Erratum: Extending the Lifetime of Organic Flow Batteries via Redox State Management (J. Am. Chem. Soc. (2019) 141: 20 (8014-8019) DOI: 10.1021/jacs.8b13295)

The additional data presented herein is intended to revisit a previous claim based upon Figure 4 of the original publication. More extensive trials now expose a clear trend between the restriction of negolyte state of charge (SOC) and the capacity fade rate of the DHAQ/Fe(CN)6 flow battery. The authors now consider the fade rate of 0.14%/day claimed in the main text to be an outlier, as depicted by the red cross in Figure 1 below, and the data in Figure S23 of the original Supporting Information to be more representative of the achievable fade rate with an 88% SOC restriction. Upon further investigation, it was determined that cell cycling history is the most likely source of the discrepancy. The cell and electrolyte used in the experiment shown in Figure 4 of the original publication had undergone prior cycling, thereby accumulating the DHA decomposition product and slowing the rate of further decomposition; in contrast, the data for Figure S23 come from a pristine cell and electrolyte. This consideration was overlooked in the original analysis and can be an important direction for future study. The overall conclusions of the original study remain unchanged. The present data affect only the degree to which the SOC restriction strategy minimizes the formation of the anthrone decomposition product of the fresh DHAQ negolyte. More specifically, the data herein indicate that a 10-fold (from 4.5 to 0.49%/day at 80% SOC restriction) or 15-fold (from 4.5 to 0.29%/day at 50% SOC restriction) decrease in capacity loss rate can be reproducibly achieved through SOC restriction alone, but not a 40-fold decrease (from 5.6 to 0.14%/day), as originally published. In addition, the aggregate data in Figure 1, collected under various conditions over an entire year, indicate the robustness of the trend to seasonal variability (∼5 °C). Although most of the data in Figure 1 were obtained with a Coulombic capacity cutoff as in Figure S23, a few of the cell runs at 50% and 80% SOC restriction were obtained with a potential-controlled SOC cutoff similar to that used in Figure 4 of the original publication. Within error, these different cutoff conditions produce indistinguishable results, thereby showing that the type of cutoff condition is not the cause of the original discrepancy. In addition, the aggregate data contain cell runs at two different electrolyte concentrations, which suggests that the lifetime extension strategy is applicable to practical battery concentrations.