Organic redox flow batteries (RFBs) have emerged as promising energy storage systems linked to intermittent renewable energy sources. The organic redox-active molecules serve as the negolyte and posolyte (active negative and positive active electrolyte, respectively) and have developed to tune redox potential, capacity, and stability. This RFB offers potentially lower costs than vanadium RFBs due to the rising vanadium expense. Nonetheless, the capacity of organic RFBs has been limited by 0.5~2 M due to the limited solubility of organic molecules and increased viscosity, resulting in restricted energy density.To mitigate this challenge, the idea of loading solid-state charge storage into the electrolyte reservoir of RFBs has been suggested. This concept establishing redox targeting flow battery (RTFB) is, therefore, composed of soluble active molecules and storage, indicated as redox mediator (RM) and solid-state redox target (RT), respectively. The RM is responsible for the electrochemical redox event at the electrode and flows to the reservoir, where spontaneous electron transfers occur from RM to RT. Subsequently, the RM is recovered to the original state and undergoes the electrochemical redox process again at the electrode. Ideally, the capacity of RTFB relies on the electron storage ability of RT rather than the solubility of RM. Several previous reports have used polymeric materials or LiFePO4 as the insoluble RTs. They focused on the one electron-transfer process and an accessible alignment of energy levels and electron-transfer kinetics between RT and RM for reversible reactions. However, two electron-transfer processes have not yet been studied, although these systems enhance the energy density further in RTFBs.Here, we present two electron-transfer processes using naphthalene diimide (NDI) utilized as both RM and RT. We previously developed stable NDI derivatives as the negolyte in RFBs thanks to their extended-conjugation structure. In the RTFB system, we utilized these soluble NDI as the mobile RM and loaded a covalent organic framework (COF) involving NDI as the insoluble storage. Well-defined nanoporous channels with in crystalline COF provide the flow route of the soluble RM of NDI so that, ideally, all NDI included in COF (NDI-COF) can be involved in the electron-transfer process from the RM. Cyclic voltammetry of NDI-COF exhibited two cathodic peaks at –0.48 V and –0.68 V vs. NHE. Accordingly, we prepared a couple of NDI designs as the RM with different bulkiness and potential gaps from NDI-COF by introducing electron-donating and -withdrawing groups. The state of charge (SoC) of RT was inferred from the reduced state of RM in this solution, which was analyzed using open circuit potential and UV-visible spectroscopy. As a result, we observed that the NDI RM's long alkyl chain and slightly positive cathodic potential led to slow electron transfer to RM through the first cathodic reaction. Further, the second cathodic event was far slower, seemingly holding the SoC at 50% (i.e., terminating only for the first cathodic process) despite the complete two cathodic processes of RMs during the galvanostatic performance. To mitigate this challenge, we introduced RMs with more negative cathodic potential and developed RTs deposited on entangled carbon nanotubes to ensure the space, allowing the approach of RMs to the NDI in COFs, and increasing the electron-transfer efficiency. In this presentation, I will discuss details of the effectiveness of two-step electron-transfer processes using NDI materials and the related capacity enhancement. Figure 1
Read full abstract