A 45.6 m NaCTFSI/NaFSI hybrid electrolyte for high-voltage aqueous sodium-ion batteries operable at subzero temperatures Authors: Thileep Kumar Kumaresan, Kyunglim Pyo, Joon Seop Kwak, Woon-Bae Park, Hyeon Su No, Choi Yujin, Myoungho Pyo, Jung Yong Seo, Kee-Sun Sohn Published: 2026 ABSTRACT Aqueous sodium-ion batteries (SIBs) offer a safe and sustainable energy storage option, but their performance is restricted by the narrow electrochemical stability window (ESW) of water and poor low-temperature operability of conventional water-in-salt electrolytes (WISEs). Here, we show that sodium cyano(trifluoromethanesulfonyl)imide (NaCTFSI) addresses these issues. A 17.6 m NaCTFSI solution forms a true WISE, providing an ESW of 2.78 volts and high-voltage compatibility, but crystallizes below room temperature, severely degrading performance. To overcome this, we systematically explore a eutectic-like hybrid electrolyte by mixing NaCTFSI (14.1 m) with sodium bis(fluorosulfonyl)imide (NaFSI; 31.5 m), representing the most concentrated Na+-based WISE to date, which remains liquid down to −20°C and extends the ESW to 3.07 volts. Incorporating CTFSI− anions mitigates FSI−-induced anodic corrosion, ensuring long-term cyclability. Electrochemical tests confirm that KMnC | hybrid | NaMnF full cells sustain over 500 charge-discharge cycles across a wide temperature range, maintaining high capacity even at −20°C, demonstrating a promising platform for safe and wide-climate aqueous SIBs. INTRODUCTION Rechargeable batteries using aqueous electrolytes are emerging as a promising alternative to those based on organic electrolytes for large-scale stationary energy storage systems. Because of their inherently nonflammable and nontoxic nature, aqueous batteries offer a safer and more sustainable option compared to conventional battery systems that rely on organic electrolytes (1, 2). However, the application of dilute aqueous electrolytes is constrained by their intrinsically narrow electrochemical stability window (ESW), which leads to low energy density and limited cycle life (3–5). As a potential solution to the challenges faced by aqueous batteries, the concept of water in-salt electrolytes (WISEs) was recently introduced using 21 m lithium bis(trifluoromethane sulfonyl)imide (LiTFSI). In this system, a reduction in water activity and the formation of a LiF-rich interphase substantially extended the ESW to 3.0 V—far beyond the thermodynamic stability window of water (1.23 V) (6). This groundbreaking study has spurred a wave of subsequent research into alternative WISE formulations, including inorganic salts [LiCl (7), LiNO3 (8, 9), LiSO4 (10, 11), NaNO3 (12), and NaClO4 (13, 14)], organic sulfonates [LiCF3SO3 (15), NaCF3SO3 (16, 17), and KCF3SO3 (18, 19)], various types of symmetric [LiN(SO2F)2 (20), LiN(SO2C2F5)2 (21), LiN(SO2CF3)2 (6, 22, 23), NaN(SO2CF3)2 (23), NaN(SO2F)2 (20, 24), and NaN(SO2C2F5)2 (25)], and asymmetric imide salts [LiN(SO2CF3)(SO2C2F5) (26), LiN(SO2CF3)(SO2F) (27), LiN(SO2CF3)(CN) (28), NaN(SO2CF3)(SO2F) (29), NaN(SO2CF3)(C2F5SO2) (30, 31), and KN(SO2CF3)(CN) (32)], as well as their mixtures (29, 33–38), including so-called “hydrate melts” (21, 26, 30, 31). Most of these systems exhibit higher solubility than 21 m and wider ESWs exceeding 3.0 V, attributed to the scarcity of free water molecules and the direct involvement of anions in cation coordination. As a result, WISEs with improved ESWs have enabled the integration of conventional anode/cathode combinations into full cells, such as Li4Ti5O12/LiCoO2 (26), Mo6S8/LiNi0.5Mn1.5O4 (39), NaTi2(PO4)3/Na3(VOPO4)2F (29), and perylene-3,4,9,10-tetracarboxylic diimide/KVPO4F (32). These systems exhibit nominal charge/discharge (C/D) voltages of 1.5 to 2.8 V, indicating the need for further advancements in WISE formulations to achieve even higher ESWs that can compete with state-of-the-art rechargeable batteries. In addition, most WISE systems reported to date suffer from a limited operational temperature range, presenting another critical challenge for practical deployment. Because the maximum solubility of the electrolyte is inherently temperature dependent, WISEs are prone to partial or complete crystallization at temperatures below room temperature (RT), which can ultimately lead to battery failure. Recently, several studies have addressed these limitations by demonstrating that the high rotational mobility of uncoordinated branches in asymmetric imide salts, or their mixtures, can kinetically suppress crystallization at low temperatures (27, 29, 33). We have also shown that a 32.2 m asymmetric KTFSI electrolyte containing C≡N groups [potassium cyano(trifluorometanesulfonyl)imide (KCTFSI)] exhibits a low liquidus temperature (3°C) and enables full-cell operation at low temperatures for over 300 C/D cycles (32). In this study, we first describe that NaCTFSI can be dissolved up to the WISE concentration regime, in sharp contrast to the limited solubility of NaTFSI. The deficiency of free water molecules in a highly concentrated NaCTFSI solution (17.6 m), resulting from preferential C≡N coordination and the asymmetry of the anions, leads to an extended ESW and enables stable redox reactions of Prussian blue–type anode and cathode materials. However, the excellent properties of NaCTFSI as a WISE electrolyte at RT are compromised at low temperatures due to partial salt precipitation or complete freezing of the solution. Next, we show that these limitations can be overcome by formulating a mixed electrolyte with the symmetric sodium bis(fluorosulfonyl)imide (NaFSI). Through a comprehensive exploration of the binary phase space, we identify a unique composition (14.1 m NaCTFSI/31.5 m NaFSI) that remains liquid at −20°C and exhibits suppressed corrosion despite the high content of FSI− anions. Using this hybrid electrolyte, we achieve a further extension of the ESW and demonstrate excellent long-term C/D performance at −20°C. The coordination environment and solution structure of the hybrid electrolyte are also examined through a combination of spectroscopic analyses and theoretical calculations. We believe that this work provides valuable insights into WISE systems and presents a promising avenue for the practical implementation of Na-based aqueous batteries. RESULTS Physicochemical and electrochemical properties of NaCTFSI It is well known that imide salts with asymmetric branches exhibit higher solubility in aqueous solutions than their symmetric counterparts (26, 30, 31). This trend also holds for CTFSI−-based salts, although the degree of solubility enhancement strongly depends on the counter cation. Previously, we have shown that, although the impact of asymmetry on solubility is marginal for Li salts (21 m for LiTFSI versus 22 m for LiCTFSI) (6, 32), KCTFSI is extremely soluble compared to KTFSI (1.5 m for KTFSI versus 