Recent Progress in Cathode Materials for Sodium-Metal Halide Batteries

Transitioning from fossil fuels to renewable energy sources is a critical goal to address greenhouse gas emissions and climate change. Major improvements have made wind and solar power increasingly cost-competitive with fossil fuels. However, the inherent intermittency of renewable power sources motivates pairing these resources with energy storage. Electrochemical energy storage in batteries is widely used in many fields and increasingly for grid-level storage, but current battery technologies still fall short of performance, safety, and cost. This review focuses on sodium metal halide (Na-MH) batteries, such as the well-known Na-NiCl2 battery, as a promising solution to safe and economical grid-level energy storage. Important features of conventional Na-MH batteries are discussed, and recent literature on the development of intermediate-temperature, low-cost cathodes for Na-MH batteries is highlighted. By employing lower cost metal halides (e.g., FeCl2, and ZnCl2, etc.) in the cathode and operating at lower temperatures (e.g., 190 °C vs. 280 °C), new Na-MH batteries have the potential to offer comparable performance at much lower overall costs, providing an exciting alternative technology to enable widespread adoption of renewables-plus-storage for the grid.


Introduction
Mass deployment of renewable energy is crucial to achieve ambitious net-zero power generation in the 21st century. While renewables such as wind and solar are increasingly cost-competitive with conventional fossil-fuel-based power generation, intermittency precludes straightforward incorporation into the current power grid and can even potentially destabilize it with supply/demand mismatch (e.g., >20% in supply), leading to negative pricing and wasted power [1]. These factors motivate large-scale, and potentially longduration, energy storage coupled with renewable electricity production, which can improve revenue generation by storing electricity during low-demand periods and delivering it to the grid during high-demand periods [2]. Of the various energy storage technologies available, electrochemical energy storage is the most generally deployable, not requiring specific geography like pumped hydro or compressed air energy storage and has a wide range of chemistries allowing tuning of energy or power density for the desired application [1,3].
Unfortunately, most electrochemical energy storage technologies coupled to a renewable power source are not economically competitive with fossil fuels at present. A recent technoeconomic analysis indicates that energy storage capacity with a cost less than $20 per kilowatt-hour (kWh) is required for meeting 100% of demand with renewables [3]. If demand only needs to be met 95% of the time, a more achievable but still very low Figure 1. Schematic illustration of a Na-NiCl2 battery with a tubular β"-Al2O3 solid low). During operation, molten NaAlCl4 acts as the secondary electrolyte in the ca Figure 1. Schematic illustration of a Na-NiCl 2 battery with a tubular β"-Al 2 O 3 solid electrolyte (yellow). During operation, molten NaAlCl 4 acts as the secondary electrolyte in the cathode to conduct sodium ions between the active materials and the solid electrolyte. Particles of Ni form a conductive network radiating from the cathode current collector. A surface coating of NiCl 2 develops on the Ni particles, as sodium ions migrate across the solid electrolyte to the anode during charging.
Finally, new cell designs based on simple planar geometry along with inexpensive surface treatments [28] effectively mitigate the problem of poor Na-wetting on the solid electrolyte at low temperatures, and also make scalable fabrication of the BASE electrolyte easier, requiring e.g., a simple tape-casting process [20] rather than complex processing of tubular electrolytes [10][11][12][13]. With these major challenges solved, intermediate-temperature, low-cost Na-MH batteries provide new opportunities for scalable, inexpensive, grid-level energy storage to address the urgent need for increased utilization of renewables and their effective incorporation. This review aims to highlight some of the recent work on intermediate-temperature Na-MH battery chemistries, focusing on Ni, as well as low-cost Fe-, Cu-, and Zn-based cathode chemistries that will be crucial to reach the low-cost targets to unlock economical and scalable electrochemical energy storage.

Basic Working Principle
As far as 40 years ago, safety issues in developing high-temperature Na-S batteries prompted active investigation into new cathode materials. While similar corrosion problems persisted, a secondary cell with SCl 4 cathode in molten NaAlCl 4 showed attractive energy density and higher battery safety [29]. In 1986, Coetzer et al. reported ZEBRA batteries with transition metal halides as the cathodes in a NaAlCl 4 secondary Materials 2021, 14, 3260 4 of 18 electrolyte (m.p. 157 • C), which transported Na + between the active materials and solidstate electrolytes [30]. The solid metal halides were more stable and less corrosive at high temperatures, resolving multiple problems in the Na-S system. Among all metal halides (NiCl 2 , FeCl 2 , ZnCl 2 , CuCl 2 , etc.) compatible with the ZEBRA battery concept, the Na-NiCl 2 battery has been most widely studied. It was a promising candidate for large-scale energy storage application, with the overall redox reaction below with E 0 = 2.58 V at 300 • C (Equation (1)) [30,31].
2NaCl + Ni (discharged state) ⇄ 2Na + NiCl2 (charged state), To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reduction of NiCl2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple assembly, and good cycle life of a Na-NiCl2 battery, research efforts have explored the electrochemical mechanism of Na-NiCl2 cells in depth and identified factors limiting the cycling performance, including passivation from nickel chloride [22,32], dissolution loss of nickel in the NaAlCl4 melt [32], and capacity decline due to Ni/NaCl particle growth. In response, various chemical additives were introduced.
Bones et al. witnessed severe early capacity loss at 250 °C (~50% in <10 cycles), which related to the growth of the nickel grains [33]. At the same time, they found that sulfur additive in NaAlCl4 melt could develop an active, metallic phase with high surface area on the cathode. As a result, the long-term cyclability improved, retaining >75% of capacity over 2000 cycles with a negligible increase in the cell resistance [33]. Ratnakumar et al. verified the effect of sulfur additives on limiting Ni growth during cycling and determined the optimal amount between 0.5 and 1.0 wt.% [34]. Prakash et al. identified the nickel chloride layer formation during charging as another factor in the nickel cathode degradation [35]. Although the NiCl2 layer initially increased the area capacity of a nonporous nickel electrode by decreasing the impedance between 200 °C and 225 °C, the highly resistive surface layer ultimately led to an accumulation of unutilized NaCl in the chloroaluminate melt and limited the areal capacity at higher temperatures.
In addition to sulfur, researchers explored other additives including NaI/NaBr, Al, FeS, and Ni3S2. By doping the NaAlCl4 melt with NaI or NaBr, NiBr2 or NiI2 with larger lattice parameters than NiCl2 formed in the surface layer. Those surface phases could offer faster mass transport and in turn improve the nickel utilization [32].
Al participated in the electrochemical reaction at around 1.58 V (Figure 2a,b) and alleviated the sudden polarization at the end-of-discharge by the following reaction, (Equation (2)) 4NaCl + Al ⇄ 3Na + NaAlCl4, Furthermore, the dissolution of Al via Equation (2) left fine pores inside the cathode, which enhanced high-rate performance of the Na-NiCl2 battery [11].
FeS showed signature intermediate plateaus in the voltage profile (Figure 2a,b) and could stabilize the nickel grain size similar to elemental sulfur additives [11]. Li et al. systematically investigated the role of FeS in initial activation and subsequent degradation in Na-NiCl2 cells. Figure 2c,d shows the first-cycle charge-discharge curves of cells with no FeS (0X) and 1 mol% FeS (1X) loadings. The voltage of a cell without FeS quickly rose to 2.8 V at 10 mA, with no observed plateau at ~2.6 V corresponding to Ni/Ni 2+ . X-ray photoelectron spectroscopy (XPS) analysis on raw Ni powder indicated a passivating surface layer (oxides and hydroxides), which not only increased the ohmic resistance but also decreased the electrochemically active surface area (Figure 2e,f). The addition of FeS largely 2Na + NiCl 2 (charged state), To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl 2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al 2 O 3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl 2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl 2 concurrent with reduction of NiCl 2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple assembly, and good cycle life of a Na-NiCl 2 battery, research efforts have explored the electrochemical mechanism of Na-NiCl 2 cells in depth and identified factors limiting the cycling performance, including passivation from nickel chloride [22,32], dissolution loss of nickel in the NaAlCl 4 melt [32], and capacity decline due to Ni/NaCl particle growth. In response, various chemical additives were introduced.
Bones et al. witnessed severe early capacity loss at 250 • C (~50% in <10 cycles), which related to the growth of the nickel grains [33]. At the same time, they found that sulfur additive in NaAlCl 4 melt could develop an active, metallic phase with high surface area on the cathode. As a result, the long-term cyclability improved, retaining >75% of capacity over 2000 cycles with a negligible increase in the cell resistance [33]. Ratnakumar et al. verified the effect of sulfur additives on limiting Ni growth during cycling and determined the optimal amount between 0.5 and 1.0 wt.% [34]. Prakash et al. identified the nickel chloride layer formation during charging as another factor in the nickel cathode degradation [35]. Although the NiCl 2 layer initially increased the area capacity of a nonporous nickel electrode by decreasing the impedance between 200 • C and 225 • C, the highly resistive surface layer ultimately led to an accumulation of unutilized NaCl in the chloroaluminate melt and limited the areal capacity at higher temperatures.
In addition to sulfur, researchers explored other additives including NaI/NaBr, Al, FeS, and Ni 3 S 2 . By doping the NaAlCl 4 melt with NaI or NaBr, NiBr 2 or NiI 2 with larger lattice parameters than NiCl 2 formed in the surface layer. Those surface phases could offer faster mass transport and in turn improve the nickel utilization [32].
Al participated in the electrochemical reaction at around 1.58 V (Figure 2a,b) and alleviated the sudden polarization at the end-of-discharge by the following reaction, (Equation (2) storage application, with the overall redox reaction below with E0 = 2.58 V at 300 °C (Equation (1)) [30,31].
To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reduction of NiCl2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple assembly, and good cycle life of a Na-NiCl2 battery, research efforts have explored the electro- removed the passivation layer on Ni powder during the low-current charging period, indicated by the shrinking peaks of Ni 2+ species in XPS. The formation of polysulfide species likely accounted for the surface activation. An optimal level of FeS was necessary to remove surface passivation completely while not inducing Ni particle growth that would later hinder cell performance [36]. Recently, Ao et al. [37,38] showed that the surface modification of Ni powder by Ni3S2, from either S or Ni3S2, can mitigate Ni particle aggregation and promote reversibility. The proposed self-repairing process and mitigation to Ni particle growth relied on in situ Ni3S2 surface layer formation, schematically illustrated in Figure 3. Although more added sulfur led to the longer sustained self-repairing in the cathode, the layer thickness in excess would degrade cycle performance, owing to the poor electrical conductivity of sulfur and Ni3S2 [38].

