Non-uniform metal deposition and dendrite formation in high density energy storage devices reduces the efficiency, safety, and life of batteries with metal anodes. Superconcentrated ionic liquid (IL) electrolytes (e.g. 1:1 IL:alkali ion) coupled with anode preconditioning at more negative potentials can completely mitigate these issues, and therefore revolutionize high density energy storage devices. However, the mechanisms by which very high salt concentration and preconditioning potential enable uniform metal deposition and prevent dendrite formation at the metal anode during cycling are poorly understood, and therefore not optimized. Here, we use atomic-force microscopy and molecular dynamics simulations to unravel the influence of these factors on the interface chemistry in a sodium electrolyte, demonstrating how a molten salt like structure at the electrode surface results in dendrite free metal cycling at higher rates. Such a structure will support the formation of a more favorable solid electrolyte interphase (SEI) accepted as being a critical factor in stable battery cycling. This new understanding will enable engineering of efficient anode electrodes by tuning interfacial nanostructure via salt concentration and high voltage preconditioning. File list (2) download file view on ChemRxiv supplementary.pdf (5.03 MiB) download file view on ChemRxiv Manuscript-V1.pdf (2.06 MiB)

We have previously reported that water addition (∼1000 ppm) to an N-methyl-N-propylpyrrolidinium bis(fluorosulfonyl)imide (C 3 mpyrFSI) superconcentrated ionic liquid electrolyte (50 mol % NaFSI) promoted the formation of a favorable solid electrolyte interphase (SEI) and resulted in enhanced cycling stability. This study reports the characterization of Na−metal anode surfaces cycled with these electrolytes containing different water concentrations (up to 5000 ppm). Morphological and spectroscopic characterization showed that water addition greatly influences the formation of the SEI and that ∼1000 ppm of water promoted the formation of an active and more uniform deposit, with larger quantities of SEI species (S, O, F, and N) present. Water addition to the electrolyte system is also proposed to promote the formation of a new complex between the FSI anions, water molecules, and sodium cations as components of the SEI. For both dry and wet (∼1000 ppm) electrolytes, the SEIs were mainly composed of NaF, metal oxide (i.e., Na 2 O), and the complex, suggested to be Na 2 [SO 3 -N-SO 2 F]•nH 2 O (n = 0−2). Postcycling SEM analysis of the Na−metal electrodes after extensive cycling (500 cycles, 1.0 mA•cm −2 , 1.0 mA•.cm −2 ) was used to estimate the minimal average cycling efficiency (ACE), which was enhanced by water addition: up to ∼99% for the 1000 ppm cell compared to ∼98% for the dry cell. Two distinct deposit morphologies, a microporous and a compact layer deposit, were evident after extended cycling in the wet and dry electrolytes. The presence of both the microporous and compact layer deposits on Na−metal surfaces cycled with the wet electrolyte, along with the distinct chemistry and morphology of the SEI, all contributed to a more stable symmetric cell voltage profile and lower cell polarization. In contrast, a higher fraction of microporous deposits and the absence of compact layer formation in the dry electrolyte were associated with higher cell polarization potentials and the occurrence of dendrites.

The effect of water on the properties of superconcentrated sodium salt solutions in ionic liquids (ILs) was investigated to design electrolytes for sodium battery applications with water as an additive. Water was added to a 50 mol % solution of NaFSI [FSI=bis(fluorosulfonyl)imide] in the ionic liquid N‐methyl‐N‐propylpyrrolidinium bis(fluorosulfonyl)imide (C3mpyrFSI). Although the thermal properties (e.g., glass transition temperature) showed little dependence on the water content, the viscosity and, in particular, the ionic conductivity were strongly affected. The Na|Na symmetrical cell cycling performance was strongly dependent on the applied current density as well as on the water content. At higher current densities (1.0 mA cm−2) the polarization profiles showed a water dependence, suggesting that water was actively involved in the formation of an improved solid electrolyte interface layer (SEI) for high‐water‐content samples (1000–5000 ppm), resulting in improved long‐term cycling stability. The initial impedance of cells cycled at 1.0 mA cm−2 (measured after 20 cycles) was elevated after water addition, and large polarizations occured for the “wet” samples. However, with further cycling the wet cells began to exhibit lower polarization and improved stability compared to the “dry” sample. The Na|NaFePO4 cell cycling performance was also demonstrated with minimal effect on the cell capacity, further highlighting the negligible activity of water in these electrolyte systems. In fact, reduced cell polarization and a more clearly defined charge profile were evident after water addition. The work shown here suggests that water may be used as a convenient and inexpensive additive for superconcentrated sodium IL electrolytes for improved device performance.

We have investigated the sodium electrochemistry and the evolution and chemistry of the solid–electrolyte interphase (SEI) upon cycling Na metal electrodes in two ionic liquid (IL) electrolytes. The effect of the IL cation chemistry was determined by examining the behavior of a phosphonium IL (P111i4FSI) in comparison to its pyrrolidinium-based counterpart (C3mpyrFSI) at near-saturated NaFSI salt concentrations (superconcentrated ILs) in their dry state and with water additive. The differences in their physical properties are reported, with the P111i4FSI system having a lower viscosity, higher conductivity, and higher ionicity in comparison to the C3mpyrFSI-based electrolyte, although the addition of 1000 ppm (0.1 wt %) of water had a more dramatic effect on these properties in the latter case. Despite these differences, there was little effect in the ability to sustain stable cycling at moderate current densities and capacities (being nearly identical at 1 mA cm–2 and 1 mAh cm–2). However, the IL based on the phosphonium cation is shown to support more demanding cycling with high stability (up to 4 mAh cm–2 at 1, 2, and 4 mA cm–2 current density), whereas C3mpyrFSI rapidly failed (at 1 mA cm–2 /4 mAh cm–2). The SEI was characterized ex situ using solid-state 23Na NMR, XPS, and SEM and showed that the presence of a Na complex, identified in our previous work on C3mpyrFSI to correlate with stable, dendrite-free Na metal cycling, was also more prominent and coexisted with a NaF-rich surface. The results here represent a significant breakthrough in the development of high-capacity Na metal anodes, clearly demonstrating the superior performance and stability of the P111i4FSI electrolyte, even after the addition of water (up to 1000 ppm (0.1 wt %)), and show great promise to enable future higher-temperature (50 °C) Na-metal-based batteries.

No abstract