32 m for KCTFSI) (32). Na salts exhibit an intermediate enhancement (8.4 m for NaTFSI versus 17.6 m NaCTFSI) (30, 32), highlighting that enthalpic and entropic contributions arising from cation interactions with H2O and anions in the liquid state is a crucial factor in determining the solubility of Na salts (24, 33, 40, 41). The melting points of NaTFSI and NaCTFSI, which qualitatively reflect lattice energy, are comparable, unlike the significant differences observed for K salts (32). This suggests that the energy loss associated with lattice breakdown is negligible for Na salts (fig. S1). Before assessing the potential use of NaCTFSI as a WISE in sodium-ion batteries (SIBs), we first investigated the changes in the physicochemical properties of its aqueous solutions at RT. The ionic conductivity (σion) exhibited a convex trend with increasing NaCTFSI concentration, indicating a transition in the dominant ion transport mechanism from vehicular transport at low concentrations to hopping transport at higher concentrations, associated with cluster formation (42, 43), which is characteristic of highly concentrated salt solutions (Fig. 1A) (29, 30). Specifically, σion increased sharply from 50.1 mS cm−1 at 1 m to 96.0 mS cm−1 at 4 m, followed by a continuous decline, ultimately reaching 30.0 mS cm−1 in the saturated solution (17.6 m). The initial increase in σion is attributed to the increase in the number of mobile charge carriers, whereas the subsequent decrease beyond the peak is due to progressively stronger ion-ion interactions, where contact ion pairs (CIPs) and aggregated cation-anion clusters (AGGs) become prevalent (44–46). Consistently, the activation energy (Ea) for ion transport did not directly correlate with σion; instead, it increased continuously with concentration, exhibiting a sigmoidal trend. The most pronounced increase occurred at intermediate concentrations (for impedance spectra and Arrhenius plots, see fig.S2). This behavior indicates that, although vehicular transport is not substantially hindered by ion-ion interactions in the low-concentration regime, transport via hopping within and between clusters becomes dominant at high concentrations. Such hopping-dominated transport is characteristic of superionic-like solutions, where mobile cations efficiently traverse the quasistatic anion or cluster network. Transport properties and solvation structure evolution. (A) σion and Ea and (B) dynamic viscosity and density as a function of NaCTFSI molality at RT. (C) Walden plot highlighting the superionic behavior of 17.6 m NaCTFSI. (D) FTIR spectra in the O─H stretching region and (E) relative fractions of H-bonded water species obtained by deconvolution of (D). (F) FTIR spectra of C≡N symmetric stretching and (G) O═S═O asymmetric stretching vibrations. (H) Evolution of SSIPs, CIPs, and AGGs with increasing molality, extracted from deconvolution of the TFSIS─N─CC≡N Raman band. (I) Comparison of various coordination structures and their relative solvation energies with respect to the hexahydrated SSIP. To probe the superionic-like behavior more directly, we measured the viscosity and density of NaCTFSI solutions. As shown in Fig.1B, although the density increases nearly linearly with concentration, the viscosity increases sharply in the highly concentrated regime, likely due to the formation of CIPs and AGGs. Notably, despite the elevated viscosity at 17.6 m (25.8 mPa·s), the σion remains relatively high (~30 mS cm−1; Fig. 1A), suggesting that ion transport is partially decoupled from the bulk viscosity. This decoupling is more clearly illustrated in the Walden plot, where the 17.6 m NaCTFSI solution lies slightly above the ideal KCl line, indicating partial decoupling between ionic mobility and viscosity and thus placing the electrolyte within the superionic regime (Fig.1C; for numerical data corresponding to Fig. 1, A and B, see table S1) (47, 48). Fourier transform infrared (FTIR) spectroscopy was used to probe changes in the hydrogen-bonding network of water and the coordination of C≡N and O═S═O groups with increasing NaCTFSI concentration. As expected, Fig. 1D shows a continuous decrease in the O─H stretching absorbance due to the depletion of free water with increasing salt content (for full FTIR spectra, see fig. S3). The spectra also exhibit a gradual blue shift, a typical feature of highly concentrated aqueous electrolytes (22, 32, 49), indicating that more water molecules coordinate with Na+ cations while the native hydrogen-bonding network of bulk water is progressively disrupted (50). Deconvolution of the normalized O─H stretching bands (fig. S4) and quantitative analysis of the hydrogen bond types (Fig. 1E) reveal that the fractions of water clusters (WCs) and strongly H-bonded water (SHBW) decrease, whereas those of weakly H-bonded water (WHBW) and non–H-bonded water (NHBW) increase with concentration. These results demonstrate the continuous loss of free water and the progressive breakdown of the H-bonding network as the NaCTFSI concentration increases (51–54). Changes in the coordination environment with increasing concentration are also evident in the FTIR spectra of the C≡N symmetric stretching and O═S═O asymmetric stretching bands, located at ~2200 and ~1326 cm−1, respectively (Fig. 1, F and G). These two vibrational modes exhibit markedly different degrees of spectral shift with concentration. The position of the C≡N stretching band remains nearly unchanged, suggesting that the C≡N moiety experiences strong and preferential coordination to Na+ ions even at low concentrations (32, 49, 55). In contrast, the O═S═O asymmetric stretching band shows a pronounced red shift, from 1332.5 cm−1 (1 m) to 1328.7 cm−1 (17.6 m) (49, 56). This gradual red shift indicates an increasing involvement of sulfonyl oxygen atoms in Na+ coordination with rising concentration, likely due to the progressive depletion of water molecules in the cation’s solvation shell. Such coordination weakens the O═S═O bond strength, resulting in the observed shift to lower wave numbers. Given that the simultaneous coordination of both the C≡N and O═S═O groups can promote the formation of large ionic clusters such as AGGs, we further investigated the solution structure using Raman spectroscopy. The analysis focused on the symmetric vibrational mode of the TFSIS─N─CC≡N moiety in the region of 740 to 780 cm−1, which is known to be particularly sensitive to O═S═O coordination and thus indicative of cluster formation (49). With increasing NaCTFSI concentration, a blue shift of the S─N─C peak position was observed—from 759.4 cm−1 (1 m) to 760.9 cm−1 (17.6 m)—indicating enhanced O═S═O···Na+ interactions (fig. S5) (49). Deconvolution of the S─N─C bands further revealed a significant presence of ion pairs and/or clusters (Fig. 1H) (20, 57, 58). At low concentrations, CIPs were