Size effect of SO₂ receptors on the energy efficiency of Na–SO₂ batteries: gallium-based inorganic electrolytes

Abstract

Na–SO₂ batteries have garnered significant attention as alternative energy conversion/storage systems, however, their low energy efficiency, which corresponds to large polarization, is a crucial restriction in achieving high performance of the cell. Here, a new Ga-based inorganic electrolyte, NaGaCl₄·2SO₂, which effectively improves the energy efficiency by decreasing the coulombic interaction between the ionic species of NaGaCl₄ and SO₂ is proposed and its underlying working mechanism is demonstrated.

---

Introduction

Recently, alternative energy storage/conversion systems employing metallic anode materials have garnered significant attention as the concerns associated with the finite global lithium reserves increase.^1–3 Among these metallic anode materials, an attractive candidate is sodium (Na) because it can provide high theoretical energy density based on its high specific capacity with low cost.^4,5 Therefore, many types of Na-based battery systems have been investigated in the energy field, such as Na–S,^6,7 Na–NiCl₂,^8,9 Na–O₂,^10,11 and Na-ion^12–15 batteries.

Recently, our research group proposed a new Na-based battery system, the Na–SO₂ battery, which works on a reversible electrochemical reaction of a Na-based inorganic electrolyte.^16,17 The Na-based electrolyte is composed of electrochemically active sulfur dioxide (SO₂) and a counter part of SO₂ receptor, which is an inorganic complex of sodium chloride (NaCl) and aluminum chloride (AlCl₃). During the discharge state, the SO₂ ligand that is weakly coordinated on NaAlCl₄ is first reduced via an electrochemical reaction and a substantial chemical reaction of the reduced SO₂ ligand (SO₂⁻) with NaAlCl₄ leads to the formation of discharged products, NaCl and NaAlCl₂(SO₂)₂. During the charge state, the SO₂ attached to NaAlCl₂ is electrochemically oxidized and thereafter the reproduced SO₂ ligand reforms the initially developed coordinated inorganic complex, NaAlCl₄·2SO₂.¹⁷ The Na–SO₂ battery has a remarkable specific capacity over 1000 mA h g⁻¹ as well as an excellent stable cycling retention, through the reversible reaction:

2Na + NaAlCl₄·2SO₂ ↔ 2NaCl + NaAlCl₂(SO₂)₂.

Despite these advantages, the large voltage polarization of the Na–SO₂ batteries is a crucial restriction to expanding its application. This restriction results from the strong binding affinity between the reduced SO₂⁻ and the SO₂ receptors, NaAlCl₂. The reduced nucleophilic SO₂⁻ species readily combines with the electrophilic Al-based SO₂ receptor via the formation of strong chemical bonding in the discharge process. That is, the cleavage of the chemical bond that corresponds to Al–SO₂ in the charge process is too difficult. Therefore, it requires significantly higher electrochemical energy to progress its reverse reaction. This results in the charge potential substantially increasing in the cell, which is responsible for the large polarization of the Na–SO₂ batteries. Note that the energy density of the cell is proportional to the capacity and voltage of the cell. Therefore, the large overpotential in Na–SO₂ batteries must be alleviated in order to achieve high energy density.

Thus, in this paper, a gallium (Ga)-based inorganic electrolyte is proposed that can effectively reduce the large overpotential behavior in Na–SO₂ cells. Note that the binding affinity between the ionic species is primarily governed by the coulombic force, which is proportional to the charge of the ionic species and inversely proportional to the square of the distance between the ionic species. This means that using larger SO₂ acceptors than NaAlCl₂ would be effective in decreasing the coulombic interaction as the distance between the SO₂ receptor and reduced SO₂⁻ increases, which would facilitate a reverse electrochemical reaction, which determines the polarization behavior of the Na–SO₂ cell. Therefore, a Ga-based SO₂ receptor that can compensate the loss of energy efficiency for Na–SO₂ batteries was employed. Because the large size of the Ga element decreases the coulombic interaction between the Ga-based SO₂ receptor and the reduced SO₂⁻ compared with the Al-based inorganic electrolyte, it is expected that the SO₂ ligand is easily detached from the SO₂ receptor, which would allow a reversible electrochemical reaction with less polarization. In practice, the cell composed of the Ga-based SO₂ receptor exhibited a considerable decrease in the voltage polarization compared with the Al-based electrolyte. In addition, the underlying reaction mechanisms are clarified for the Ga-based Na–SO₂ batteries through a combination of experimental and computational analyses, which are effective in understanding the electrochemical aspects of advanced conversion-based batteries.

