systems involve focus on organic species, due to their chemical abundance, and non-aqueous solvent systems, due to an expanded electrochemical stability window. Currently, RFBs suffer from limitations that prevent them from being economically competitive when scaled. Among the critical properties hindering RFB expansion, one limitation is low system energy densities. The energy density of a RFB system is dependent on voltage, electrons transferred, and concentration of the anolyte and catholyte. Electrolyte advancements have focused on optimizing energy density by targeting species that maximize divergence of anolyte/catholyte redox potentials, increase species solubility, and feature reversible multi-electron transfer. Strategic structural engineering of redox-active materials is necessary to tune these distinct qualities. Understanding the relationship between molecular design and these variables, and then developing strategies to predict structures with optimal characteristics could help identify promising electrolyte candidates. Our work is focused on understanding and predicting the solubility trends of pyridinium anolyte materials. The pyridiniums in this series all feature low reduction potentials and a broad range of solubility in acetonitrile. By carefully aligning experimental data to DFT and other modeled parameters we are investigating the predictive parameters involved in controlling electrolyte solubility.

While the use of nonaqueous solvents in redox flow batteries (RFBs) offers the promise of higher cell voltages than can typically be obtained in aqueous electrolytes, suitable compounds that are sufficiently soluble and stable to permit extended operation has proven challenging. Viologen anolytes have been successfully employed in aqueous systems, but their first reduction occurs at very modest potentials, thus limiting their advantage in nonaqueous systems. We have previously reported flow battery chemistry employing a series of extended bis(pyridinium) species with reduction potentials ca. 300-400 mV more negative than viologens and very long-lived, durable reduced states. However, these materials were difficult to access via common synthetic routes and often exhibited limited solubility. In this presentation, we describe an extension of those studies to new pyridinium compounds that provide still more negative reduction potentials (ca. -1.6 to -1.7 V vs Fc/Fc+). Moreover, these compounds are readily synthesized in good yield from inexpensive raw materials, are highly soluble, display excellent electrochemical kinetics, and are extremely persistent in the reduced state. The synthesis and electrochemical characterization of compounds incorporating these groups will be presented, along with preliminary RFB cycling results.