predominant, with a minor contribution from solvent-separated ion pairs (SSIPs), reflecting the preferential coordination through the C≡N groups. As the concentration increased, the fraction of AGGs rose steadily, accompanied by a gradual decrease in both CIPs and SSIPs. In the saturated solution (17.6 m), AGGs dominated the population, comprising 82%, whereas SSIPs were no longer present (0%). To theoretically support the proposed solvation structure, we performed density functional theory (DFT) calculations to determine the solvation free energies of the most probable coordination configurations. All calculations were conducted using the B3LYP (59) hybrid exchange-correlation functional and the 6-311++G(d,p) basis set (60, 61), as implemented in the Gaussian 16 software package. The solvent environment beyond the first solvation shell was modeled using the SMD implicit solvation model (62), which treated bulk solvent as a dielectric continuum with surface tension at the solute-solvent interface. Figure 1I shows the possible configurations in which a single Na+ ion is surrounded by one CTFSI− ion and six H2O molecules. The solvation free energies of each structure—given relative to that of a hexahydrate Na+ without direct coordination by CTFSI−—are listed below the corresponding geometries. The results reveal that all CIP configurations are energetically more favorable than the SSIP. In particular, the CIP coordinated by five water molecules and one C≡N group exhibits the lowest energy (−6.69 kJ mol−1), supporting the experimentally observed preferential coordination of C≡N even at low NaCTFSI concentrations. The bidentate structure, in which both the C≡N and O═S═O moieties coordinate to Na+, was also found to be relatively stable (−5.44 kJ mol−1), although less so than the C≡N-only coordinated CIP. The ESW of aqueous NaCTFSI electrolytes was evaluated using linear sweep voltammetry (LSV). Measurements were conducted on stainless steel (SS) and Al electrodes at a scan rate of 0.1 mV s−1. The potential was scanned from the open-circuit potential (OCP), using Ag/AgCl and Pt foil as the reference and counter electrodes, respectively. The OCP remained nearly constant at −2 mV versus Ag/AgCl, regardless of the electrolyte concentration and electrode material. As shown in Fig. 2A, both the cathodic (Elc) and anodic (Ela) limits shift continuously with increasing concentration, resulting in a significant expansion of the ESW beyond the thermodynamic water-splitting limit (1.23 V). On SS, the ESW expanded from 1.71 V (Elc = −0.52 V and Ela = +1.19 V) at 1 m to 2.34 V (Elc = −1.02 V and Ela = +1.32 V) in the saturated solution. When Al was used as the working electrode, Elc shifted further in the negative direction (−1.46 V), yielding a wider ESW of 2.78 V. The enhanced cathodic stability on Al is attributed to the formation of a more compact and robust solid electrolyte interphase (SEI) on the native Al2O3 surface, which likely acts as an effective kinetic barrier to the hydrogen evolution reaction. Similar behavior has been reported in previous studies (32, 63, 64). The Elc and Ela were defined at a current density cutoff of ±10 μA cm−2. Electrochemical stability and low-temperature performance of NaCTFSI electrolytes. (A) LSV curves of NaCTFSI at various concentrations recorded on SS and Al electrodes at 0.2 mV s−1. Elc and Ela were determined by current levels smaller than ±10 μA cm−2. (B) C/D profiles of NaMnF and KMnC at 1C (144 mA g−1 for NaMnF and 78 mA g−1 for KMnC). (C) CVs of NaMnF and KMnC in 17.6 m NaCTFSI at 0.2 mV s−1, measured at RT and 10°, 0°, and −10°C. Corresponding LSV traces on SS and Al electrodes are overlaid (dashed lines). (D) DSC curves of 17.6 m NaCTFSI recorded during cooling (blue) and subsequent heating (red) between 60° and −90°C at 1°C min−1. Inset: Optical photographs of 17.6 m electrolytes after store at 25°, 15°, 0°, −10°, and −20°C for 48 hours. Next, we evaluated the compatibility of 17.6 m NaCTFSI with Prussian blue–type electrode materials. Na1.90Mn[Fe(CN)6]0.98·2.6H2O (NaMnF) and K0.10Mn[Cr(CN)6]0.70·2.0H2O (KMnC) were selected as cathode and anode materials, respectively, due to the suitability of their redox potentials within the ESW of 17.6 m NaCTFSI (13, 36). Phase purity was confirmed by x-ray diffraction (XRD), showing the crystallization of NaMnF (light gray) and KMnC (pale yellow) in the P21/n and Fm-3m space groups, respectively (figs. S6 and S7). The chemical compositions were verified using energy-dispersive x-ray spectroscopy (EDX) coupled with field-emission scanning electron microscopy (FESEM; figs. S8 and S9). To further validate the composition, the relative ratios of the metallic species were confirmed by inductively coupled plasma atomic emission spectroscopy (ICP-AES). The atomic ratios of Na and Fe relative to Mn in NaMnF (1.93 and 1.02, respectively), and K and Cr relative to Mn in KMnC (0.09 and 0.68, respectively), were consistent with those obtained from FESEM-EDX analysis within experimental error. The exact compositions of NaMnF and KMnC differed slightly from those reported previously (13, 36), indicating theoretical reversible capacities of 144 mA·hour g−1 {Na1.90Mn[Fe(CN)6]0.98 ↔ Mn[Fe(CN)6]0.98} and 78 mA·hour g−1 {K0.10Mn[Cr(CN)6]0.70 ↔ K0.10Na0.70Mn[Cr(CN)6]0.70} for NaMnF and KMnC, respectively. Figure 2B shows the C/D profiles of NaMnF and KMnC in 17.6 m NaCTFSI cycled at 1C (RT). Although the reversible capacity of the KMnC electrode closely approaches its theoretical value (77 mA·hour g−1), only ~70% of the theoretical capacity is achieved for NaMnF (101 mA·hour g−1), due to the high-voltage cutoff of 1.30 V. Increasing the upper cutoff initially improved capacities, but rapid capacity decay occurred over subsequent cycles. When cycled within −1.30 to 0.0 V for KMnC and 0.0 to 1.30 V for NaMnF, the capacity decay over 1000 cycles was less than 7%, suggesting the feasibility of constructing 1.7 V-class aqueous SIBs. The excellent electrochemical performance of NaMnF and KMnC in 17.6 m NaCTFSI at RT, however, was severely compromised at lower temperatures (Fig. 2C). Comparison of cyclic voltammograms (CVs) revealed a sudden decrease in current responses when the temperature was lowered from RT to 10°C. Although both NaMnF and KMnC exhibit well-defined redox pairs at RT, the peak currents were substantially reduced at 10°C and this trend continued with further decreases in temperatures. Because it is unlikely that this abrupt suppression of electroactivity is due to sluggish sodiation/desodiation kinetics, we attributed the phenomenon to a sudden change in the physical state of 17.6 m NaCTFSI at low temperatures. Differential scanning calorimetry (DSC) analysis of 17.6 m NaCTFSI between 60° and −90°C revealed a liquidus temperature near RT (Fig. 2D). Upon cooling, the solution exhibited two exothermic events: a small peak at 11°C, likely corresponding to partial structural ordering or precrystallization of ion clusters, and a major peak at −42°C, reflecting the onset of bulk crystallization, i.e., a phase transition from the supercooled liquid to the solid state. During subsequent heating, an endothermic peak appeared at −26°C, indicating partial melting of previously crystallized domains. Another sharp endothermic peak at 23°C was attributed to complete melting of the crystalline phase, which corresponds to the liquidus temperature of the system. Visual observation on the freezing behavior corroborated the DSC results: The solution remained liquid at RT for 48 hours, began forming significant crystallites at 10°C, and was completely frozen at −20°C. Quick crystallization without a supercooled state in a standard glass tube (contrary to the DSC results during cooling) is likely caused by external perturbations, such as mechanical vibrations or heterogeneous nucleation at the container walls, which can substantially reduce the degree of supercooling. A 45.6 m NaCTFSI/NaFSI hybrid electrolyte with viable physicochemical properties Although the NaCTFSI-based WISE exhibited favorable electrochemical performance at RT, its practical application remains limited due to crystallization at lower temperatures. To address this issue, we explored a second imide salt suitable for formulating a hybrid electrolyte with NaCTFSI (26, 29, 30). As a first attempt, we selected the symmetric NaTFSI as the second imide salt. A slight enhancement in the total Na+ concentration was observed at the eutectic composition of Na(CTFSI)0.8(TFSI)0.2·2.6H2O; however, this mixture failed to remain in the liquid state at −20°C (fig. S10). We then investigated an asymmetric/asymmetric combination using the sodium salt of (fluorosulfonyl)(trifluorometanesulfonyl)imide (NaFTFSI). The NaCTFSI/NaFTFSI mixture also displayed eutectic behavior, showing the lowest water content at the composition of Na(CTFSI)0.60(FTFSI)0.40·2.0H2O (fig. S11). Although the maximum Na+ concentration in NaFTFSI-only solution was 19 m, it increased markedly to 28.4 m upon the addition of NaCTFSI. However, as with the NaCTFSI/NaTFSI system, the NaCTFSI/NaFTFSI solution also rapidly solidified at −20°C within 12 hours. These results suggest that, although asymmetric anions can enhance configurational entropy—thereby improving salt solubility and inhibiting crystallization—such benefits only become significant at sufficiently high salt concentrations, especially at low temperatures. A previously reported mixture containing symmetric NaFSI (25 m) and asymmetric NaFTFSI (10 m), with a water content less than 2.0 [Na(FSI)0.71(FTFSI)0.29·1.59H2O], remained fully liquid at −20°C (29). We also investigated a hybrid electrolyte system composed of NaCTFSI and NaFSI by first examining the solubility of their binary compositions at RT (Fig. 3A). This analysis revealed that the hybrid electrolyte is soluble across a wide compositional range (sky blue area), except when both NaCTFSI and NaFSI are present at relatively high concentrations. The highest Na+ concentration of 47.3 m was achieved at a composition of 12.3 m NaCTFSI (α = 7) and 35 m NaFSI (β = 10). However, this eutectic mixture failed to remain in the liquid state for more than 6 hours. In contrast, a nearby composition—containing 45.6 m Na+ (α = 8 and β = 9, marked in red)—remained liquid at −20°C. All other tested compositions, including the 35 m NaFSI single-salt solution, solidified under low-temperature conditions, indicating their unsuitability for aqueous SIBs in cold environments. This particular composition exhibited exceptional stability, showing no signs of crystallization even when the container was vigorously shaken or a nucleating agent (21 nm TiO2) was introduced. The solution remained fully liquid after extended storage at −20°C for 1 month. The existence of this unique composition—distinct from the eutectic point—is intriguing, suggesting that the role of asymmetric anions in enabling low-temperature stability may differ somewhat from their role in enhancing solubility. We note that the microscopic structural origin of this enhanced supercooling stability remains to be fully elucidated and may involve kinetic frustration of ionic clusters rather than purely thermodynamic effects. The excellent low-temperature performance of the 45.6 m NaCTFSI/NaFSI electrolyte was further confirmed by repeated cycling over a temperature range (−40° to 60°C), which is relevant for practical battery applications (Fig. 3B and fig. S12). DSC conducted over five cooling/heating cycles showed no discernible thermal events, validating the superior stability of the supercooled state in the 45.6 m NaCTFSI/NaFSI electrolyte. Phase stability, interfacial behavior, and electrochemical properties of hybrid electrolytes. (A) Binary phase map, showing the solubility of NaCTFSI/NaFSI at RT and crystallization behavior at −20°C. Soluble compositions at RT are shown in sky blue, whereas frozen compositions at −20°C are denoted by the capital letter “F.” The hybrid solution at α = 8 and β = 9 is the only composition that remained unfrozen (marked “N” in red). Inset photographs illustrate the crystallization behavior over time (1 to 48 hours) for representative compositions stored at −20°C. h, hours. (B) DSC curves of the hybrid solution during repeated heating and cooling cycles between −40° and 60°C at 1°C min−1. (C) OCP variation over time on SS electrodes in 35 m NaFSI, 17.6 m NaCTFSI, and hybrid electrolytes. Inset photos show color differences of a methyl red indicator. (D) Potentiodynamic polarization curves on SS electrodes and corresponding FESEM images of SS surfaces after anodic polarization in the respective electrolytes. Scan rate = 1 mV s−1. (E) EIS spectra at various temperatures and the corresponding Arrhenius plot (inset) in the 45.6 m hybrid electrolyte. (F) LSV curves of the 45.6 m hybrid electrolyte on SS and Al electrodes at 0.2 mV s−1 (cutoff current = ±10 μA cm−2). (G) Atomic compositions relative to Al, determined by EDX analysis of Al surfaces subjected to various negative potentials. Abnormally high “O” and “C” contents, arising from the surface oxide layer (Al2O3) and surface-adsorbed CO2, are excluded. We also found that the hydrolysis of FSI− anions is suppressed considerably in the hybrid electrolyte. The electrophilic S─F bond in FSI− is highly susceptible to nucleophilic attack by H2O molecules, leading to the formation of HF (65, 66). HF is a particularly problematic acid because it not only dissolves metal oxides but also forms stable metal-fluoride complexes, resulting in continuous corrosion of passivating metals such as SS (67). Because corrosion is strongly influenced by the HF concentration and the solution pH, we compared these parameters among three systems: a 35 m