Experimental and computational methods

The electrolyte (NaMCl₄·2SO₂, M: Al or Ga) was prepared in a homemade glass/Teflon vessel. In detail, sodium chloride (NaCl, >99.9%, Alfa Aesar) and metal chloride precursor (MCl₃, 99.999%, Alfa Aesar) was placed in vessel with a molar ratio of 1.1 to 1.0 and thereafter SO₂ gas was blown into reactor for two hours. The mixture was then converted to transparent liquids. After preparation of electrolytes was completed, we calculated the mass of SO₂ that participated in the formation of the electrolyte complex by weighing the vessel reactor in order to determine number of substituted SO₂ participated in the formation of the electrolyte. The chemical structure of electrolyte was measured by Fourier transform infrared spectrometer (FT-IR, Bruker, VERTEX 70) and ionic conductivity was observed by conductivity meter (TOA) at room temperature.

For evaluating electrochemical performances, mixture of 90% carbon (Ketjen black, EC-600JD) with 10% poly(tetrafluoroethylene) was finely agitated and pastes were subjected to coating onto Ni mesh. The loading level of the electrode was 3.0 mg cm⁻² and the electrode density was 0.2 g cm⁻³. The 2032 coin-type cell was assembled with Na metal as the anode, carbon-coated Ni mesh as the cathode, electrolyte, and a glass filter (GC50, 190 μm, Advantec) as a separator. The cells were then rested for 12 h and thereafter they were subjected to galvanostatically discharging/charging using an electrochemical tester (Toyo, TOSCAT3100) in the range of 2.00 to 4.05 V (vs. Na/Na⁺) with a 0.1C current density (150 mA g⁻¹) for pre-cycling (two cycles) and 0.2C current for 50 cycles. Energy efficiency of the cell was calculated at 1 cycle by equation shown in below.

Energy efficiency (%) = (average discharge voltage × specific discharge capacity) × 100/(average charge voltage × specific charge capacity)

After evaluating electrochemical performances, the discharged cells were disassembled in an Ar-filled glove box and the electrodes were recovered. For characterization of surface morphologies of discharged electrodes, the discharge electrodes were subjected to analyses by scanning electron microscopy (SEM, JSM-7000F, JEOL) and X-ray diffractometry (PANalytical) to characterize the discharge products developed on the electrode surface. For analysis of chemical composition corresponds to discharge products, recovered discharged electrodes were measured by FT-IR spectrometer and X-ray photoelectron spectroscopy (XPS, Thermo-Scientific, K alpha) under an N₂ atmosphere in a dry room in which the dew point was less than −60 °C.

To characterize the discharge product of the Ga-based electrolyte, ten 2032 coin-type cells were assembled and were discharged with a 0.1C current density (150 mA g⁻¹). The end condition was controlled by the specific capacity depending on the depth of discharge (DOD): five cells were finished if the DOD reached 30% and the other five cells stopped when the DOD reached 60%. The partially discharged cells were disassembled in an Ar-filled glove box and the electrodes were recovered. The mass changes of the recovered discharged electrodes were carefully measured by a scale, and the average values of the mass changes were used for the determination of the discharge products.

Density functional theory (DFT) calculations were performed using the Gaussian 09 software package.¹⁸ The geometry optimization calculations employed Becke three-parameter hybrid exchange functional combined with Lee–Yang–Parr correlation functional (B3LYP)^19,20 and the standard 6-31G* basis set. All stationary points were characterized by vibrational frequency analysis with Hessian matrices.

Results and discussion

Fig. 1 depicts the basic physicochemical properties for the Ga-based electrolyte. The NaCl and GaCl₃ are solid at room temperature, however, the mixture of solids became a transparent liquid when they were reacted with SO₂ gas (Fig. 1a). This indicates that the chemical environment of NaCl and GaCl₃ changed and resulted in forming a new inorganic complex. The FT-IR analysis of electrolyte supported this explanation and demonstrated that the electrolyte exhibited new absorbance peaks at 1323 and 1155 cm⁻¹, which can be assigned to the vibrational peaks of the S–O chemical connectivity in the new inorganic complex (Fig. 1b).²¹

Fig. 1 (a) NaGaCl₄-based electrolyte after exposing to SO₂ gas, (b) FT-IR analysis for NaGaCl₄-based electrolyte, (c) ionic conductivities for Al and Ga-based electrolytes at room temperature, and (d) degree of SO₂ ligand substituted on metallic SO₂ receptor (Al: black, and Ga: blue).

The prepared Ga-based inorganic electrolyte exhibited a moderate ionic conductivity of 67.2 mS cm⁻¹, which was a similar value to that of NaAlCl₄·2SO₂ (Fig. 1c). In addition, its chemical composition was determined through evaluating the calculation of the SO₂ mass that participated in the formation of the new inorganic electrolyte (more details are explained in the Experimental section) (Fig. 1d). It was observed that the 2.35 mol of SO₂ was attached to the NaGaCl₄·SO₂ receptor, which indicated that the primary composition of the electrolyte was NaGaCl₄·2SO₂, where a minor portion of NaGaCl₄·3SO₂ might also be present.

In order to evaluate the size effect of the SO₂ receptor, the electrolytes were subjected to electrochemical performance evaluations as presented in Fig. 2. The initial potential profiles differed depending on the SO₂ receptor: the average discharge potential for NaGaCl₄·2SO₂ decreased slightly to 2.65 V (vs. Na/Na⁺) compared with NaAlCl₄·2SO₂ (2.87 V, vs. Na/Na⁺). However, the potential difference in the charge/discharge was significantly improved in NaGaCl₄·2SO₂, where the ΔV for NaGaCl₄·2SO₂ was 0.498 V, while that for NaAlCl₄·2SO₂ was 1.097 V, which corresponded to a large overpotential (Fig. 2b). Note that it requires a high potential to oxidize the SO₂ ligand attached to the NaAlCl₄ (3.97 V, vs. Na/Na⁺, energy efficiency of 71.5%). In contrast, it is relatively easy to promote the oxidation reaction of the SO₂ ligand chelated with NaGaCl₄ (3.14 V, vs. Na/Na⁺, energy efficiency of 82.1%). This implies that using Ga is effective in promoting the reverse electrochemical reaction of SO₂ through reducing the coulombic interaction between the ionic species. In addition, the NaGaCl₄·2SO₂ exhibited excellent cycling performance (Fig. 2c). The NaGaCl₄·2SO₂ exhibited 99.9% cycling retention based on its remarkable average coulombic efficiency (99.9%) for 50 cycles, while the cycling performance of NaAlCl₄·2SO₂ faded at several initial cycles (at a retention ratio of 88.2% with an average coulombic efficiency of 99.3% for 50 cycles). This indicated that employing the Ga-based SO₂ receptor was effective in not only improving the energy efficiency corresponding to the polarization behaviors, but also in ensuring the cycling performance of the Na–SO₂ batteries.

Fig. 2 (a) Initial potential profiles for electrolytes, (b) degree of voltage polarization (ΔV) at 50% state of charge, and (c) cycling performances for electrolytes (Al: black, and Ga: blue).

In order to verify the detailed reaction mechanism for the electrochemical reaction of NaGaCl₄·2SO₂, the discharged electrode was characterized using scanning electron microscopy (SEM) and X-ray powder diffractometry (XRD) in order to determine the discharge products (Fig. 3). The solid product appeared in the discharged electrode in the SEM image, and its chemical compositions verified that the Na and Cl elements were observed in the energy-dispersive spectroscopy (EDS) analyses (Fig. 3a). The XRD analyses for the discharged electrode exhibited the chemical structure of the discharge product. That is, clear spectroscopic evidence corresponding to the formation of crystalline NaCl was observed in the discharged electrode cycled with NaGaCl₄·2SO₂ (Fig. 3b). These indicate that once the SO₂ ligand was first reduced in the initial discharge state, the electrochemically produced SO₂⁻ was further chemically reacted with NaGaCl₄, leading to the formation of NaCl as a discharge product similar to the Na–SO₂ batteries with the NaAlCl₄·2SO₂ electrolyte.^16,17

Fig. 3 (a) SEM analysis result for discharged electrode cycled with Ga-based electrolyte (up) and its EDS analysis result (down), and (b) XRD analyses for pristine electrode (black), NaCl (red), and discharged electrode cycled with Ga-based electrolyte (blue).

In order to clarify the additional discharge product that is a counterpart of NaCl, the discharged electrode was further analyzed using FT-IR and XPS spectrometry (Fig. 4). Note that the XRD analyses only allow prediction for crystalline compounds and thus, analyzing the discharged electrode using FT-IR and XPS can provide information to estimate the discharge product with an amorphous structure. The electrode exhibited new absorbance peaks associated with the existence of chemical bonding of S–O and S=O (1384, 1329, 1263, 1219, 1157, 1045, 1022, and 897 cm⁻¹) (Fig. 4a).²¹ The XPS analyses also indicated four new peaks in the discharged electrode: an Na peak was observed at 1073.3 eV together with NaCl (1072.3 eV),²² a Cl peak appeared at 201.4 eV as well as NaCl (199.7 eV),²³ an S peak was found at 162.3 eV, and an O peak occurred at 536.8 eV. These additional spectroscopic peaks implied that additional discharge product also existed in the electrode as a counterpart of NaCl.

Fig. 4 Analysis for discharged electrode cycled with Ga-based electrolyte by (a) FT-IR, and (b) XPS.

For more details regarding the characterization of the discharge product, the mass changes (Δm) of the discharged electrode were also monitored depending on the depth of discharge (DOD). Note that the mass changes of the discharged electrode were highly associated with the degree of SO₂ substitution on the Ga-based SO₂ receptor, and this provides useful information to determine the discharge product in Na–SO₂ batteries. Several hypotheses were considered depending on the numbers of SO₂ ligands substituted in the Ga-based SO₂ receptor as well as the formation of Na₂S₂O₄. The theoretical mass-to-charge (m/Q) relationships are displayed as lines and the experimental results are plotted by scatterplots in Fig. 5a. The Δm values for the electrode discharged with NaGaCl₄·2SO₂ with a variation of DOD were measured. Then, they were subjected to matching with theoretical Δm values, which were estimated using a simple calculation depending on the number of SO₂ ligands substituted. The Q over Δm plot indicated that the chemical structure of the discharge product appeared to be NaGaCl₂(SO₂)₂ in which two moles of SO₂⁻ were replaced with two moles of Cl⁻ attached to Ga. The experimental m/Q value from our measurements was approximately 120 mA h g⁻¹, which is close to the value for the formation of NaGaCl₂(SO₂)₂ and 2NaCl. The formation of NaGaCl₂(SO₂)₂ and 2NaCl would consume two moles of SO₂ ligands with two electrons per mole of NaGaCl₄, resulting in a cathode mass of 131.2 mA h g⁻¹, as indicated by the slope of the blue line. This is well supported by the DFT calculations for several types of discharged products for SO₂ ligands. The lowest reaction energy was observed in the formation of NaGaCl₂(SO₂)₂ (ΔG = −6.91 kcal mol⁻¹), while the other discharge products exhibited higher reaction energies (Fig. 5b). In addition, the reaction energy of NaGaCl₂(SO₂)₂ was higher than that of NaAlCl₂(SO₂)₂ (ΔG = −9.73 kcal mol⁻¹). This indicates that NaGaCl₄ had a lower coulombic interaction than the NaAlCl₄ with increases in the size of the SO₂ receptor, which contributed to lowering the severe polarization behavior in Na–SO₂ batteries. According to these results, a working reaction mechanism for Ga-based Na–SO₂ batteries can be established, as below.

2Na + NaGaCl₄·2SO₂ ↔ 2NaCl + NaGaCl₂(SO₂)₂

Fig. 5 (a) Relationship of capacity versus mass of discharged electrode depending on depth of discharge at initial cycle, and (b) theoretical thermodynamic energies for discharge products depending on number of substituted SO₂ ligands (kcal mol⁻¹).

Substitution of most SO₂ ligands on Ga-based SO₂ receptors would enhance the safety performance of Na–SO₂ batteries because there are few free SO₂ ligands in the cell. Preliminary studies support the explanation that the Li–SO₂ batteries which use LiAlCl₄·6SO₂ as an electrolyte are subjected to high vapor pressure in the cell because only three moles of SO₂ ligands participate in the discharge reaction while the remaining SO₂ ligands (3SO₂) continuously increase the internal pressure associated with toxic gas generation.^24,25 On the other hand, most of the SO₂ ligands on the Ga-based electrolyte are electrochemically reduced during discharge process, resulting in the formation of discharge products via a further chemical reaction. It can therefore be anticipated that the use of the Ga-based SO₂ receptor is effective for not only improving the electrochemical performance capabilities but also for enhancing the safety characteristics of Na–SO₂ batteries.

Conclusions

The effect of Ga-based SO₂ receptors on the electrochemical performance of Na–SO₂ batteries was demonstrated and its working mechanism was clarified. The chemical structure of the Ga-based inorganic electrolyte was NaGaCl₄·2SO₂, and it exhibited excellent ionic conductivity of 67.2 mS cm⁻¹. Compared with NaAlCl₄·2SO₂, NaGaCl₄·2SO₂ significantly enhanced the energy efficiency associated with the large overpotential (from 1.097 to 0.498 V) as well as the cycling retention (from 88.2% to 99.9% at 50 cycles). NaCl and NaGaCl₂(SO₂)₂ were formed during the discharge process, and the reverse reaction allows reversible electrochemical reactions for Na–SO₂ batteries, which lead to efficient electrochemical performances. It is expected that these insights correspond to a size effect in the SO₂ receptors and the overall reaction mechanism would be beneficial in making Na–SO₂ batteries more attractive as an alternative post lithium-ion batteries system.

Acknowledgements

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (2016R1C1B1009452 and 2016R1A2B4013374). In addition, this work was partially supported by the Energy Efficiency & Resources Core Technology Program of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) granted financial resource from the Ministry of Trade, Industry & Energy, Republic of Korea (No. 20132020000260)

References

1. H. Wu and Y. Cui, Nano Today, 2012, 7, 414.
2. L. Ji, Z. Lin, M. Alcoutlabi and X. Zhang, Energy Environ. Sci., 2011, 4, 2682.
3. W.-J. Zhang, J. Power Sources, 2011, 196, 13.
4. Y. Kim, K.-H. Ha, S. M. Oh and K. T. Lee, Chem.–Eur. J., 2014, 38, 11980.
5. Y. Kim, Y. Park, A. Choi, N.-S. Choi, J. Kim, J. Lee, J. H. Ryu, S. M. Oh and K. T. Lee, Adv. Mater., 2013, 25, 3045.
6. J. Wang, J. Yang, Y. Nuli and R. Holze, Electrochem. Commun., 2007, 9, 31.
7. X. Lu, B. W. Kirby, W. Xu, G. Li, J. Y. Kim, J. P. Lemmon, V. L. Sprenkle and Z. Yang, Energy Environ. Sci., 2013, 6, 299.
8. C.-W. Ahn, M. Kim, B.-D. Hahn, I. Hong, W. Kim, G. Moon, H. Lee, K. Jung, Y.-C. Park and J.-H. Choi, J. Power Sources, 2016, 329, 50.
9. G. Li, X. Lu, J. Y. Kim, V. V. Viswanathan, K. D. Meinhardt, M. H. Engelhard and V. L. Sprenkle, Adv. Energy Mater., 2015, 5, 1500357.
10. W.-J. Kwak, Z. Chen, C. S. Yoon, J.-K. Lee, K. Amine and Y.-K. Sun, Nano Energy, 2015, 12, 123.
11. S. H. Sahgong, S. T. Senthilkumar, K. Kim, S. M. Hwang and Y. Kim, Electrochem. Commun., 2015, 61, 53.
12. J. Y. Jang, Y. Lee, Y. Kim, J. Lee, S.-M. Lee, K. T. Lee and N.-S. Choi, J. Mater. