Hybrid Membrane Leading to Improved Ionic Selectivity for Non-Aqueous Redox Flow Battery

Abstract

The efficient operation of the non-aqueous redox flow battery (NAqRFB) requires a selective membrane that prevents the crossover of the redox species to minimize self-discharge because the crossover-induced capacity loss remains one of the most challenging problems for this battery. At the same time, the membrane must enable high ionic conductivity (i.e. exchange of ions of supporting electrolyte) to lower the ohmic losses. The identification of a highly conductive, yet selective membrane is of paramount importance for market penetration of NAqRFBs. Previously, Nafion 115 was often chosen among the Nafion series to get a trade-off between mechanical strength, ion permeability, and ohmic resistance. The thick membranes have the advantage of higher mechanical strength and lower crossover rates. However, the use of thick membranes induces high ohmic resistance and a large amount of precious Nafion ionomer involved, leading to low performance of the battery and high material cost.This cost and conductivity issue was addressed using porous separators, which are widely used in lithium-ion batteries and flooded lead-acid batteries, as the alternative membrane in RFBs. Unlike Nafion and hydrocarbon membranes, the porous separators have pores at the submicron-size level. These separators usually have no inherent ion-exchange sites but become conductive when soaked in organic solvents because of the formation of an interfacing layer consisting of solvent molecules. The high ionic conductivity combined with remarkably low cost makes it promising for RFB applications. However, there are two main disadvantages in the application of porous separators for the RFB systems. The first one is their inability to absorb the effect of hydraulic permeability, which means that even a small pressure difference between two sides will result in a substantial convective flow causing the volumetric imbalance of electrolyte between the reservoirs. The second one is the lack of selectivity because of the non-existence of any ion-exchange sites. These are the main reasons that porous separators show lower coulumbic efficiency and cycling performance than ion-exchanging membranes like Nafion.Interestingly, these disadvantages can be overcome by the incorporation of a thin ion-exchange (IE) layer sandwiched between microporous separators. A proof-of-concept hybrid membrane for the non-aqueous redox flow battery is introduced in this study. This sandwiching allows us to use the thinnest reinforced IE membrane (like Nafion HP) because it is getting mechanical strength from the layers of micro-porous separators (Celgard 4560 & Celgard 2500). The relationship between the membrane conductivity, redox species crossover, and selectivity was explored to determine the applicability of hybrid membranes for non-aqueous redox flow batteries. Vanadium acetylacetonate-based electrolyte is used as a model to assess the performance of membranes. The hybrid membrane C45-HP-C45 performed best during the battery cycling and C25-HP-C25 allowed to achieve a high limiting current density of 98 mA.cm-2 from the cell which corresponds to the high discharge response CRATE = 13C as compared to the highest achieved for aqueous VRFB (i.e. CRATE = 8.5C). The ideal trend of energy efficiency as a function of selectivity highlights the margin for improvement that exists for the performance of hybrid membranes. Future approaches can be to fabricate the hybrid membrane having a more compact combination of the sandwiched layers either by using any chemical binder or by applying a pressure-based technique. This work is still in progress for in-depth study covering other aspects as well. Figure 1