NaFSI solution, a 17.6 m NaCTFSI solution, and the 45.6 m NaCTFSI/NaFSI hybrid solution (Fig. 3C). The 35 m NaFSI solution exhibited a low pH (2.0) and a high fluoride concentration {([F−] = 185 parts per million (ppm)}. In contrast, the hybrid solution showed a markedly higher pH of 4.8 and a considerably lower [F−] of 22.4 ppm. Although these values are slightly less favorable than those of the 17.6 m NaCTFSI solution (pH = 6.2; [F−] = 7.4 ppm), they clearly indicate that FSI− hydrolysis is no longer a major concern in the hybrid system. The mechanism underlying the suppression of FSI− hydrolysis upon mixing with CTFSI− remains unclear. However, we speculate that this effect may be associated with the difference in acidity between HFSI and HCTFSI. The combination of the relatively basic FSI− anion [pKb (where Kb is the base dissociation constant) of HFSI = ~12.72] (68) with the nearly neutral CTFSI− anion—considering the superacidity of HTFSI [pKa (where Ka is the acid dissociation constant) = −10] (69)—may impart a buffering effect to the hybrid solution. The suppression of FSI− hydrolysis in the hybrid electrolyte resulted in only marginal OCP changes over time (Fig. 3C). The OCP on SS in the hybrid electrolyte remained relatively stable for 60 min, decreasing slightly from 0.13 and 0.09 V. A similarly small change was observed in the 17.6 m NaCTFSI solution (0.18 V → 0.15 V), confirming the noncorrosive nature of CTFSI−-containing electrolytes. In contrast, the initially high OCP of 0.42 V in the 35 m NaFSI solution rapidly dropped to −0.02 V, likely due to the breakdown of the passivation layer and the onset of active corrosion (70). The distinct corrosion behavior of the hybrid electrolyte was further confirmed by potentiodynamic polarization measurements (Fig. 3D). These measurements were conducted over a potential range of ±250 mV relative to the corrosion potential (Ecorr) at a scan rate of 1.0 mV s−1. Among the tested electrolytes, the 17.6 m NaCTFSI solution exhibited the highest Ecorr (0.18 V) and the lowest corrosion current density (icorr = 0.21 mA cm−2), whereas the 35 m NaFSI solution showed a markedly lower Ecorr (−0.05 V) and a higher icorr (0.63 mA cm−2). The 45.6 m hybrid solution showed intermediate values (Ecorr = 0.08 V and icorr = 0.34 mA cm−2), further highlighting the beneficial effect of CTFSI− in suppressing FSI− hydrolysis. Moreover, the hybrid electrolyte exhibited intermediate Tafel slopes (βa = 567 mV dec−1 and βc = −230 mV dec−1), indicating a favorable balance between passivation layer stability and electron transfer kinetics (table S2) (71). Consistent with these electrochemical results, the surface morphologies of SS electrodes after anodic polarization in the 17.6 m NaCTFSI, 45.6 m hybrid, and 35 m NaFSI electrolytes revealed notable differences. FESEM images of SS polarized in CTFSI−-containing electrolytes showed relatively smooth and crack-free surfaces, indicative of a stable and intact passivation layer. In contrast, irregular pitting—characteristic of localized corrosion—was clearly observed on SS electrodes treated in NaFSI. The absence or mitigation of corrosive behavior in CTFSI−-containing electrolytes is not attributed to the surface adsorption of CTFSI− on SS as no detectable CTFSI− species were observed on SS electrodes immersed in NaCTFSI, according to x-ray photoelectron microscopy (XPS) studies (fig. S13). Comparable Tafel behavior and surface morphologies were also observed for Al electrodes (fig. S14 and table S3). It is worth noting that, although NaFSI was previously reported to be inert in aqueous solutions—unlike LiFSI—we did not observe such inertness in the present study (20). This discrepancy may be attributed to the initial pH adjustment performed in the earlier work. In general, when the salt concentration in WISEs exceeds a critical threshold, the σion begins to decrease while the viscosity increases exponentially (32, 72). Despite this, certain WISE systems exhibit superionic behavior in the extremely concentrated regime (30, 33, 73), likely due to the transport of loosely bound charge carriers within and between ionic clusters (46). Our 45.6 m NaCTFSI/NaFSI solution also falls into this category. It exhibited a σion of 7.8 mS cm−1 at RT—lower than that of 35 m NaFSI (12.3 mS cm−1) (34) and 17.6 m NaCTFSI (30 mS cm−1) but higher than those of other mixed Na-salt systems, such as 27.8 m NaN(SO2CF3)(SO2C2F5)/NaN(SO2C3F7)(SO2CF3) (3.3 mS cm−1) (31) and 40 m NaFSI/CH3COONa (7.1 mS cm−1) (34). Moreover, the viscosity of the 45.6 m hybrid electrolyte (110 mPa·s) is somewhat higher than that of saturated NaFSI (75.9 mPa·s at 35 m) (34) but considerably lower than those of other mixed-anion Na-WISE systems, such as 27.8 m NaN(SO2CF3)(SO2C2F5)/NaN(SO2C3F7)(SO2CF3) (293 mPa·s) (31) and 40 m NaFSI/CH3COONa (121 mPa·s) (34). Owing to its relatively high σion, low viscosity, and moderate density (1.910 g cm−3), our hybrid solution exhibited superionic behavior [log(hour−1) = −0.041; log(L) = −0.033], suggesting the presence of facile Na+ transport pathways both within and between clusters. Accordingly, the temperature-dependent measurements of σion revealed a relatively moderate variation, yielding an Ea of 0.24 eV for the hybrid electrolyte—lower than those of the saturated single-salt counterparts (0.27 eV for 17.6 m NaCTFSI and 0.33 eV for 35 m NaFSI; fig. S15). Electrochemical impedance spectroscopy (EIS) showed a gradual and consistent decrease in σion from 29.9 mS cm−1 at 60°C to 2.1 mS cm−1 at −20°C, corresponding to an Ea of 0.24 eV (Fig. 3E). Notably, the σion value at −20°C (2.1 mS cm−1)—comparable to the RT σion of some WISE systems (26, 31, 38)—suggests that the NaCTFSI/NaFSI hybrid electrolyte holds promise for subzero battery applications. Compared to the ESW of 35 m NaFSI (Elc = −1.30 V on Al, Ela = 1.44 V on SS, and ESW = 2.74 V; fig. S16), the 45.6 m NaCTFSI/NaFSI hybrid electrolyte exhibited a substantially broader ESW (Fig. 3F). The LSV curves revealed an expansion to 3.07 V (Elc = −1.58 V on Al and Ela = 1.49 V on SS), indicating that the slight positive shift in Ela (1.44 V → 1.49 V) contributes to the widening, but the primary factor is the negative shift in Elc (−1.30 V → −1.58 V). This suggests that NaCTFSI plays a beneficial role in forming a compact and robust SEI, which effectively suppresses hydrogen evolution on Al (fig. S17). This hypothesis was further supported by EDX composition analysis of Al surfaces polarized to various negative potentials (Fig. 3G and fig. S18). As the potential decreased from −1.0 to −1.3 V, the concentrations of SEI-related electrolyte components increased markedly and remained high down to −1.6 V. At more negative potentials, however, these concentrations dropped sharply, suggesting that the SEI remains protective down to ~−1.6 V, close to Elc. Coordination