Improving the Performance of Flame-Retarding Sodium Bis(Oxalato)Borate in Triethyl Phosphate Electrolyte for Sodium Ion Batteries

Abstract

An ever-growing demand for energy storage has highlighted potential future resource constraints for Li-ion batteries (LIBs), in particular with respect to Li.1-3 With this in mind Na-ion batteries (NIBs) hold promise as a complementary technology to LIBs, particularly for stationary energy storage.4 To be competitive with LIBs, NIBs need to become fully competitive on cost and on capacity density.5-7 NIBs can however offer more than simply a cost competitive option to LIBs, with potential to offer improvements on both safety and sustainability. A number of cathode materials exist that can avoid many of the unanswered question around the sustainability of state-of-the-art LIB cathodes.3 However, to address safety concerns of LIBs it is impossible to ignore the electrolyte. State of the art electrolyte in LIBs, regularly LiPF6 in a mix of carbonates, is both toxic (by virtue of LiPF6 and to a limited extent ethylene carbonate) and flammable (by virtue of the electrolyte solvents). Importing the Na analogue of state-of-the-art LIB electrolytes (i.e. NaPF6 in carbonates) creates the same problems of flammability and toxicity currently seen in LIBs (albeit the toxicity of NaPF6 is lower than that of LiPF6).Recently Mogensen et al.8 proposed sodium bis(oxalato)borate (NaBOB) in trimethyl phosphate (TMP) electrolyte that overcame the flammability concerns of NaPF6 in carbonate electrolytes. In addition to non-flammability, the use of NaBOB instead of NaPF6 allowed for higher temperatures to be used and produced an electrolyte that was halogen free. Switching from TMP to triethyl phosphate (TEP) allowed for a relatively less-toxic electrolyte (NaBOB and TEP are listed by the European Chemical Agency as ‘harmful if swallowed’ in comparison to ‘harmful if swallowed, in contact with skin or inhaled’ for NaPF6 in carbonates, although limited research has been conducted on the toxicity of NaBOB and so this should be received with limited certainty) without sacrificing cell performance.9 However, hurdles remain for NaBOB in organophosphate electrolytes. Principal amongst these has been the poor solubility of NaBOB in most solvents and the difficulties transferring promising low mass-loading full cell results to higher mass-loading full cells. The low solubility of NaBOB limits the concentration of NaBOB in organophosphate electrolytes to < 0.4 M, producing electrolytes with relatively low conductivities (for NaBOB in TMP, 4.5 mS/cm, and in TEP, 5 mS/cm). N-methyl pyrrolidone (NMP) has previously been shown to increase the solubility of NaBOB when used as a cosolvent with TMP, with conductivity improved to 6.5 mS/cm without sacrificing non-flammability and promising cycling shown in low mass-loading full-cells.10 A range of alternative cosolvents were here investigated to increase the solubility of NaBOB in TEP. Of these, NMP, N,N-dimethyl formamide (DMF) and N,N-dimethyl acetamide (DMAc) were evaluated as cosolvents with TEP for near-commercial mass loading full-cells (Prussian white||hard carbon at ~ 2 mAh/cm2). To maintain non-flammable characteristics, the cosolvents had to be kept below 40% (v/v) for DMAc and DMF and below 60% (v/v) for NMP, resulting in electrolyte conductivities of 5.7, 6.9 and 6.0 mS/cm respectively. These nearly doubled the conductivity we observed for NaBOB in TEP of 3.5 mS/cm. However, all electrolytes examined showed large polarisation, suspected to be due a thick, resistive SEI. A range of strategies have been employed in order to reduce this polarisation, with particular focus on improved formation protocols (studying the effects of temperature, current density and electrochemical method of formation) and their use in concert with film forming additives. Future work will aim to understand the interplay between formation conditions, the resultant SEI formation and the cell polarization in an effort to translate promising low mass-loading full cell results to near-commercial mass-loading full cells.