Effects of Operating Temperature and Electrode Morphology
The operating temperature and electrode morphology were two key aspects dictating the interfacial solid-solid conversion, an interplay of thermodynamic and kinetic forces behind performance optimization. Early study by Hosseinifar et al. correlated the performance and aging of ZEBRA cells with operating temperatures of 260 °C and 350 °C [39]. At 350 °C, a combination of high temperature, accelerated NiCl2 dissolution in molten NaAlCl4, and the formation of the Ni3S2 phase enabled nickel grain growth, which resulted in a significant capacity loss. Lu et al. constructed a planar-type Na-NiCl2 cell that could be cycled at C/3 at a low temperature of 175 °C [40]. Despite the mitigated particle growth, high cell resistance and thus large cell polarization appeared, arising from either slow Na + transport within the NaAlCl4 melt or poor wettability of sodium on the BASE at lower operating temperatures. Furthermore, the dissolution of Al via Equation (2) left fine pores inside the cathode, which enhanced high-rate performance of the Na-NiCl 2 battery [11].
FeS showed signature intermediate plateaus in the voltage profile (Figure 2a,b) and could stabilize the nickel grain size similar to elemental sulfur additives [11]. Li et al. systematically investigated the role of FeS in initial activation and subsequent degradation in Na-NiCl 2 cells. Figure 2c,d shows the first-cycle charge-discharge curves of cells with no FeS (0X) and 1 mol% FeS (1X) loadings. The voltage of a cell without FeS quickly rose to 2.8 V at 10 mA, with no observed plateau at~2.6 V corresponding to Ni/Ni 2+ . X-ray photoelectron spectroscopy (XPS) analysis on raw Ni powder indicated a passivating surface layer (oxides and hydroxides), which not only increased the ohmic resistance but also decreased the electrochemically active surface area (Figure 2e,f). The addition of FeS largely removed the passivation layer on Ni powder during the low-current charging period, indicated by the shrinking peaks of Ni 2+ species in XPS. The formation of polysulfide species likely accounted for the surface activation. An optimal level of FeS was necessary to remove surface passivation completely while not inducing Ni particle growth that would later hinder cell performance [36].
Recently, Ao et al. [37,38] showed that the surface modification of Ni powder by Ni 3 S 2 , from either S or Ni 3 S 2 , can mitigate Ni particle aggregation and promote reversibility. The proposed self-repairing process and mitigation to Ni particle growth relied on in situ Ni 3 S 2 surface layer formation, schematically illustrated in Figure 3. Although more added sulfur led to the longer sustained self-repairing in the cathode, the layer thickness in excess would degrade cycle performance, owing to the poor electrical conductivity of sulfur and Ni 3 S 2 [38].  Later, Li et al. demonstrated that a planar Na-NiCl2 battery operating at an intermediate temperature of 190 °C with ultra-high energy density ( Figure 4) [20]. Discharged at C/5, the cell showed a capacity increase for the first 100 cycles before stabilizing at 137 mAh/g. In contrast, the cell cycled at 280 °C, despite a high initial capacity, exhibited major capacity fading after 100 cycles. Furthermore, the intermediate-temperature cell delivered an impressive energy density of 350 Wh/kg over 1000 cycles with negligible capacity decay. The higher initial capacity observed at 280 °C than at 190 °C was ascribed to the better sodium wetting on the BASE. From the voltage profiles vs. state of charge (SOC) ( Figure  4a,b), the cell operated at 280 °C had a stable SOC at end of discharge (SOCEOD). However, its SOC at end of charge (SOCEOC) dropped rapidly over 200 cycles, indicating that cathode degradation was mainly responsible for capacity fade. This assertion was further supported by morphology analysis using scanning electron microscopy (SEM) with energy dispersive X-ray (EDX) spectroscopy, where the particle growth of both NaCl and Ni at 280 °C was much faster than at 190 °C (Figure 4c,d). Conceptually, lower operating temperatures would not only decrease the solubilities of cathode materials (NaCl and NiCl2) but also the diffusion coefficients of the corresponding ions, due to an increased viscosity of the melt. This combined effect hampers diffusion-influenced growth [41]. As a result, Ostwald ripening, previously identified as the underlying particle growth mechanism at 280 °C [22], is significantly suppressed at lower operating temperatures to maintain metal grain size and prolong cell life [20]. . Schematic illustration of the Ni 3 S 2 layer self-repairing process and the proposed mechanism of sulfur additive via preventing Ni particle growth (adapted with the permission from ACS).

Effects of Operating Temperature and Electrode Morphology
The operating temperature and electrode morphology were two key aspects dictating the interfacial solid-solid conversion, an interplay of thermodynamic and kinetic forces behind performance optimization. Early study by Hosseinifar et al. correlated the performance and aging of ZEBRA cells with operating temperatures of 260 • C and 350 • C [39]. At 350 • C, a combination of high temperature, accelerated NiCl 2 dissolution in molten NaAlCl 4 , and the formation of the Ni 3 S 2 phase enabled nickel grain growth, which resulted in a significant capacity loss. Lu et al. constructed a planar-type Na-NiCl 2 cell that could be cycled at C/3 at a low temperature of 175 • C [40]. Despite the mitigated particle growth, high cell resistance and thus large cell polarization appeared, arising from either slow Na + transport within the NaAlCl 4 melt or poor wettability of sodium on the BASE at lower operating temperatures.
Later, Li et al. demonstrated that a planar Na-NiCl 2 battery operating at an intermediate temperature of 190 • C with ultra-high energy density (Figure 4) [20]. Discharged at C/5, the cell showed a capacity increase for the first 100 cycles before stabilizing at 137 mAh/g. In contrast, the cell cycled at 280 • C, despite a high initial capacity, exhibited major capacity fading after 100 cycles. Furthermore, the intermediate-temperature cell delivered an impressive energy density of 350 Wh/kg over 1000 cycles with negligible capacity decay. The higher initial capacity observed at 280 • C than at 190 • C was ascribed to the better sodium wetting on the BASE. From the voltage profiles vs. state of charge (SOC) (Figure 4a,b), the cell operated at 280 • C had a stable SOC at end of discharge (SOC EOD ). However, its SOC at end of charge (SOC EOC ) dropped rapidly over 200 cycles, indicating that cathode degradation was mainly responsible for capacity fade. This assertion was further supported by morphology analysis using scanning electron microscopy (SEM) with energy dispersive X-ray (EDX) spectroscopy, where the particle growth of both NaCl and Ni at 280 • C was much faster than at 190 • C (Figure 4c,d). Conceptually, lower operating temperatures would not only decrease the solubilities of cathode materials (NaCl and NiCl 2 ) but also the diffusion coefficients of the corresponding ions, due to an increased viscosity of the melt. This combined effect hampers diffusion-influenced growth [41]. As a result, Ostwald ripening, previously identified as the underlying particle growth mechanism at 280 • C [22], is significantly suppressed at lower operating temperatures to maintain metal grain size and prolong cell life [20].

framework.
Li et al. carried out extensive studies of the cell performance in correlation to morphology evolution of NaCl and Ni particles, by varying Ni/NaCl ratios and cut-off voltages for EOC [22]. A high current density and a low Ni/NaCl ratio favored the Ni particle growth, while a wide cycling capacity window was mainly responsible for the growth of NaCl. More importantly, excess Ni alleviated performance degradation due to particle growth by offering an additional active electrode surface and electron pathways. As a result, the cell with a cathode Ni/NaCl ratio of 1.8 was able to maintain high capacity over extended cycling. To further lower cell cost by cutting expensive Ni, a recent work by Chang et al. reduced the Ni/NaCl ratio to 1.25 without losing the cell performance at 190 °C [18]. Among many Ni/NaCl ratios, a Ni/NaCl ratio of 1.25 delivered a cell with the highest specific energy density of 405 Wh/kg, with almost no capacity fading over 300 cycles. Morphology analysis showed no significant particle growth after 300 cycles for cells with Ni/NaCl ≥ 1.25. Although the cell did not have the highest capacity at high currents due to low Ni content, it was the most cost-effective with a 30% lower cost of Ni.  Since Na-MH batteries rely on solid-solid conversion chemistry, a large, active cathode interface between NaCl and Ni is important to achieve good electrochemical performance. Prakash et al. investigated three different cathode morphologies (a nonporous nickel substrate, a nickel felt, and a porous sintered nickel electrode) and achieved the best cell performance with the porous sintered electrode. They demonstrated that both surface area and the pore accessibility were critical to performance optimization [42]. Kim et al. compared the electrochemical performance of Na-NiCl 2 cells using Ni powders with different initial particle sizes (0.5, 6, and 30 µm), showing that smaller particle size (i.e., larger surface area) allowed higher initial charge capacities and more complete utilization of the NaCl [43]. However, capacity fade occurred quickly in cells initially assembled in the discharged state with smaller Ni particles, apparently resulting from particles becoming disconnected from the electron conduction framework due to volume expansion from the formation of surface NiCl 2 on charging followed by volume contraction on the reduction back to Ni [43]. This problem was mitigated by using 'reticular' Ni to maintain the large surface area of the active material while ensuring an interconnected electron conducting framework.
Li et al. carried out extensive studies of the cell performance in correlation to morphology evolution of NaCl and Ni particles, by varying Ni/NaCl ratios and cut-off voltages for EOC [22]. A high current density and a low Ni/NaCl ratio favored the Ni particle growth, while a wide cycling capacity window was mainly responsible for the growth of NaCl. More importantly, excess Ni alleviated performance degradation due to particle growth by offering an additional active electrode surface and electron pathways. As a result, the cell with a cathode Ni/NaCl ratio of 1.8 was able to maintain high capacity over extended cycling. To further lower cell cost by cutting expensive Ni, a recent work by Chang et al. reduced the Ni/NaCl ratio to 1.25 without losing the cell performance at 190 • C [18]. Among many Ni/NaCl ratios, a Ni/NaCl ratio of 1.25 delivered a cell with the highest specific energy density of 405 Wh/kg, with almost no capacity fading over 300 cycles. Morphology analysis showed no significant particle growth after 300 cycles for cells with Ni/NaCl ≥ 1.25. Although the cell did not have the highest capacity at high currents due to low Ni content, it was the most cost-effective with a 30% lower cost of Ni.
Clearly, the microstructure of the cathode material, which enables electrical connection to small, high-surface-area particles, has a significant influence on electrochemical performance over extended cycling and is therefore crucial to optimize. In another study, a core-shell microarchitecture of nickel-coated graphite reduced the Ni loading by 40%. The cell operating at 190 • C showed high initial energy density (133 Wh/kg at~C/4) and excellent capacity retention (94% over 150 cycles). The decent electrochemical performance was ascribed to the fact that the graphite core successfully retained active surface and structural integrity for fast electron transport [17].
The incorporation of carbon-based current collectors into the Ni cathode was further developed by Wen's group through nano engineering [19,21,44]. This approach was effective in providing a continuous electron-conducting network and mitigating particle growth of NaCl and Ni. For example, a free-standing Ni-less cathode with Ni/NaCl particles uniformly distributed in the three-dimensional (3D) conductive matrix constructed from carbon fiber (CF) and multiwalled carbon nanotubes (MWCNTs) is shown in Figure 5a,b [44]. The 3D hierarchical structure via a low-cost vacuum infiltration method ( Figure 5a) offered sufficient void space to accommodate the NaCl particle growth and effectively enhance the electronic conductivity of the composite cathode. As a result, at a Ni/NaCl ratio of 1, the MWCNTs/CF/Ni/NaCl-1 cathode achieved an initial reversible capacity of 172 mAh/g, up to 96.5% of its theoretical capacity. In stark contrast, the traditional Ni/NaCl cathode only delivered around 40% of its theoretical capacity in the first conditioning cycle. At a lower ratio of Ni/NaCl of 0.5, the MWCNTs/CF/Ni/NaCl-1 cathode exhibited a capacity of 215 mAh/g, which further confirmed the excellent electronic conductivity within the 3D porous network (Figure 5c). Recently, Gao et al. reported another high-rate and longlife intermediate-temperature Na-NiCl 2 battery enabled by a dual-functional Ni-carbon composite nanofiber (NCCN) network. The NCCNs could not only offer a continuous conductivity network but also limit the growth of Ni and NaCl grains [19]. The porosity of the cathode, which was controlled primarily by the grain size of the cathode metal, can now be tuned with precise structural engineering on a nanoscale.

Cathode Reaction Kinetics with Respect to Halides
Reducing the operating temperature of Na-NiCl2 battery can provide higher longterm cyclability and the adoption of inexpensive polymer materials for cell sealing [20]. However, the reduction in operating temperature results in inferior rate performance. Specifically, the Na-ion transport within the BASE, within the secondary electrolyte, and through the BASE/Na interface, is slower at lower temperatures. More significantly, the cathode reaction processes (such as the formation of NiCl2 and dissolution of NaCl) critical to the battery rate performance are also more sluggish. If Na-MH batteries are to become commercially competitive with increased market penetration, the reaction kinetics at lower temperatures require significant improvement [4].
Zhan et al. recently adopted a strategy of manipulating the halides in the Na-MH battery cathodes [45]. Comparing the typical voltage profiles of NaCl/Ni, NaBr/Ni, and NaI/Ni cathodes shown in Figure 6a-c, NaCl/Ni and NaI/Ni cathodes exhibited a single plateau for charging and the discharging process. On the contrary, the NaBr/Ni cathode showed an extra discharge plateau at ~2.47 V with a larger overpotential, which was a result of phase relaxations among multiple NaBrxCl1-x based on high-resolution synchrotron X-ray diffraction (XRD) and ion chromatography analysis. This cathode showed a significant improvement in rate capacity among the three compositions (Figure 6d-e) and excellent cycling performance over 400 cycles (>100 days) at 30 mA (10 mA/cm 2 ) with negligible capacity fading in the example of NaBr/Ni (Figure 6f). Combining inductively coupled plasma-optical emission spectroscopy (ICP-OES), cyclic voltammetry (CV), and SEM characterizations, the observed rate improvement was attributed to faster NaBr dissolution in molten NaAlCl4.

Cathode Reaction Kinetics with Respect to Halides
Reducing the operating temperature of Na-NiCl 2 battery can provide higher longterm cyclability and the adoption of inexpensive polymer materials for cell sealing [20]. However, the reduction in operating temperature results in inferior rate performance. Specifically, the Na-ion transport within the BASE, within the secondary electrolyte, and through the BASE/Na interface, is slower at lower temperatures. More significantly, the cathode reaction processes (such as the formation of NiCl 2 and dissolution of NaCl) critical to the battery rate performance are also more sluggish. If Na-MH batteries are to become commercially competitive with increased market penetration, the reaction kinetics at lower temperatures require significant improvement [4].
Zhan et al. recently adopted a strategy of manipulating the halides in the Na-MH battery cathodes [45]. Comparing the typical voltage profiles of NaCl/Ni, NaBr/Ni, and NaI/Ni cathodes shown in Figure 6a-c, NaCl/Ni and NaI/Ni cathodes exhibited a single plateau for charging and the discharging process. On the contrary, the NaBr/Ni cathode showed an extra discharge plateau at~2.47 V with a larger overpotential, which was a result of phase relaxations among multiple NaBr x Cl 1-x based on high-resolution synchrotron X-ray diffraction (XRD) and ion chromatography analysis. This cathode showed a significant improvement in rate capacity among the three compositions (Figure 6d,e) and excellent cycling performance over 400 cycles (>100 days) at 30 mA (10 mA/cm 2 ) with negligible capacity fading in the example of NaBr/Ni (Figure 6f). Combining inductively coupled plasma-optical emission spectroscopy (ICP-OES), cyclic voltammetry (CV), and SEM characterizations, the observed rate improvement was attributed to faster NaBr dissolution in molten NaAlCl 4 .
This finding also strongly suggested that the rate-limiting step of the intermediatetemperature Na-MH batteries was the dissolution of sodium salts (e.g., NaCl) rather than metal halide (e.g., NiCl 2 ) formation during the charging process (Figure 6g). While the ion diffusion coefficient of NiBr 2 was the lowest among the three halide salts obtained through cyclic voltammetry, the solubility of NaBr was the highest obtained through ICP-OES. If the rate-limiting step was the formation of NiX 2 (X = Cl, Br, or I) during the charging process, then the cell would expect passivation by NiBr 2 deposition at 100% SOC due to a high concentration of standing Br − and low mobility of NiBr 2 . Instead, no significant features appeared in backscatter SEM images. Furthermore, the low-rate sensitivity of the specific discharge capacity implied sufficient supply of standing Br − at all times, cementing NaX salt dissolution as the rate-limiting step. Figure 6. (a-c) Typical voltage profiles during the conditioning cycles for the (a) NaCl/Ni, (b) NaBr/Ni, and (c) NaI/Ni cathodes. The current density is 3.3 mA/cm 2 for both charge and discharge, equivalent to ~C/16, ~C/13, and ~C/10 for NaCl/Ni, NaBr/Ni, and NaI/Ni, respectively, and (d) plots of the corresponding rate performance in terms of discharge capacity. (e) Capacity retention vs. C-rate for NaCl/Ni, NaBr/Ni, and NaI/Ni cells. (f) Long-term performance of (c) NaBr/Ni cell cycled at a current density of 10 mA/cm 2 . (g) Schematic view of reaction mechanisms for a Na-NiCl2 battery during charging (adapted with the permission from Elsevier).
This finding also strongly suggested that the rate-limiting step of the intermediatetemperature Na-MH batteries was the dissolution of sodium salts (e.g., NaCl) rather than metal halide (e.g., NiCl2) formation during the charging process (Figure 6g). While the ion diffusion coefficient of NiBr2 was the lowest among the three halide salts obtained through cyclic voltammetry, the solubility of NaBr was the highest obtained through ICP-OES. If the rate-limiting step was the formation of NiX2 (X = Cl, Br, or I) during the charging process, then the cell would expect passivation by NiBr2 deposition at 100% SOC due to a high concentration of standing Br − and low mobility of NiBr2. Instead, no significant features appeared in backscatter SEM images. Furthermore, the low-rate sensitivity of the specific discharge capacity implied sufficient supply of standing Br − at all times, cementing NaX salt dissolution as the rate-limiting step.

Basic Working Principle and Mechanism
The lower cost of Fe has motivated replacing Ni with Fe as the cathode material during the development of the Na-MH batteries. Examples of solid-state FeCl2 cathodes in conjunction with NaAlCl4 molten salt as the electrolyte date back to 1980 [30,[46][47][48], with an open-circuit voltage of 2.35 V at 250 °C (Figure 7a, red dash line). Like the Na-NiCl2 cell, the Na-FeCl2 cell was also loaded in the discharged state with Fe and NaCl in the cathode compartment for the ease of assembly. When the cell was overcharged (Figure 7a, purple dash line), the resulting FeCl3 that dissolved in NaAlCl4 could poison the BASE, leading to a rise in resistance [47]. This undesired side reaction could be mitigated by Ni additives, which offer an overcharge protection by undergoing a reversible cell reaction at 2.59 V (black dash line, Figure 7a) and forming insoluble NiCl2. On the other hand, the cell reaction at 1.60 V (blue dash line, Figure 7a) corresponds to the reduction of NaAlCl4 to Al and NaCl, serving as a good over-discharge protection. It was also shown to be important to keep the NaAlCl4 melt basic (NaCl/AlCl3 ≥ 1), ensuring an intimate contact amongst Fe, FeCl2, and NaCl in the cathode [46,47]. (a-c) Typical voltage profiles during the conditioning cycles for the (a) NaCl/Ni, (b) NaBr/Ni, and (c) NaI/Ni cathodes. The current density is 3.3 mA/cm 2 for both charge and discharge, equivalent to~C/16,~C/13, and~C/10 for NaCl/Ni, NaBr/Ni, and NaI/Ni, respectively, and (d) plots of the corresponding rate performance in terms of discharge capacity. (e) Capacity retention vs. C-rate for NaCl/Ni, NaBr/Ni, and NaI/Ni cells. (f) Long-term performance of (c) NaBr/Ni cell cycled at a current density of 10 mA/cm 2 . (g) Schematic view of reaction mechanisms for a Na-NiCl 2 battery during charging (adapted with the permission from Elsevier).

Basic Working Principle and Mechanism
The lower cost of Fe has motivated replacing Ni with Fe as the cathode material during the development of the Na-MH batteries. Examples of solid-state FeCl 2 cathodes in conjunction with NaAlCl 4 molten salt as the electrolyte date back to 1980 [30,[46][47][48], with an open-circuit voltage of 2.35 V at 250 • C (Figure 7a, red dash line). Like the Na-NiCl 2 cell, the Na-FeCl 2 cell was also loaded in the discharged state with Fe and NaCl in the cathode compartment for the ease of assembly. When the cell was overcharged (Figure 7a, purple dash line), the resulting FeCl 3 that dissolved in NaAlCl 4 could poison the BASE, leading to a rise in resistance [47]. This undesired side reaction could be mitigated by Ni additives, which offer an overcharge protection by undergoing a reversible cell reaction at 2.59 V (black dash line, Figure 7a) and forming insoluble NiCl 2 . On the other hand, the cell reaction at 1.60 V (blue dash line, Figure 7a) corresponds to the reduction of NaAlCl 4 to Al and NaCl, serving as a good over-discharge protection. It was also shown to be important to keep the NaAlCl 4 melt basic (NaCl/AlCl 3 ≥ 1), ensuring an intimate contact amongst Fe, FeCl 2 , and NaCl in the cathode [46,47].
Unlike a single-stage reaction in Na-NiCl 2 cell, the partially discharged Na-FeCl 2 cell contained various intermediate phases with a general formula of Na 8-x Fe x Cl 8 [48]. At 250 • C, Adendorff et al. found that the electrochemical reaction proceeded in a two-stage manner (Equations (3) and (4) To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reduction of NiCl2 to Ni. 3Fe + Na 6 FeCl 8 , To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reduction of NiCl2 to Ni.
The intermediate Na 6 FeCl 8 crystalized into a defected rock-salt structure (Fm3 m) where 1/8 of the octahedral sites were empty. Raising the operating temperature to 350 • C, two other phases, orthorhombic Na 2 FeCl 4 (Pbam) and trigonal Na 2 Fe 3 Cl 8 (R3 m), also formed. Unlike FeCl 3 , the formation of these complex intermediate phases showed neither significant poisoning of the BASE by Fe 2+ ions, nor structural disintegration of the sintered iron electrode even after hundreds of cycles [48]. Moseley et al. investigated the potential effects of iron incorporation on BASE stability and cell performance in Na/FeCl 2 cells. Their findings suggested that cells operated under the normal temperature (i.e., 250 • C) with overcharge/over-discharge protections were free from the concerns of iron incorporation into β" grains [46]. Ratnakumar et al. confirmed the abovementioned stepwise conversion of FeCl 2 cathode based on cyclic voltammetry results. The potentials of reaction (3) and (4) were determined to be 2.353 V and 2.341 V, respectively, with a separation of 12 mV. Slight asymmetry between the oxidation and reduction peaks was also observed, corresponding to the relative kinetics of individual charge transfer steps and the relative amounts of the intermediate species [49]. Coetzer et al. evaluated the electrochemical behaviors of iron electrodes with NaCl-saturated NaAlCl 4 in the temperature range 175-400 • C [50]. Remarkably, the results pointed to a predominantly solid-state reaction mechanism near 250 • C, which was further confirmed by the successful operation of a solid-state Na-FeCl 2 cell with no liquid electrolyte. At high temperatures (300-400 • C), dissolved species such as FeCl 2 contributed increasingly to the cathode reaction mechanism. Orchard et al. modeled the discharge behavior of the Na-FeCl 2 cells [51]. Their calculations, where the solubilities of FeCl 2 and NaCl were disregarded, agreed with the cathode reaction as effectively a solid-state process. Unlike a single-stage reaction in Na-NiCl2 cell, the partially discharged Na-FeCl2 cell contained various intermediate phases with a general formula of Na8-xFexCl8 [48]. At 250 °C, Adendorff et al. found that the electrochemical reaction proceeded in a two-stage manner (Equations (3) and (4)): 2Na + Na6FeCl8 ⇄ Fe + 8NaCl, The intermediate Na6FeCl8 crystalized into a defected rock-salt structure (Fm3 m) where 1/8 of the octahedral sites were empty. Raising the operating temperature to 350 °C, two other phases, orthorhombic Na2FeCl4 (Pbam) and trigonal Na2Fe3Cl8 (R3 m), also formed. Unlike FeCl3, the formation of these complex intermediate phases showed neither Battery activation and ease of assembly were also important to Na-FeCl 2 cells. An early approach by Coetzer et al. used pre-sintering of the Fe cathode in a reducing environment to form electron-conducting paths for initial charging of batteries assembled in the discharged state [30,50]. Alternatively, Bones et al. exploited gaseous chlorination of the Fe cathode by Cl 2 and then assembled the battery with a partially charged Fe/FeCl 2 cathode and a sodium metal anode [47]. Both approaches introduced complications due to the handling of hazardous/sensitive materials like reduced Fe, FeCl 2 , or metallic sodium. Recently, Li et al. proposed an easy discharged-state assembly of a Na-FeCl 2 battery utilizing sulfur-based additives [26]. As shown in Figure 7b, the cell with 4 mol.% S addition showed a significant reduction of Fe 3+ related (Fe 2 O 3 ) peaks by XPS, indicating the successful removal of oxide layers on the Fe cathode during the cell conditional cycling. The oxide removal not only freed Fe metal surface for Fe/FeCl 2 electrochemical conversion (Figure 7c), but also reduced cell resistance to lower overall cell polarization. As a result, the cell using an optimal amount of sulfur additives exhibited excellent cycling performance with no capacity fade and minimum end-of-charge (EOD) voltage degradation up to 100 cycles [26].

Fast-Charging Capability
In addition to the low cost and abundance of raw materials (Fe and NaCl), Na-FeCl 2 batteries have also been demonstrated with fast-charging capability [27]. Following the successful self-activation of a Fe powder cathode enabled by sulfur-based additives, Zhan et al. recently reported a high-rate Na-FeCl 2 battery operated at 190 • C with voltage profile shown in Figure 8a. NaCl first transformed to Na 6 FeCl 8 (Equation (5)) then to FeCl 2 (Equation (6)) during the charging process and overall reaction can be written as Equation (7) To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reduction of NiCl2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple assembly, and good cycle life of a Na-NiCl2 battery, research efforts have explored the electrochemical mechanism of Na-NiCl2 cells in depth and identified factors limiting the cycling performance, including passivation from nickel chloride [22,32], dissolution loss of nickel in the NaAlCl4 melt [32], and capacity decline due to Ni/NaCl particle growth. In response, various chemical additives were introduced.
Bones et al. witnessed severe early capacity loss at 250 °C (~50% in <10 cycles), which related to the growth of the nickel grains [33]. At the same time, they found that sulfur additive in NaAlCl4 melt could develop an active, metallic phase with high surface area on the cathode. As a result, the long-term cyclability improved, retaining >75% of capacity over 2000 cycles with a negligible increase in the cell resistance [33]. Ratnakumar et al. verified the effect of sulfur additives on limiting Ni growth during cycling and determined the optimal amount between 0.5 and 1.0 wt.% [34]. Prakash et al. identified the nickel chloride layer formation during charging as another factor in the nickel cathode degradation [35]. Although the NiCl2 layer initially increased the area capacity of a nonporous nickel electrode by decreasing the impedance between 200 °C and 225 °C, the highly resistive surface layer ultimately led to an accumulation of unutilized NaCl in the chloroaluminate melt and limited the areal capacity at higher temperatures.
In addition to sulfur, researchers explored other additives including NaI/NaBr, Al, FeS, and Ni3S2. By doping the NaAlCl4 melt with NaI or NaBr, NiBr2 or NiI2 with larger lattice parameters than NiCl2 formed in the surface layer. Those surface phases could offer faster mass transport and in turn improve the nickel utilization [32].
Al participated in the electrochemical reaction at around 1.58 V (Figure 2a,b) and alleviated the sudden polarization at the end-of-discharge by the following reaction, (Equation (2)) 4NaCl + Al ⇄ 3Na + NaAlCl4, Furthermore, the dissolution of Al via Equation (2) left fine pores inside the cathode, which enhanced high-rate performance of the Na-NiCl2 battery [11].
FeS showed signature intermediate plateaus in the voltage profile (Figure 2a,b) and could stabilize the nickel grain size similar to elemental sulfur additives [11]. Li et al. systematically investigated the role of FeS in initial activation and subsequent degradation in 2Na + Na 6 FeCl 8 , 0 25% SOC, To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reduction of NiCl2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple assembly, and good cycle life of a Na-NiCl2 battery, research efforts have explored the electrochemical mechanism of Na-NiCl2 cells in depth and identified factors limiting the cycling performance, including passivation from nickel chloride [22,32], dissolution loss of nickel in the NaAlCl4 melt [32], and capacity decline due to Ni/NaCl particle growth. In response, various chemical additives were introduced.
Bones et al. witnessed severe early capacity loss at 250 °C (~50% in <10 cycles), which related to the growth of the nickel grains [33]. At the same time, they found that sulfur additive in NaAlCl4 melt could develop an active, metallic phase with high surface area on the cathode. As a result, the long-term cyclability improved, retaining >75% of capacity over 2000 cycles with a negligible increase in the cell resistance [33]. Ratnakumar et al. verified the effect of sulfur additives on limiting Ni growth during cycling and determined the optimal amount between 0.5 and 1.0 wt.% [34]. Prakash et al. identified the nickel chloride layer formation during charging as another factor in the nickel cathode degradation [35]. Although the NiCl2 layer initially increased the area capacity of a nonporous nickel electrode by decreasing the impedance between 200 °C and 225 °C, the highly resistive surface layer ultimately led to an accumulation of unutilized NaCl in the chloroaluminate melt and limited the areal capacity at higher temperatures.
In addition to sulfur, researchers explored other additives including NaI/NaBr, Al, FeS, and Ni3S2.
By doping the NaAlCl4 melt with NaI or NaBr, NiBr2 or NiI2 with larger lattice parameters than NiCl2 formed in the surface layer. Those surface phases could offer faster mass transport and in turn improve the nickel utilization [32].
Al participated in the electrochemical reaction at around 1.58 V (Figure 2a,b) and alleviated the sudden polarization at the end-of-discharge by the following reaction, (Equation (2) Furthermore, the dissolution of Al via Equation (2) left fine pores inside the cathode, which enhanced high-rate performance of the Na-NiCl2 battery [11]. FeS showed signature intermediate plateaus in the voltage profile (Figure 2a,b) and 6Na + 4FeCl 2 (charged state), 25 100% SOC, Overall reaction: 2NaCl + Fe To circumvent the handling of anhydrous nickel chloride and highly reactive sod associated with the charged state, a typical Na-NiCl2 battery is usually assembled i discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid trolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back thro the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with redu of NiCl2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple as bly, and good cycle life of a Na-NiCl2 battery, research efforts have explored the ele chemical mechanism of Na-NiCl2 cells in depth and identified factors limiting the cy performance, including passivation from nickel chloride [22,32], dissolution loss of n in the NaAlCl4 melt [32], and capacity decline due to Ni/NaCl particle growth. In respo various chemical additives were introduced.
Bones et al. witnessed severe early capacity loss at 250 °C (~50% in <10 cycles), w related to the growth of the nickel grains [33]. At the same time, they found that s additive in NaAlCl4 melt could develop an active, metallic phase with high surface on the cathode. As a result, the long-term cyclability improved, retaining >75% of cap over 2000 cycles with a negligible increase in the cell resistance [33]. Ratnakumar verified the effect of sulfur additives on limiting Ni growth during cycling and determ the optimal amount between 0.5 and 1.0 wt.% [34]. Prakash et al. identified the n chloride layer formation during charging as another factor in the nickel cathode degrad [35]. Although the NiCl2 layer initially increased the area capacity of a nonporous n electrode by decreasing the impedance between 200 °C and 225 °C, the highly resi surface layer ultimately led to an accumulation of unutilized NaCl in the chloroalum melt and limited the areal capacity at higher temperatures.
In addition to sulfur, researchers explored other additives including NaI/NaBr FeS, and Ni3S2.
By doping the NaAlCl4 melt with NaI or NaBr, NiBr2 or NiI2 with la lattice parameters than NiCl2 formed in the surface layer. Those surface phases could faster mass transport and in turn improve the nickel utilization [32].
Al participated in the electrochemical reaction at around 1.58 V (Figure 2a,b) alleviated the sudden polarization at the end-of-discharge by the following reac (Equation (2)) 4NaCl + Al ⇄ 3Na + NaAlCl4, Furthermore, the dissolution of Al via Equation (2) left fine pores inside the cath which enhanced high-rate performance of the Na-NiCl2 battery [11].
2Na + FeCl 2 , (7) to the handling of hazardous/sensitive materials like reduced Fe, FeCl2, or metallic sodium. Recently, Li et al. proposed an easy discharged-state assembly of a Na-FeCl2 battery utilizing sulfur-based additives [26]. As shown in Figure 7b, the cell with 4 mol.% S addition showed a significant reduction of Fe 3+ related (Fe2O3) peaks by XPS, indicating the successful removal of oxide layers on the Fe cathode during the cell conditional cycling. The oxide removal not only freed Fe metal surface for Fe/FeCl2 electrochemical conversion (Figure 7c), but also reduced cell resistance to lower overall cell polarization. As a result, the cell using an optimal amount of sulfur additives exhibited excellent cycling performance with no capacity fade and minimum end-of-charge (EOD) voltage degradation up to 100 cycles [26].

Fast-Charging Capability
In addition to the low cost and abundance of raw materials (Fe and NaCl), Na-FeCl2 batteries have also been demonstrated with fast-charging capability [27]. Following the successful self-activation of a Fe powder cathode enabled by sulfur-based additives, Zhan et al. recently reported a high-rate Na-FeCl2 battery operated at 190 °C with voltage profile shown in Figure 8a. NaCl first transformed to Na6FeCl8 (Equation (5)) then to FeCl2 (Equation (6)) during the charging process and overall reaction can be written as Equation (7): Na6FeCl8 + 3Fe ⇄ 6Na + 4FeCl2 (charged state), 25100% SOC, Overall reaction: 2NaCl + Fe ⇄ 2Na + FeCl2,  As shown in Figure 8b, NaCl and Na 6 FeCl 8 shared the same rock-salt structure, despite the latter having a unit cell twice that of the former. Meanwhile, Na 6 FeCl 8 and FeCl 2 resemble each other in the arrangement of Cl − ions and alternating Fe 2+ occupancy, which in part led to fast Na 6 FeCl 8 /FeCl 2 phase transitions with minimal rearrangement of the atoms. These structural similarities among the three phases facilitated the charging reaction. In addition, CV and Tafel measurements indicated that the Fe/Fe 2+ redox couple possessed intrinsically faster kinetics than the Ni/Ni 2+ . As a result, the Na-FeCl 2 battery delivered unprecedented high-rate performance (Figure 8c), with a capacity of 116 mAh/g at an extremely high current of 100 mA (33.3 mA/cm 2 ; 0.63C). When expressed as normalized capacity, this Na-FeCl 2 battery doubled the conventional Na-NiCl 2 battery at the same rate (Figure 8d). More encouragingly, the optimized Fe/NaCl cathode exhibited excellent cycling stability (Figure 8e) by maintaining a discharge energy density of over 135 mAh/g (295 Wh/kg) for 200 cycles at 10 mA/cm 2 (≈C/5) [27].

Effect of Electrode Morphology and Fe/Ni-Mixed Cathodes
Like the Na-NiCl 2 system, controlling the morphology of Fe particles was vital to maintaining a good electron conducting network and ensuring high active material utilization in a Na-FeCl 2 battery. Ahn et al. prepared Na-(Ni, Fe)Cl 2 cells with different microstructure designs and investigated the electrochemical behaviors of the composites [52]. Their findings suggested that the conducting metal (higher electrochemical potential, Ni in this case) should have smaller particle sizes than the active metal (Fe in this case) in order to preserve electrical connection in the cathode. With a higher weight ratio of Fe in the composite, the resulting thinner passivating surface NiCl 2 layer led to better cycling performance [52]. In the aforementioned work conducted by Zhan et al., the pristine Fe cathode (with only Fe and NaCl) showed very early capacity fading. SEM studies revealed considerable pulverization of Fe particles [27]. ICP-OES tests on multiple relevant phases revealed that FeCl 2 had the highest solubility (0.06 mol%) in NaAlCl 4 , which was twice that of Na 6 FeCl 8 (0.03 mol%) and 20 times that of NiCl 2 (0.002 mol%). It was thus proposed during the solid-solid conversion from FeCl 2 to Na 6 FeCl 8 and finally to Fe at discharge, that reduction of dissolved species from the liquid electrolyte formed fine Fe particles. These Fe particles, smaller in size with undefined precipitation locations, were more likely to depart from the electron-conduction network and become electrochemically inactive. Consequently, the proportion of active Fe particles gradually decreased with cycling, leading to the observed capacity fading. Adding 10 wt% Ni strengthened the electron-conducting chain (Figure 7c), which successfully inhibited the Fe particle pulverization and enabled stable long-term cycling (Figure 8e) [27].

Other Transition Metal Halide Cathode Materials
In addition to extensive development of cathodes based on ferrous and nickel chlorides, other transition-metals have also been explored. In an early study, Ratnakumar et al. demonstrated that the feasibility of a metal chloride as a cathode material was predictable by qualitatively estimating the solubility in NaAlCl 4 molten electrolyte through CV [49]. Specifically, the higher oxidation peak current than the corresponding reduction implied either a porous deposit or high solubility of the oxidation product [49]. Based on this criterion, they conducted screening tests for several candidates including manganese, chromium, aluminum, silver, titanium, molybdenum, and cobalt, among which only Co and Mo chlorides displayed suitability at 250 • C with open-circuit voltages of 2.55 V and 2.64 V, respectively [53]. Parthasarathy et al. first reported a high-temperature (350 • C) Na-ZnCl 2 battery using Zn cathode and a molten NaCl-ZnCl 2 with over 33 mol% of ZnCl 2 as the catholyte/cathode. Although the electrochemical study was limited, there was no evidence of ion exchange with BASE, which implied the viability of zinc chloride as a cathode. In 2013, Lu et al. investigated the behavior and performance of a low-cost Na-ZnCl 2 battery in a planar configuration with BASE [25]. Based on the phase diagram in Figure 9a and the voltage profiles of Na-ZnCl 2 cells in Figure 9b, the following four reversible stepwise reactions (Equations (8)-(11)) were proposed and validated by phase diagram calculations (NaCl utilizations) and XRD phases analysis.
2NaCl + Ni (discharged state) ⇄ 2Na + NiCl2 (charged state), To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reduction of NiCl2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple assembly, and good cycle life of a Na-NiCl2 battery, research efforts have explored the electrochemical mechanism of Na-NiCl2 cells in depth and identified factors limiting the cycling performance, including passivation from nickel chloride [22,32], dissolution loss of nickel in the NaAlCl4 melt [32], and capacity decline due to Ni/NaCl particle growth. In response, various chemical additives were introduced.
Bones et al. witnessed severe early capacity loss at 250 °C (~50% in <10 cycles), which related to the growth of the nickel grains [33]. At the same time, they found that sulfur additive in NaAlCl4 melt could develop an active, metallic phase with high surface area on the cathode. As a result, the long-term cyclability improved, retaining >75% of capacity over 2000 cycles with a negligible increase in the cell resistance [33]. Ratnakumar et al. verified the effect of sulfur additives on limiting Ni growth during cycling and determined the optimal amount between 0.5 and 1.0 wt.% [34]. Prakash et al. identified the nickel chloride layer formation during charging as another factor in the nickel cathode degradation [35]. Although the NiCl2 layer initially increased the area capacity of a nonporous nickel electrode by decreasing the impedance between 200 °C and 225 °C, the highly resistive surface layer ultimately led to an accumulation of unutilized NaCl in the chloroaluminate melt and limited the areal capacity at higher temperatures.
In addition to sulfur, researchers explored other additives including NaI/NaBr, Al, FeS, and Ni3S2. By doping the NaAlCl4 melt with NaI or NaBr, NiBr2 or NiI2 with larger lattice parameters than NiCl2 formed in the surface layer. Those surface phases could offer faster mass transport and in turn improve the nickel utilization [32].
Al participated in the electrochemical reaction at around 1.58 V (Figure 2a,b) and alleviated the sudden polarization at the end-of-discharge by the following reaction, (Equation (2)) Furthermore, the dissolution of Al via Equation (2) left fine pores inside the cathode, which enhanced high-rate performance of the Na-NiCl2 battery [11].
FeS showed signature intermediate plateaus in the voltage profile (Figure 2a,b) and could stabilize the nickel grain size similar to elemental sulfur additives [11]. Li et al. systematically investigated the role of FeS in initial activation and subsequent degradation in Na-NiCl2 cells. Figure 2c,d shows the first-cycle charge-discharge curves of cells with no FeS (0X) and 1 mol% FeS (1X) loadings. The voltage of a cell without FeS quickly rose to 2.8 V at 10 mA, with no observed plateau at ~2.6 V corresponding to Ni/Ni 2+ . X-ray photoelectron spectroscopy (XPS) analysis on raw Ni powder indicated a passivating surface layer (oxides and hydroxides), which not only increased the ohmic resistance but also decreased the electrochemically active surface area (Figure 2e,f). The addition of FeS largely storage application, with the overall redox reaction below with E0 = 2.58 V at 300 °C (Equation (1)) [30,31].
To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reduction of NiCl2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple assembly, and good cycle life of a Na-NiCl2 battery, research efforts have explored the electrochemical mechanism of Na-NiCl2 cells in depth and identified factors limiting the cycling performance, including passivation from nickel chloride [22,32], dissolution loss of nickel in the NaAlCl4 melt [32], and capacity decline due to Ni/NaCl particle growth. In response, various chemical additives were introduced.
Bones et al. witnessed severe early capacity loss at 250 °C (~50% in <10 cycles), which related to the growth of the nickel grains [33]. At the same time, they found that sulfur additive in NaAlCl4 melt could develop an active, metallic phase with high surface area on the cathode. As a result, the long-term cyclability improved, retaining >75% of capacity over 2000 cycles with a negligible increase in the cell resistance [33]. Ratnakumar et al. verified the effect of sulfur additives on limiting Ni growth during cycling and determined the optimal amount between 0.5 and 1.0 wt.% [34]. Prakash et al. identified the nickel chloride layer formation during charging as another factor in the nickel cathode degradation [35]. Although the NiCl2 layer initially increased the area capacity of a nonporous nickel electrode by decreasing the impedance between 200 °C and 225 °C, the highly resistive surface layer ultimately led to an accumulation of unutilized NaCl in the chloroaluminate melt and limited the areal capacity at higher temperatures.
In addition to sulfur, researchers explored other additives including NaI/NaBr, Al, FeS, and Ni3S2.
By doping the NaAlCl4 melt with NaI or NaBr, NiBr2 or NiI2 with larger lattice parameters than NiCl2 formed in the surface layer. Those surface phases could offer faster mass transport and in turn improve the nickel utilization [32].
Al participated in the electrochemical reaction at around 1.58 V (Figure 2a,b) and alleviated the sudden polarization at the end-of-discharge by the following reaction, (Equation (2)) 4NaCl + Al ⇄ 3Na + NaAlCl4, Furthermore, the dissolution of Al via Equation (2) left fine pores inside the cathode, which enhanced high-rate performance of the Na-NiCl2 battery [11]. FeS showed signature intermediate plateaus in the voltage profile (Figure 2a,b) and could stabilize the nickel grain size similar to elemental sulfur additives [11]. Li et al. systematically investigated the role of FeS in initial activation and subsequent degradation in Na-NiCl2 cells. Figure 2c,d shows the first-cycle charge-discharge curves of cells with no FeS (0X) and 1 mol% FeS (1X) loadings. The voltage of a cell without FeS quickly rose to 2.8 V at 10 mA, with no observed plateau at ~2.6 V corresponding to Ni/Ni 2+ . X-ray photoelectron spectroscopy (XPS) analysis on raw Ni powder indicated a passivating surface layer (oxides and hydroxides), which not only increased the ohmic resistance but also decreased the electrochemically active surface area (Figure 2e,f). The addition of FeS largely Salt liquid (ZnCl 2 : 62 mol%) + Na (E~2.07 V), storage application, with the overall redox reaction below with E0 = 2.58 V at 300 °C (Equation (1)) [30,31].
To circumvent the handling of anhydrous nickel chloride and highly reactive sodium associated with the charged state, a typical Na-NiCl2 battery is usually assembled in the discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, and a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid electrolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode side, oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back through the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reduction of NiCl2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple assembly, and good cycle life of a Na-NiCl2 battery, research efforts have explored the electrochemical mechanism of Na-NiCl2 cells in depth and identified factors limiting the cycling performance, including passivation from nickel chloride [22,32], dissolution loss of nickel in the NaAlCl4 melt [32], and capacity decline due to Ni/NaCl particle growth. In response, various chemical additives were introduced.
Bones et al. witnessed severe early capacity loss at 250 °C (~50% in <10 cycles), which related to the growth of the nickel grains [33]. At the same time, they found that sulfur additive in NaAlCl4 melt could develop an active, metallic phase with high surface area on the cathode. As a result, the long-term cyclability improved, retaining >75% of capacity over 2000 cycles with a negligible increase in the cell resistance [33]. Ratnakumar et al. verified the effect of sulfur additives on limiting Ni growth during cycling and determined the optimal amount between 0.5 and 1.0 wt.% [34]. Prakash et al. identified the nickel chloride layer formation during charging as another factor in the nickel cathode degradation [35]. Although the NiCl2 layer initially increased the area capacity of a nonporous nickel electrode by decreasing the impedance between 200 °C and 225 °C, the highly resistive surface layer ultimately led to an accumulation of unutilized NaCl in the chloroaluminate melt and limited the areal capacity at higher temperatures.
In addition to sulfur, researchers explored other additives including NaI/NaBr, Al, FeS, and Ni3S2.
By doping the NaAlCl4 melt with NaI or NaBr, NiBr2 or NiI2 with larger lattice parameters than NiCl2 formed in the surface layer. Those surface phases could offer faster mass transport and in turn improve the nickel utilization [32].
Al participated in the electrochemical reaction at around 1.58 V (Figure 2a,b) and alleviated the sudden polarization at the end-of-discharge by the following reaction, (Equation (2)) 4NaCl + Al ⇄ 3Na + NaAlCl4, Furthermore, the dissolution of Al via Equation (2) left fine pores inside the cathode, which enhanced high-rate performance of the Na-NiCl2 battery [11]. FeS showed signature intermediate plateaus in the voltage profile (Figure 2a,b) and could stabilize the nickel grain size similar to elemental sulfur additives [11]. Li et al. systematically investigated the role of FeS in initial activation and subsequent degradation in Na-NiCl2 cells. Figure 2c,d shows the first-cycle charge-discharge curves of cells with no FeS (0X) and 1 mol% FeS (1X) loadings. The voltage of a cell without FeS quickly rose to 2.8 V at 10 mA, with no observed plateau at ~2.6 V corresponding to Ni/Ni 2+ . X-ray photoelectron spectroscopy (XPS) analysis on raw Ni powder indicated a passivating surface layer (oxides and hydroxides), which not only increased the ohmic resistance but also decreased the electrochemically active surface area (Figure 2e,f) (1)) [30,31].
To circumvent the handling of anhydrous nickel chloride and highly reactive sodiu associated with the charged state, a typical Na-NiCl2 battery is usually assembled in t discharged state with nickel powder, sodium chloride, sodium tetrachloroaluminate, an a few additives [10,11]. During charging, Na + ions diffuse through the β"-Al2O3 solid ele trolyte (BASE) and are reduced at the anode side as sodium metal. On the cathode sid oxidized NiCl2 coats the Ni surface. During discharging, Na + ions shuttle back throug the BASE electrolyte and form NaCl with Cl − ions from NiCl2 concurrent with reductio of NiCl2 to Ni.

Electrochemical Mechanism and Additives
To take advantage of the high cell voltage, superior energy efficiency, simple assem bly, and good cycle life of a Na-NiCl2 battery, research efforts have explored the electr chemical mechanism of Na-NiCl2 cells in depth and identified factors limiting the cyclin performance, including passivation from nickel chloride [22,32], dissolution loss of nick in the NaAlCl4 melt [32], and capacity decline due to Ni/NaCl particle growth. In respons various chemical additives were introduced.
Bones et al. witnessed severe early capacity loss at 250 °C (~50% in <10 cycles), whi related to the growth of the nickel grains [33]. At the same time, they found that sulf additive in NaAlCl4 melt could develop an active, metallic phase with high surface ar on the cathode. As a result, the long-term cyclability improved, retaining >75% of capaci over 2000 cycles with a negligible increase in the cell resistance [33]. Ratnakumar et verified the effect of sulfur additives on limiting Ni growth during cycling and determin the optimal amount between 0.5 and 1.0 wt.% [34]. Prakash et al. identified the nick chloride layer formation during charging as another factor in the nickel cathode degradati [35]. Although the NiCl2 layer initially increased the area capacity of a nonporous nick electrode by decreasing the impedance between 200 °C and 225 °C, the highly resisti surface layer ultimately led to an accumulation of unutilized NaCl in the chloroalumina melt and limited the areal capacity at higher temperatures.
In addition to sulfur, researchers explored other additives including NaI/NaBr, A FeS, and Ni3S2.
By doping the NaAlCl4 melt with NaI or NaBr, NiBr2 or NiI2 with larg lattice parameters than NiCl2 formed in the surface layer. Those surface phases could off faster mass transport and in turn improve the nickel utilization [32].
Al participated in the electrochemical reaction at around 1.58 V (Figure 2a,b) an alleviated the sudden polarization at the end-of-discharge by the following reactio (Equation (2)) 4NaCl + Al ⇄ 3Na + NaAlCl4, Furthermore, the dissolution of Al via Equation (2) left fine pores inside the cathod which enhanced high-rate performance of the Na-NiCl2 battery [11]. FeS showed signature intermediate plateaus in the voltage profile (Figure 2a,b) an could stabilize the nickel grain size similar to elemental sulfur additives [11]. Li et al. sy tematically investigated the role of FeS in initial activation and subsequent degradation Na-NiCl2 cells. Figure 2c,d shows the first-cycle charge-discharge curves of cells with n FeS (0X) and 1 mol% FeS (1X) loadings. The voltage of a cell without FeS quickly rose 2.8 V at 10 mA, with no observed plateau at ~2.6 V corresponding to Ni/Ni 2+ . X-ray ph toelectron spectroscopy (XPS) analysis on raw Ni powder indicated a passivating surfa layer (oxides and hydroxides), which not only increased the ohmic resistance but also d creased the electrochemically active surface area (Figure 2e,f). The addition of FeS large CuCl2 is also an attractive cathode material because of its high theoretical capacity of 400 mAh/g, with equilibrium potentials of 3.40 V for the CuCl2/CuCl redox couple and 2.74 V for the CuCl/Cu redox couple in Na-based battery systems [24,55,56]. However, CuCl2 is soluble in conventional organic-based liquid electrolytes and high-temperature molten salt electrolytes such as NaAlCl4, hindering its practical application. In 2014, Yang et al. proposed an advanced intermediate-temperature Na/CuCl2 battery using a NaCl-EMIC-AlCl3 catholyte and a 500 mm thick BASE film separator operating at only 150 °C [57]. Unfortunately, the study did not report cyclability or identify the fundamental parameters such as the stability of CuCl2. In the search for suitable electrolytes compatible with CuCl2, Kim et al. constructed a new ZEBRA-type room temperature Na-CuCl2 battery system with SO2-based nonflammable inorganic electrolyte. The Na-CuCl2 cell showed outstanding battery performance with a high energy density of ~580 Wh/kg based on the cathode material, a more than 97% round-trip efficiency, as well as a superior capacity retention over 1000 cycles. These excellent performances were attributed to the SO2based inorganic electrolyte, which exhibited high ionic conductivity and high chemical/electrochemical compatibility with the CuCl2 cathode material [55].
As a final comment, a major driving force behind the intermediate-temperature sodium battery is that sodium is significantly more abundant than lithium, which provides an economic incentive for long-duration and thus low-cycle battery applications. While other metals can participate in the conversion chemistry as shown above, the scale of the Upon charging, NaCl first reacted with Zn to form Na on the anode side and Na 2 ZnCl 4 on the cathode side. Once all the NaCl was consumed, Na 2 ZnCl 4 and Zn reacted to form the NaCl-ZnCl 2 liquid phase. At the end of charge, the liquid phase reacted with Zn to generate solid ZnCl 2 . The authors further identified the effect of the liquid-phase formation on electrochemical performance by testing cells at 280 • C and 240 • C. At 280 • C, cells revealed stable cyclability with the liquid phase formation during cycling. However, a rapid rise in polarization was observed at 240 • C where only a solid-state reaction occurred. SEM analysis indicated that the growth of Zn and NaCl particles was suppressed at 280 • C, where Zn and NaCl precipitated during discharge [25]. Later, Lu et al. investigated the performance and reaction mechanism of a Na-ZnCl 2 battery at 190 • C. As predicted by the phase diagram (Figure 9a), two stepwise reversible reactions ( Figure 9d) were identified, and the corresponding reaction mechanism was illustrated in Figure 9e. First, NaCl reacted with Zn to produce a ribbon-like Na 2 ZnCl 4 layer. This layer existed at the NaCl-Zn interface rather than covering the Zn particles' surfaces, resulting in excellent rate capability (Figure 9c). Subsequently, the Na 2 ZnCl 4 transformed gradually into ZnCl 2 , which covered the surfaces of Zn particles. The authors suggested that the passivation by ZnCl 2 during the second step blocked the electron pathway of the NaCl/Zn cathodes and limited the overall rate performance of the battery [24]. More recently, Lee et al. reported an electrochemically activated Na-ZnCl 2 battery at 260 • C using a carbon matrix in the cathode compartment. By using inexpensive carbon felt to maintain efficient electron percolation in the cathode, the cell realized lower charge transfer resistance and higher capacity, as well as lower materials cost [54].
CuCl 2 is also an attractive cathode material because of its high theoretical capacity of 400 mAh/g, with equilibrium potentials of 3.40 V for the CuCl 2 /CuCl redox couple and 2.74 V for the CuCl/Cu redox couple in Na-based battery systems [24,55,56]. However, CuCl 2 is soluble in conventional organic-based liquid electrolytes and high-temperature molten salt electrolytes such as NaAlCl 4 , hindering its practical application. In 2014, Yang et al. proposed an advanced intermediate-temperature Na/CuCl 2 battery using a NaCl-EMIC-AlCl 3 catholyte and a 500 mm thick BASE film separator operating at only 150 • C [57]. Unfortunately, the study did not report cyclability or identify the fundamental parameters such as the stability of CuCl 2 . In the search for suitable electrolytes compatible with CuCl 2 , Kim et al. constructed a new ZEBRA-type room temperature Na-CuCl 2 battery system with SO 2 -based nonflammable inorganic electrolyte. The Na-CuCl 2 cell showed outstanding battery performance with a high energy density of~580 Wh/kg based on the cathode material, a more than 97% round-trip efficiency, as well as a superior capacity retention over 1000 cycles. These excellent performances were attributed to the SO 2 -based inorganic electrolyte, which exhibited high ionic conductivity and high chemical/electrochemical compatibility with the CuCl 2 cathode material [55].
As a final comment, a major driving force behind the intermediate-temperature sodium battery is that sodium is significantly more abundant than lithium, which provides an economic incentive for long-duration and thus low-cycle battery applications. While other metals can participate in the conversion chemistry as shown above, the scale of the upcoming energy storage target prefers an inexpensive cathode metal where the chemistry is well-understood, such as Ni or Fe.

Conclusions
Replacing fossil fuel-based power generation with widespread deployment of renewable energy sources is a crucial technological goal for reducing greenhouse gas emission. For this to occur, the inherent intermittency of renewables such as wind and solar must be addressed by some form of energy storage. Electrochemical energy storage using batteries is a promising option, with various storage technologies such as LIBs, Na-S batteries, RFB, and also Na-MH batteries. For such technologies to be adopted, they must have desirable performance metrics such as high voltage, high specific energy density, high-power density, and superior safety and reliability while also being inexpensive enough to be economical. Of these options, Na-MH batteries are promising due to their high operating voltage, low cost of materials, good safety in the case of cell failure, and ability to be assembled in the discharged state. However, challenges regarding high operating temperature, high cost of materials of construction, and the high cost of Ni has motivated developing Na-MH batteries with low-cost transition metal cathodes and intermediate-temperature operation.
In this review, Na-NiCl 2 batteries have been discussed with a particular focus on recent work that addresses some of the challenges of conventional Na-MH batteries, specifically by using lower operating temperatures (e.g., 190 • C rather than >265 • C or higher), identifying methods to reduce the amount of expensive Ni required, employing lower-cost materials for sealing (enabled by lower temperatures), additives that improve otherwise sluggish kinetics at lower operating temperatures, and others.
Moving beyond expensive Ni, other transition metal cathodes such as Fe, Zn, or Cu, are of great interest due to the greater abundance and lower cost of raw materials. Some of these battery chemistries have been discussed in this review, showing comparable performance to widely studied Na-NiCl 2 batteries but with lower-cost active materials. In some cases, these chemistries even offer superior performance compared to Na-NiCl 2 cells such as the high rate capability of Na-FeCl 2 cells or the high theoretical specific capacity of the Na-CuCl 2 cells discussed herein. While each of these transition metal battery chemistries have some general similarities to the Na-NiCl 2 system, each also has unique benefits and challenges. Although there have been valuable explorations of these non-Ni cathode systems to date, there is still much room for continued investigation to identify further potential improvements. For example, further reduction in operating temperature (e.g., <120 • C) of Na-MH batteries can expand the potential materials of construction while requiring less thermal energy management and improving safety. Such temperature reductions will require new innovations to improve Na-wettability on the anode side of BASE as well as improvements in catholyte composition to reduce the melting point while maintaining good kinetics and mass transfer on the cathode side. Potentially, interface engineering of the Na-BASE solid interface could enable high-performance solid-sodium Na-MH batteries operating at temperatures below 100 • C. Finally, the demonstration of large area planar Na-MH cells is a crucial step to scaling up this promising battery concept from the lab to the grid scale.
We expect that continued research efforts to enable intermediate-temperature Na-MH batteries with low-cost cathodes will allow battery performance that can surpass current state-of-the-art molten Na batteries while also providing much lower overall energy storage costs. By improving performance and reducing cost, not to mention taking advantage of future economies of scale, intermediate-temperature Na-MH batteries will be a critical component of an economical, renewable power grid.