in the 45.6 m NaCTFSI/NaFSI hybrid electrolyte To gain insight into the solvation structure of the 45.6 m hybrid electrolyte (14.1 m NaCTFSI/31.5 m NaFSI), we performed a suite of spectroscopic analyses, including 1H and 19F nuclear magnetic resonance (NMR), as well as FTIR. For the NMR measurements, a coaxial insert containing D2O was placed inside the sample tube to provide the field lock (22). The 1H NMR spectrum of 14.1 m NaCTFSI shows the water resonance at 4.73 ppm, notably downfield relative to 4.21 ppm in 31.5 m NaFSI, indicating stronger hydrogen bonding among water molecules coordinated to Na+ ions (Fig. 4A). Upon mixing with 31.5 m NaFSI, we anticipated a large upfield shift compared with the resonance of the single-salt NaFSI solution, owing to the reduced water content in the coordination shell and thus weaker hydrogen bonding in the hybrid electrolyte. However, the observed resonance appeared at 4.28 ppm, shifted upfield from 4.73 ppm (NaCTFSI) yet still downfield relative to 4.21 ppm (NaFSI). This suggests that, in the highly concentrated hybrid solution, water molecules coordinated to Na+ ions still experience substantial hydrogen bonding with the O═S═O groups of the anions (vide infra). Solvation structure and ion coordination in hybrid electrolytes. (A) 1H NMR and (B) 19F NMR spectra of 14.1 m NaCTFSI, 31.5 m NaFSI, and 45.6 m hybrid electrolytes. (C) Comparison of FTIR spectra in the C≡N stretching region between 14.1 m NaCTFSI and 45.6 m hybrid electrolytes. (D) Possible coordination configurations and corresponding solvation energies for a single Na+ ion surrounded by one CTFSI−, two FSI−, and four H2O molecules. RDFs [G(r)] and CNs for (E) Na+···Owater, (F) Na+···N≡CCTFSI and Na+···OCTFSI, and (G) Na+···OFSI. (H) Comparison of G(r) for OFSI···Hwater and OCTFSI···Hwater. The 19F NMR spectra show resonance peaks at 53.24 ppm for 31.5 m NaFSI and −77.40 ppm for 14.1 m NaCTFSI, corresponding to the SO2F group in FSI− and the CF3 group in CTFSI−, respectively (Fig. 4B). In the hybrid electrolyte, two distinct signals appear at 53.30 and −77.15 ppm, each slightly shifted relative to those of the single-anion systems. These subtle shifts can be attributed to changes in coordination and electronic environments in the hybrid solution. For FSI−, the symmetric SO2F groups at both ends can participate in coordination, and as the water content decreases, the probability of simultaneous coordination to multiple Na+ ions increases. Consequently, in the hybrid electrolyte, more SO2F groups are on average deshielded, leading to a small downfield shift (from 53.24 to 53.30 ppm). A similar trend is observed for the CF3 groups in CTFSI−: Their incorporation into larger ionic clusters reduces shielding, giving rise to the downfield shift of the 19F peak (from −77.40 to −77.15 ppm). The strength of C≡N coordination was also modestly affected by the addition of 31.5 m NaFSI (Fig. 4C). FTIR spectra in the C≡N stretching region exhibit a slight blue shift compared to the single-salt solution. Although the characteristic C≡N band appears at 2200 cm−1 in 14.1 m NaCTFSI, it shifts to 2208 cm−1 in the hybrid electrolyte, suggesting a partial weakening of the Na+···N≡C interaction, although the coordination remains strong. This blue shift likely reflects a more dynamic solvation environment, where the slightly reduced interaction strength facilitates faster Na+ transport within and between clusters, thereby contributing to the observed superionic behavior. Next, we investigated the coordination environment using a theoretical approach to validate the spectroscopic interpretations and gain deeper insights into the solvation structure. A model system was constructed, in which a single Na+ ion is surrounded by one CTFSI−, two FSI−, and four H2O molecules, giving a H2O-to-anion molar ratio (1.33) comparable to that of the 45.6 m hybrid electrolyte (1.22). Using the Gaussian 16 software package, we calculated the solvation energies of various possible configurations, using the same methodology as in Fig. 1I. Figure 4D presents seven plausible configurations along with their relative solvation energies. The results clearly indicate that configurations lacking coordination between CTFSI− and Na+ are energetically unfavorable. The most stable configuration, with a solvation energy of −20.0 kJ mol−1, features Na+ coordinated by one CTFSI−, one FSI− (cis to CTFSI−), and four H2O molecules. The second FSI− is stabilized via H2O-mediated bridging, suggesting a role in cluster formation. This most stable configuration is also consistent with the H-bonding nature of the hybrid electrolyte, as inferred from the 1H NMR spectra. We further evaluated the solution structures using ab initio molecular dynamics (AIMD) simulations (74, 75). The simulation was carried out for a periodic cell containing 3 NaCTFSI, 6 NaFSI, and 12 H2O molecules, closely matching the composition of the 45.6 m hybrid electrolyte (NaCTFSI/NaFSI/H2O = 3.0/6.6/11.8). Detailed information on the simulation methodology and the simulation box is provided in Supplementary Text and table S4, respectively. Figure 4 (E to G) shows the pair radial distribution functions [RDFs; G(r)] and coordination numbers (CNs) of specific atoms surrounding Na+ ion. The G(r) between Na+ ions and oxygen atoms of water (Na+···Owater) exhibits a primary peak at 2.38 Å, indicating that water molecules occupy the first solvation sheath (Fig. 4E). Moreover, the peak area matches the CN of 1.33 (the ratio of Owater atoms to Na+ ions in the simulation box), suggesting that all water oxygen atoms are coordinated to Na+ ions. A similar trend is observed for Na+ ions and CTFSI− anions (Fig. 4F). The dominant peaks of nitrogen and oxygen atoms appear within the first solvation sheath (2.53 and 2.38 Å), and their areas correspond to the theoretical CN values (0.33 and 0.67). This indicate that, although Na+···N≡C coordination is energetically favored, Na+···O═S═O···Na+ coordination is unavoidable in the highly concentrated hybrid electrolyte. In contrast, the G(r) between Na+ ions and oxygen atoms of FSI− (Na+···OFSI) exhibits slightly different behavior (Fig. 4G). Although the primary peak position is same (2.38 Å), the CN calculated from this peak is slightly lower than the theoretical value of 2.67, implying that some FSI− oxygen atoms are not directly coordinated to Na+. G(r) profiles between oxygen atoms of CTFSI− and FSI− and the hydrogen atoms of water (OCTFSI···Hwater and OFSI···Hwater, respectively) reveal a clear distinction (Fig. 4H). No characteristic H-bonding feature is observed for OCTFSI···Hwater, whereas OFSI···Hwater displays a peak at ~2 Å, corresponding to a typical H-bond distance (1.7 to 2.2 Å) (76). This result clearly indicates that some FSI− anions are bridged to Na+ ions through H2O molecules, which is consistent with the results