Introduction

In order to address recent concerns on the limited resources of lithium and the localized reserves, Na rechargeable batteries have gained much attention as alternative power sources to replace Li rechargeable ones. Up to date, several types of Na rechargeable batteries have been investigated, such as, high-temperature Na–S (NAS) and Na–NiCl2 (ZEBRA) batteries, room-temperature Na-ion and Na–O2 batteries and each system has the pros and cons; more detailed materials and technology for the Na rechargeable batteries are well discussed in other recent works1,2,3,4,5,6,7,8. Herein, we demonstrate a new type of Na rechargeable battery using an SO2-based inorganic molten complex as both (i) a Na+-conducting medium and (ii) cathode material, i.e. catholyte, suggesting as an alternative room-temperature and high-energy Na rechargeable battery. In the history of batteries, SO2 is not a strange material for Li batteries. Primary Li–SO2 batteries in which liquefied SO2 serves as the active cathode material have been commercialized for military and industrial applications9,10 and ongoing interest for further development is still found11. About 30 years ago, there were also intensive studies on Li–SO2 rechargeable batteries based on a LiAlCl4⋅xSO2 inorganic molten complex catholyte, which shows completely different reaction chemistry from the primary SO2 battery8,12,13,14,15,16. Rechargeable Li–SO2 battery showed a discharge capacity of ~1000 mAh g−1 based on the carbon electrode (theoretical catholyte capacity of 144 mAh g−1 for LiAlCl4⋅6SO2) with an operating voltage of 3.2 V and Duracell demonstrated the performance of prototype C-size Li–SO2 rechargeable batteries12. One of the key advantages of Li–SO2 rechargeable battery is the use of a highly conductive electrolyte (~0.1 S cm−1 at room temperature)17, which is almost same to typical ionic conductivity of aqueous electrolytes. This excellent conductivity of Li+ ensures a high electrochemical reversibility and rate capability of rechargeable Li–SO2 battery system. Moreover, the inorganic electrolyte employed in Li–SO2 battery is non-flammable18, offering additional attractive feature to Li–SO2 battery over other flammable organic electrolyte-based Li batteries. The Li–SO2 rechargeable battery, however, could not succeed in its commercialization for consumer application, mainly because the use of LiAlCl4⋅6SO2 resulted into high internal cell pressure, which raised the safety concerns about cell venting even under moderate cycling condition12,19; specifically, the equilibrium vapor pressure of SO2 for LiAlCl4⋅6SO2 is about 2 bar at 20 °C (~7 bar at 60 °C) and, moreover, the LiAlCl4⋅6SO2 catholyte releases SO2 gas when the cell discharges in accordance with the reaction chemistry12,13,17. Otherwise, LiAlCl4⋅3SO2 (not 6SO2) shows relatively low equilibrium vapor pressure of ~1 bar at 20 °C (~2 bar at 60 °C) and involves no change in the cell internal pressure during discharge, which could be a better composition having a safety advantage13,17. It, however, crystallizes when cooled to about –10 °C and remains a solid even after heated up to 25 °C, because a solid phase of LiAlCl4⋅3SO2 is more stable for than its liquid phase at ambient temperature17,20.

NaAlCl4⋅xSO2, a homologue of LiAlCl4⋅xSO2, was introduced by Kühnl et al. in the 1970s21,22 as a highly conductive Na+ electrolyte (~0.1 S cm−1); however, this electrolyte has never been explored for Na rechargeable batteries although it has distinguishable properties from LiAlCl4⋅xSO2. NaAlCl4⋅2SO2, known as a stable composition under ambient conditions22, exhibits the equilibrium vapor pressure of ~1 bar and remains as a liquid phase up to –40 °C without freezing, thus alleviating our safety concerns regarding cell venting that was a critical issue in the past Li–SO2 battery17,21,22. These physicochemical properties of NaAlCl4⋅2SO2 motivated us to study and develop a Na–SO2 battery system, particularly for low-cost stationary power storage applications. Here we report a Na–SO2 rechargeable battery system using NaAlCl4⋅2SO2 electrolyte. We found that the optimized carbon cathode enables a reversible reaction of the catholyte with high capacity, good rate capability, a long life-span over 300 cycles and an estimated theoretical energy density of 407 Wh kg−1 (based on the discharged product including carbon cathode). This value is comparable with those of other high-energy Na rechargeable batteries. Moreover, non-flammability, low vapor pressure and unique self-regeneration mechanism of the inorganic electrolyte presented here would be noteworthy merits of Na–SO2 system over other Na rechargeable battery systems.

Results

In this work, we constructed a 2032 coin-type Na–SO2 cell using a Na-metal anode and a porous carbon cathode with NaAlCl4⋅2SO2 as a catholyte. The carbon cathode was prepared by roll-pressing of ketjenblack/polytetrafluoroethylene paste on a Ni-mesh and NaAlCl4⋅2SO2 was synthesized by blowing SO2 gas through a mixture of NaCl and AlCl3 powders. Details of materials and experimental methods are described in Methods. Figure 1a shows the first and second voltage profiles of the Na–SO2 cell that delivers a discharge capacity of ~1800 mAh g−1 based on the carbon cathode at a rate of 0.1C (=150 mA g−1 or 0.34 mA cm−2). This corresponds to an areal capacity of 4.1 mAh cm−2, which is comparable to typical values of commercial Li-ion batteries (3–5 mAh cm−2) and much higher than those of reported Li–O2 and Na–O2 batteries6,23,24,25,26. The Na–SO2 cell also showed an encouraging rate capability exhibited in Fig. 1b, where a high capacity of 897 mAh g−1 is observed even at a significantly high current density of 5C (7500 mA g−1 or 17 mA cm−2). Given that the rate capability is one of the most challenging issues in NAS, ZEBRA and Na–O2 batteries1,23,24, the excellent power capability could give the Na–SO2 battery a critical edge over other Na rechargeable batteries previously reported. It should be also noted that the operating voltage of the Na–SO2 cell was ~3.0 V at 0.1 C. It is higher than those of NAS (2.0 V), ZEBRA (2.58 V) and Na–O2 (2.2–2.5 V) and also comparable to those of most Na-ion battery cathodes1,2,3,4,5,6,7,26. However, the voltage gap between discharge and charge was evident in the Na–SO2 system and the low round-trip energy efficiency (~80%) needs to be further ameliorated. The Na–SO2 cell showed relatively good capacity retention during cycling, i.e. 75% of the initial capacity after 100 cycles (Fig. 1c), even under full depth-of-discharge condition, accompanied by high columbic efficiencies during cycling (average of ~99%).

Figure 1
figure 1

Electrochemical performance of Na–SO2 cells.

(a) The first and second galvanostatic voltage profiles of a Na–SO2 cell at 0.1C (=150 mA g−1 and 0.34 mA cm−2). The cutoff voltage for charge is 4.05 V. (b) Discharge rate capability of a Na–SO2 cell. The charging rate is fixed at 0.2C. The capacity retention during the rate capability test was displayed in Supplementary Figure S1. (c) Cycle performance of a Na–SO2 cell at a rate of 0.5C discharge and 0.2C charge for 100 cycles.

Full size image

It is generally accepted that the underlying reaction mechanism of the SO2-based catholyte is reversible changes in the oxidation state of sulfur in SO2 between +4 and +3 (ref. 13 and 14). The detailed reaction chemistry is, however, still unclear. To elucidate the electrochemical reaction responsible for the exceptional performance of the Na–SO2 cell, we performed various in situ and ex situ analyses. Figure 2a shows in situ X-ray diffraction (XRD) patterns of the carbon cathode during the first cycle. On discharge, new peaks corresponding to NaCl started to appear and their intensity increased as discharge continued. On subsequent charge, the NaCl peaks diminished gradually until they completely disappeared at the end of the charge. This reversible behavior of NaCl was further confirmed by scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) observations (Fig. 2b–e and Supplementary Figure S2). Well-defined cubic solid discharge products with a size of about 2 μm appeared after the discharge and they vanished gradually during the successive charge. The NaCl crystals formed during discharge are regarded as the products of the electrochemical reduction of SO2 to SO2.− radical anions which displace Cl− from AlCl4− to form NaCl. In a Li–SO2 battery, homologue of Na–SO2, LiCl and LiAlCl(SO2)3 have been considered as discharge products13: 3 moles of reduced SO2.− anions react with 1 mole of AlCl4− sequentially, thereby forming insoluble 3 moles of LiCl and 1 mole of LiAlCl(SO2)3, which precipitate at the carbon cathode. While LiCl was confirmed by XRD analysis, the SO2-substituted second form of the discharge product has not been identified clearly despite several efforts made in various experimental analyses13.

Figure 2
figure 2

In situ XRD and SEM analysis on the Na–SO2 cell cathodes.

(a) In situ XRD patterns of the carbon cathode in a Na–SO2 cell during the first cycle. The referred NaCl corresponds to JCPDS #780751, indicated by arrows. The SEM images of the carbon cathode in a Na–SO2 cell (b) before discharge (pristine), (c) after full discharge, (d) after 50% charge and (e) after full charge.

Full size image

In the Na–SO2 cell, we also detected another type of product composed of Na, Al, Cl, S and O elements at the carbon surface by using SEM-EDS (Supplementary Figure S2). To clarify the reaction pathway of the reduced SO2.− anions and the chemical structure of the resulting discharge products in a Na–SO2 system, we performed an ab initio molecular dynamics (AIMD) simulation, combined with an experimental analysis of surface-enhanced Raman spectroscopy (SERS). As shown in the simulation snapshots of the reaction products that have minimum energy during discharge (Fig. 3a,b), the most stable structure of the discharge products is quadra-coordinated Al species bonded by an oxygen atom to SO2, i.e., NaAlCl2(SO2)2 with NaCl, while penta-coordinated Na2AlCl3(SO2)2 was occasionally observed during AIMD simulation and is considered as a minor discharge product. Consequently, the AIMD simulation sheds light onto how SO2.− radical anions stabilize themselves by displacing chlorine anions from tetrachloroaluminates to form NaCl and SO2-complexes. Regarding to the possible mole number of SO2.− displacing Cl− from 1 mole of AlCl4− during discharge, static first-principles calculations of the substitution reactions (Fig. 3c) support the above AIMD result that the substitution of 2 SO2 into AlCl4−, i.e., the formation of NaAlCl2(SO2)2 is the most feasible reaction. This stoichiometric behavior was experimentally confirmed by investigating the mass-to-charge ratio (m/Q) at the cathode after discharge, where the m/Q is defined as weight gain at the cathode per discharge capacity. The experimental m/Q value from our many repeated measurements was 6.89 ± 0.14 mg mAh−1 which is quite close to the value expected for the exclusive formation of NaAlCl2(SO2)2 and 2NaCl; the formation of NaAlCl2(SO2)2 and 2NaCl would consume 2 SO2 with 2 e− per NaAlCl4 and result in a 6.82 mg mAh−1 cathode weight gain, as indicated by the slope of the red line in Fig. 3d.

Figure 3
figure 3

Theoretical calculation and SERS analysis on the Na–SO2 cell cathodes.

The atomic structures of the discharge products with minimum energy during the AIMD simulation (grey: Al, green: Cl, red: O, yellow: S, purple: Na): (a) NaAlCl2(SO2)2 + 2NaCl, (b) Na2AlCl3(SO2)2 + NaCl. (c) The SO2.− to Cl− substitution reaction energies of NaAlCl4 from static first-principles calculations. (d) Relationship between cathode weight gain and discharge capacity in Na–SO2 cells. Several hypotheses depending on the numbers of SO2 reduced at the cathode were considered, including also Na2S2O4 formation. The theoretical mass-to-charge relationships were displayed as lines and the experimental measurements were plotted by scatters. (e) The observed SERS spectra of the discharged Na–SO2 cell cathode and comparison with the calculated Raman frequency. The numbered Raman frequencies with the vibrational assignment are given in Supplementary Table S1.

Full size image

To identify the SO2-substituted discharge product, we also carried out ex situ SERS measurement to probe the chemical structure of the discharge product and compared the calculated Raman spectra based on the aforementioned reaction mechanism, as presented in Fig. 3e. The observed Raman peaks in the spectral range between 400 and 600 cm−1 and at around 620, 800 and 920 cm−1 (the corresponding vibration modes are described in Supplementary Table S1) are relatively well matched with the calculated ones corresponding to NaAlCl2(SO2)2, suggesting that NaAlCl2(SO2)2 is the most plausible second discharge product formed at the cathode. Putting all the above results together, the full cell reaction scheme of the Na–SO2 rechargeable battery is proposed as follows:

Based on the above reaction, the theoretical capacity of NaAlCl4⋅2SO2 is 168 mAh g−1 (or 147 mAh g−1 based on the discharge products). Nonetheless, since the reaction of the electroactive material is highly dependent on the physicochemical properties of the carbon cathode such as its surface area and pore structure, we estimated the theoretical energy density of a Na–SO2 battery based on the mass of discharge products including carbon cathode. The evaluated energy density is 407 Wh kg−1 (for details, see Supplementary Table S2), which is comparable to other high-energy Na rechargeable batteries6,7.

The capacity fading of the Na–SO2 cell shown in Fig. 1c is mainly attributed to residual insulating discharge products that passivate carbon surface and/or block the pore entrance in the electrode, thereby reducing reaction site and increasing the impedance of the carbon cathode. We observed that NaCl did not disappear completely in the carbon cathode during repeated cycling, so that the accumulated discharge products increased the impedance of the cathode (Supplementary Figure S3 and S4). It is interesting that a tetrachloroaluminate:SO2 complex has an intrinsic self-regeneration mechanism12,27 which can be utilized to remove residual NaCl from the carbon cathode. When a Na–SO2 cell was overcharged to above 4.05 V, the recombination reaction took place as like a Li–SO2. According to the proposed overcharging mechanism for a Li–SO2 system12,27, the oxidation of AlCl4− produces Cl2 and AlCl3 during overcharge. The highly soluble Cl2 gas dissolves into the electrolyte and reacts with the Na-metal anode to form NaCl, which further reacts with AlCl3 to regenerate NaAlCl4. The produced AlCl3 can also react with residual NaCl at the cathode to regenerate NaAlCl4. These recombination reactions during overcharge can facilitate the reactivation of the surface and the pore structure of the carbon cathode, thereby restoring capacity of the Na–SO2 cell. Figure 4a shows the cycle performance before and after overcharging (See also Supplementary Figure S5 for the corresponding voltage profiles). At the 98th cycle under the normal charge/discharge condition, the capacity was below 1000 mAh g−1. Surprisingly, it jumped up about 1250 mAh g−1 after overcharging up to 4.3 V at the 99th cycle and then showed stable cycle performance during the subsequent 100 cycles (~80% of capacity retention for the subsequent 100 cycles). The cell was further cycled up to 350 cycles with another two overcharging processes. After each overcharging process, the capacity came back its initial value and finally retained 80% of the initial capacity (1000 mAh g−1) at the 350th cycle, which exhibits the remarkable long-term cycle performance of Na–SO2 cell. In XRD analysis of the cathode, NaCl peaks were observed after the 50th charge due to the accumulation of NaCl on carbon over repeated cycles. However, these NaCl peaks receded dramatically after the overcharge (Fig. 4b), supporting a cathode recuperation by the above-stated recombination reactions. SEM observations also gave a solid proof for this reaction (Supplementary Figure S6).

Figure 4
figure 4

Long-term cycle performance of Na–SO2 cells with overcharging processes.

(a) The cycle performance of a Na–SO2 cell for 350 cycles including overcharging processes. Overcharging was forced at the 99th, 206th and 249th cycles for a given time. The corresponding voltage profiles before and after the 99th overcharging are given in Supplementary Figure S5. (b) Ex situ XRD patterns of the carbon cathode in Na–SO2 cells before and after overcharging: (i) after the first discharge, (ii) after the first charge, (iii) after the 50th charge and (iv) after the 51st overcharge (■ NaCl, *SUS holder, # Ni mesh).

Full size image

Finally, it should be emphasized that the reliability of a Na–SO2 battery is a major attractive feature over other Na rechargeable batteries. First of all, a Na–SO2 battery is working at ambient temperature. Considering NAS and ZEBRA batteries need the complicated implementation to ensure durability and safety due to high temperature (~300 °C) operation1, there would be no extra high capital cost for the system construction and also no safety concern about seriously-reactive molten Na anode for the Na–SO2 battery. In comparison with other room-temperature Na-ion or Na–O2 batteries in which flammable organic solvents are normally used, the SO2-based inorganic electrolyte for the Na–SO2 battery is nonflammable, even in direct contact with an open flame (Fig. 5a,b). This self-extinguishing property of the electrolyte could significantly relieve the safety concerns about cell ignition or explosion of a Na–SO2 battery. Another important feature of a Na–SO2 battery stems from the still high Na+ conductivity of NaAlCl4⋅2SO2 at low temperatures (Fig. 5c). Owing to the excellent conductivity, a Na–SO2 cell could deliver a capacity of 1270 mAh g−1 and 830 mAh g−1 at 0 °C and –20 °C, respectively (Supplementary Figure S7). This reasonable low temperature performance with the NaAlCl4⋅2SO2 obviates a need for further increase of SO2 in case of a Na–SO2 system unlike a Li–SO2 and therefore, another safety concern regarding cell venting that was a critical issue in the past Li–SO2 battery could be relieved in a Na–SO2 battery. Figure 5d exhibits the vapor pressure of a NaAlCl4⋅2SO2 electrolyte at various temperatures. The vapor pressure at room-temperature is <1 bar (also, ~2 bar at 60 °C) and significantly lower than those of LiAlCl4⋅6SO2 and pure liquid SO217.

Figure 5
figure 5

Physical and electrical properties of NaAlCl4.2SO2 inorganic electrolyte.

Photo-snapshots of flammability-tests for (a) NaAlCl4⋅2SO2 inorganic electrolyte and (b) 1 M NaClO4 in the mixture of propylene carbonate (PC) and ethylene carbonate (EC) as one of organic-based electrolytes. The electrolyte-soaked tissue was forced to be contact with an open flame. (c) Conductivity of NaAlCl4⋅2SO2 electrolyte at various temperatures. (d) Vapor pressure of NaAlCl4⋅2SO2 electrolyte at various temperatures.

Full size image

Discussion

We presented here a 3-V-class Na–SO2 battery delivering high discharge capacity, excellent rate capability and long cycle-life. We firmly believe that these key battery performances of the Na–SO2 system are much more promising compared with other Na rechargeable batteries ever reported. We also demonstrated that the cell chemistry is based on the highly reversible redox reaction of SO2 with tetrachloroaluminate and the use of the NaAlCl4⋅2SO2 inorganic electrolyte enables highly reliable Na–SO2 system in terms of long cycle life as well as safety. For practical application, however, there still remain several problems to be resolved: the large voltage-hysteresis during discharge and charge, instability of Na-metal anodes or search for alternative anode materials, etc. Further studies for fundamental understanding of a Na–SO2 battery, such as clarifying a detailed reaction pathway during charge of a Na–SO2 battery, should be also needed. However, the recent advanced battery-technologies regarding materials, electrodes, cell engineering and also state-of-the-art analytical methods, which have remarkably developed since the advent of lithium-ion batteries, could accelerate our research and development for an advanced Na–SO2 battery, as already observed in the recent research activities for reviving Li(or Na)–O2 and Li(or Na)–S systems23,24,25,26. Considering the many favorable features and promises discussed in this report, the Na–SO2 battery can be a viable system for next cost-effective energy storage system. Further, the SO2-based inorganic electrolyte can be widely applied to battery systems adopting other metallic anodes like Ca, K, Al and Mg, which paves the way for the development of various non-lithium metal-based battery systems.

Methods

Synthesis of NaAlCl4·xSO2 electrolyte

NaCl (>99.9%, Alfa Aesar) was vacuum-dried at 120 °C for 24 hours before using, while anhydrous AlCl3, (99.999%, Alfa Aesar) was used without any purification. The electrolyte was prepared by blowing SO2 gas (anhydrous, Fluka) through a mixture of NaCl and AlCl3 in a glass/Teflon vessel. The molar ratio of NaCl to AlCl3 was 1.1 to avoid the presence of free AlCl3, which is known to be corrosive to alkali metals. As soon as SO2 gas contacted with the mixture, it became liquid of transparent light ocher color. The SO2 gas was blown until the desired SO2 concentration, which was determined by weighing the electrolyte vessel, was reached. The reaction-completed electrolyte vessel was transferred back into the Ar-filled glove box and placed in a glass-bottle containing small pieces of Na metal to remove the possible AlCl3 residue or H2O.

Electrode/cell fabrication

A carbon cathode was made of Ketjenblack (KB, EC-600JD) with 10% polytetrafluoroethylene (PTFE) binder. The paste was roll-pressed on Ni mesh and vacuum-dried at 200 °C for 1 hour. The loading level was 2.0–2.5 mg cm−2 and the electrode density was 0.2 g cm−3. A Na metal sheet as an anode was prepared by flattening a Na metal piece (Sigma-Aldrich) in an Ar-filled glove box. A glass microfiber filter of 190 μm thickness (GC50, Advantec) was used as a separator. A 2032 coin cell consisting the electrodes, separator and NaAlCl4-2SO2 electrolyte was assembled in an Ar-filled glove box for discharge/charge tests. Beaker-type or swagelock-type cells were used for some occasions.

Electrochemical test

The assembled cells were aged for 12 hours at room temperature and then electrochemically tested using a TOSCAT battery measurement system under the following protocols. The first and second cycles were operated galvanostatically at 0.1C (=150 mA g−1 or 0.34 mA cm−2) within the voltage window of 2.0–4.05 V. In the following cycles the current was set to be at 0.5C and 0.2C for discharge and charge, respectively. For rate capability test, the discharge rate was varied from 0.2C to 5C with a fixed charge rate at 0.2C. For an overcharging test, a Na–SO2 cell was charged up to 4.3 V and/or limited time. To investigate the impedance behavior of the carbon cathode, a 3-electrode electrochemical cell was constructed, where a Na metal reference electrode was positioned closely to the carbon cathode. Electrochemical impedance spectroscopic measurements conducted in the frequency range of 100 kHz to 10 mHz, with an amplitude of 5 mV at every end of charge and discharge step during cycling (VSP-300, BioLogic).

Characterization

XRD (both in situ and ex situ) patterns were obtained using an Empyrean diffractometer (PANalytical) equipped with monochromated Cu Kα radiation (λ = 1.54056 Å). A lab-made swagelok-type in situ XRD cell was composed of KB-PTFE(10%) cathode, Na metal sheet anode and glass fiber separator, with a beryllium (Be) disk on the cathode side for a X-ray window as well as a current collector. For ex situ analyses, a gas-tight sample holder filled with Ar and covered with a polyimide (Kapton) tape was used. After cell reacted up to certain level, the cathode was carefully disassembled from the cell and then rinsed with SOCl2 in an Ar-filled glove box to remove residual electrolyte since the SOCl2 is known to dissolve SO2 and NaAlCl4 well28. The morphology change of electrode after cycling was analyzed by SEM (JSM-7000F, JEOL). Weight gain at the cathode as a result of products formation was measured by weighing the carefully washed and dried cathode before and after discharge in an Ar-filled glove box. For the ex situ SERS measurement, we employed the gold(Au)-nanoparticles-anchored carbon black (Au@Vulcan XC-72) as the cathode, which was received from Nara Cell Tech Corp., Korea. The size and content of the Au nanoparticles were about 10–30 nm and 60 wt%, respectively and the generation of SERS effect from the Au@C nanocomposite was confirmed before the measurement. The carefully washed and dried cathode was placed within a sealed sample holder where a quartz window was applied to the top cap. All procedures were undertaken in an Ar-filled glove box with O2 and H2O levels maintained at <1 ppm. SERS spectra were collected using a micro-Raman spectrometer (Bruker Senterra Grating 400) with a He–Ne laser at a wavelength of 532 nm. The power of the laser beam was less than 5 mW and the spectrum acquisition time was 10 s with 10 accumulations to avoid degradation to the standards or electrodes. The TEM image showing the size and population of Au-nanoparticles in the Au@C nanocomposite and the discharge voltage profile of the Au@C cathode for the SERS measurement in Fig. 3e, are presented in Supplementary Figure S8.

Computational Details

We conducted ab initio molecular dynamics (AIMD) simulations using the Vienna Ab initio Simulation Package29 (VASP) with the projector augmented-wave30 (PAW) approach for electrochemical reaction calculations. For the total energy calculation, the Perdew−Burke−Ernzerhof (PBE) generalized-gradient approximation (GGA) functional31 was used. On electrochemical reaction calculation, the electrons were added first into the initial structures composed of 2NaAlCl4 and 4SO2 and then Na atoms were added. We used Parrinello−Rahman dynamics for NPT ensemble and Minimal Γ-centered 1 × 1 × 1 k-point grid. Two or three snap shots, which had the minimum energy, were selected from AIMD simulations. Static first-principles calculations were calculated with the B3LYP functional and 6-31G(d) basis sets. The reaction energy and Raman frequency calculations were performed using the Gaussian09 program package.

Additional Information

How to cite this article: Jeong, G. et al. A room-temperature sodium rechargeable battery using an SO2-based nonflammable inorganic liquid catholyte. Sci. Rep. 5, 12827; doi: 10.1038/srep12827 (2015).