shown in Fig. 4 (A and D). Despite this consistency, we note that the relatively small simulation cell limits the description of long-range clustering and extended solvation structures; therefore, the AIMD results are primarily intended to elucidate the local first-solvation-shell coordination environment. Coordination involving other atoms, such as peripheral F or the central N of the anions, is relatively negligible (fig. S19). Electrochemical performance of full cells at subzero temperature Last, we evaluated the electrochemical performance of aqueous cells using the 45.6 m hybrid electrolyte, anticipating higher capacity and superior stability due to its wider ESW and noncorrosive property. Excellent cyclability at low temperatures was also expected, owing to its phase stability and moderate σion. Before examining full-cell performance, we first tested the compatibility of the hybrid electrolyte with individual NaMnF and KMnC electrodes at RT. The galvanostatic C/D profiles of NaMnF revealed a reversible capacity of 139 mA·hour g−1 at 1C, with two pseudoplateaus at 0.75 and 1.29 V versus Ag/AgCl when charged from 0.0 to 1.45 V, highlighting the benefit of the wider ESW in the hybrid electrolyte (fig. S20). In comparison, the reversible capacity of NaMnF in 17.6 m NaCTFSI was only 101 mA·hour g−1 at 1C, limited by the narrower ESW and a high-voltage cutoff restricted to 1.30 V (Fig. 2B). NaMnF also demonstrated excellent long-term cycling stability in the hybrid electrolyte, retaining ~85% of its initial capacity after 500 C/D cycles at 5C, with discharge capacity fading from 105 to 89 mA·hour g−1 (fading rate = 0.032 mA·hour g−1 cycle−1). The C/D profiles of KMnC in the hybrid electrolyte were nearly identical to those in 17.6 m NaCTFSI, exhibiting a gentle slope between −1.3 and −1.0 V and a reversible capacity of 76 mA·hour g−1 at 1C (fig. S21). KMnC also showed excellent cyclability, retaining 80% of its initial capacity over 500 C/D cycles at 5C (fading rate = 0.008 mA·hour g−1 cycle−1 from the second C/D value). After confirming the compatibility of the individual electrode materials with the hybrid electrolyte, we assembled a full cell consisting of KMnC (4.4 mg) | 45.6 m hybrid | NaMnF (2.3 mg) and evaluated its electrochemical performance at RT. The C/D profiles exhibit a gradual decrease in capacity and a progressive increase in voltage hysteresis with increasing C-rates (Fig. 5A). The capacity declined from 134 mA·hour g−1cathode at 1C to 43 mA·hour g−1cathode at 20C, indicating that, although limited by kinetics, the cell remains operable even under extremely high C-rate conditions at RT. Full-cell electrochemical performance enabled by hybrid electrolytes. (A) C/D profiles of a KMnC | 45.6 m hybrid | NaMnF full cell at different C-rates. (B) Capacity retention of full cells in NaFSI, NaCTFSI, and the hybrid electrolytes at RT under 5C cycling. Solid symbols: charge; open symbols: discharge. (C) C/D profiles of a KMnC | 45.6 m hybrid | NaMnF full cell at low temperatures under 1C cycling. (D) Capacity retention of full cells in the hybrid electrolyte at different temperatures under 1C cycling. Solid symbols: charge; open symbols: discharge. (E) Photographs showing the brightness of LED lamps connected to pouch cells that were freshly charged, fully discharged, and charged after 500 cycles. The long-term cyclability test of the KMnC | 45.6 m hybrid | NaMnF full cell also demonstrated excellent capacity retention during 500 C/D cycles at RT (Fig. 5B). The initial discharge capacity of 103 mA·hour g−1cathode gradually decreased with cycling, finally reaching 87 mA·hour g−1cathode after 500 cycles. This degree of fading was comparable to that observed in 17.6 m NaCTFSI, which exhibited a capacity drop from 76 to 59 mA·hour g−1cathode over 500 cycles. In contrast, the KMnC | 35 m NaFSI | NaMnF full cell showed much faster capacity degradation, with discharge capacity falling from 69 to 28 mA·hour g−1cathode (retention = 40%; fading rate = 0.082 mA·hour g−1 cycle−1). Because the high-voltage cutoff was limited in both NaCTFSI and NaFSI electrolytes due to their narrower ESWs, the observed inferiority of cyclability in NaFSI is unlikely to originate from electrolyte decomposition. Instead, the superior long-term stability in the CTFSI−-based electrolytes (NaCTFSI and hybrid) appears to be primarily associated with corrosion suppression, as discussed in the “A 45.6 m NaCTFSI/NaFSI hybrid electrolyte with viable physicochemical properties” section. To highlight the key advantage of using the hybrid electrolyte in aqueous SIBs, we operated KMnC | 45.6 m hybrid | NaMnF full cells at 1C under subzero temperature conditions (0°, −10°, and −20°C). As the temperature decreased, the discharge capacity declined and polarization increased, reflecting slower Na+ transport kinetics (Fig. 5C). Nevertheless, the capacity remained notable: The full cells delivered 104, 90, and 75 mA·hour g−1cathode at 0°, −10°, and −20°C, respectively, corresponding to 77, 67, and 56% of the capacity delivered at RT (135 mA·hour g−1cathode). The voltage profiles also retained two pseudoplateaus during charging even at −20°C, indicating that the hybrid electrolyte can effectively mediate charge carrier transport despite kinetic limitations. Even when repeatedly cycled under challenging environmental conditions, this battery system maintains high-capacity retention. Figure 5D compares the capacity fading of full cells operated at 0°, −10°, and −20°C with that at RT under 1C cycling. Compared to a 19% capacity reduction after 500 cycles at RT, the extent of fading was slightly greater at low temperatures. The capacity loss gradually increased as the temperature decreased—10, 15, and 28% at 0,° −10°, and −20°C, respectively—yet the cell still delivered 54 mA·hour g−1cathode after 500 cycles at −20°C. The absence of an abrupt capacity drop further supports the retention of a supercooled state down to −20°C (for C/D profiles after 500 cycles, see fig. S22). Both NaMnF and KMnC preserved their original morphologies and compositions after 500 cycles at −20°C, as confirmed by FESEM and EDX analyses (figs. S8 and S9). The low-temperature performance of the 45.6 m hybrid full cells is compared with previously reported aqueous Na+-based energy storage systems in table S5. To visually demonstrate the practical applicability of the hybrid electrolyte, we fabricated a pouch cell [energy density ≈ 75 mA·hour g−1cathode x ~10 mgcathode x 1.75 V ≈ 1.3 mWh (milliwatt-hour)] and tested its ability to power two light-emitting diode (LED) lamps (~2 mW each) at −20°C (Fig. 5E). The fully charged cell initially lit the LEDs brightly, which gradually dimmed over time and became barely visible after ~20 min. This illumination time reasonably matched the value estimated from the cell’s energy and the lamp’s power consumption. Notably, even after 500 C/D cycles, the LEDs—although less bright than initially—could still be illuminated, confirming that the cell remains practically usable even after prolonged cycling at −20°C. DISCUSSION We have demonstrated that a concentrated NaCTFSI aqueous solution (17.6 m) functions as a true WISE, enabling a relatively wide ESW (2.78 V) and compatibility with high-voltage electrodes. However, despite these WISE characteristics, NaCTFSI suffers from poor low-temperature performance due to crystallization below RT, which limits its practical applicability under subzero conditions. To address this limitation, we developed a hybrid aqueous electrolyte by mixing NaCTFSI with NaFSI. The resulting formulation (14.1 m NaCTFSI and 31.5 m NaFSI) exhibits eutectic-like behavior, forming a deeply supercooled liquid phase that remains stable down to −20°C. This hybrid system not only improves low-temperature phase stability but also further extend the ESW to 3.07 V, thereby enabling the construction of high-voltage aqueous SIBs. Notably, CTFSI− anions in the hybrid electrolyte play a crucial role in suppressing anodic corrosion—common challenge in FSI−-based aqueous systems—thus ensuring long-term cycling stability. Spectroscopic analyses and DFT calculations confirmed that both CTFSI− and FSI− participate in the Na+ solvation shell, with water-mediated interactions facilitating the integration of multiple anion species. Electrochemical evaluation further verified that KMnC | 45.6 m hybrid | NaMnF full cells operate reliably for over 500 C/D cycles even under subzero conditions (down to −20°C), with acceptable capacity fading. The combination of enhanced low-temperature resilience, wide ESW, and noncorrosive behavior highlights the NaCTFSI-NaFSI hybrid electrolyte as a highly promising candidate for safe, durable, and wide-climate aqueous SIBs. MATERIALS AND METHODS Materials and synthesis NaFSI and NaTFSI (Solvionic, France) and NaCTFSI and NaFTFSI (Provisco, Czech Republic) were used as received. All other chemicals were purchased from Merck and used without further purification. Electrolyte solutions were prepared by dissolving the salts in high-purity deionized water (HPLC grade, Merck), followed by N2 degassing. KMnC was synthesized via a coprecipitation method. First, potassium hexacyanochromate (K3[Cr(CN)6]) was prepared as follows. An aqueous solution of chromium(III) acetate was slowly added to a boiling aqueous solution of KCN under vigorous stirring in a fume hood, maintaining a 1:6 stoichiometric ratio of chromium to cyanide. The resulting dark red solution was concentrated by evaporation on a steam bath until brownish-black precipitates were formed. The precipitates were collected by filtration and subsequently vacuum dried to obtain bright yellow crystals (K3[Cr(CN)6]). The crystals adhering to the container walls were collected and recrystallized twice. Next, an aqueous solution of MnCl2·4H2O was added dropwise to a vigorously stirred K3[Cr(CN)6] solution in a fume hood, maintaining a 1:1 stoichiometric ratio of Mn2+ to Cr(CN)6]3−. This resulted in the formation of light green precipitates. Light green precipitates were formed, separated by centrifugation, repeatedly washed with deionized water, and dried under vacuum at 120°C for 12 hours. The final product obtained was ochre in color. NaMnF was also synthesized by coprecipitation. Separate aqueous solutions of sodium hexacyanoferrate decahydrate (Na4[Fe(CN)6]·10H2O) and manganese sulfate monohydrate (MnSO4·H2O) were prepared in a saturated NaCl solution. The two solutions were mixed dropwise under continuous magnetic stirring at 60°C, maintaining a 1:1 stoichiometric ratio of Mn2+ to [Fe(CN)6]4−. The reaction was allowed to proceed for 12 hours, yielding precipitates that were collected by centrifugation, thoroughly washed with deionized water, and air-dried at 80°C. Characterization methods The σion of the electrolyte solution was determined using a two-electrode cell consisting of Al plates (0.7850 cm2) separated by 0.142 cm. Impedance spectra were recorded by applying a sinusoidal ac signal with an amplitude of ±10 mV over a frequency range of 1 MHz to 0.1 Hz (SP2, WonATech, Korea). The solution resistance was obtained from the high-frequency intercept of the Nyquist plot. Dynamic viscosity of the electrolyte solutions was measured using a rolling ball viscometer (Lovis 2000 M/ME, Anton Paar), and their density was determined with a density meter (DMA 5000M, Anton Paar). Both instruments were calibrated using deionized water at 30°C. Before each measurement, the U-tube was rinsed with anhydrous ethanol. The fluoride concentration was determined using a fluoride ion-selective electrode (combination electrode, 003B, iSTEK), capable of detecting [F−] in the range of 0.02 to 19,000 ppm. A series of standard solutions (0.1, 1.0, 10, 100, 1000, and 10,000 ppm) was prepared for calibration. Before analysis, the ionic strength and pH of the test solutions were adjusted using a buffer solution (TISAB II, F00ISO2). FTIR spectra were collected on a JASCO FT/IR-4200 spectrometer (Japan) equipped with a diamond attenuated total reflectance (ATR) accessory (iD1 ATR). A single drop of the solution was placed on the ATR crystal, and spectra were collected over the range of 4000 to 400 cm−1 at a resolution of 4 cm−1. A total of 16 scans were averaged per measurement to improve the signal-to-noise ratio. The melting points of NaTFSI and NaCTFSI was examined using a thermal analysis system (HITACHI-STA7200, Japan), whereas the thermal events of the aqueous electrolytes were investigated with a differential scanning calorimeter (DSC 214 Polyma, NETZSCH, Germany). Before analysis, electrolyte solutions were equilibrated at 60°C for 30 min. 1H and 19F NMR spectra were recorded on a Bruker Avance NEO 600 spectrometer (600 MHz). Aqueous samples (~300 μl) were loaded into the outer compartment of a coaxial NMR tube (Wilmad). DMSO-d6 (dimethyl sulfoxide-d6; d = 2.5 ppm) and trifluorotoluene (d = −63.72 ppm) were used as internal references for 1H and 19F, respectively. Acquisition parameters included 2 dummy scans, 16 scans, and a 1-s recycle delay for 1H (n0 = 600.25 MHz) and 4 dummy scans, 64 scans, and a 2-s delay for 19F (n0 = 564.74 MHz). All experiments were performed at 298 K, and spectral data were processed using the Bruker Topspin software. The crystal structures of the synthesized NaMnF and KMnC were analyzed by powder XRD (ULTIMA 4, Rigaku, Japan) using Cu Kα radiation source (l = 1.5405 Å). Data were collected over a 2θ range of 10° to 80° at a scan rate of 1° min−1. The elemental composition of the synthesized electrode materials was determined by EDX coupled with FESEM (JSM-7610F Plus, JEOL, Japan). The surface composition of electrodes was analyzed by XPS (K-Alpha, Thermo Fisher Scientific, USA).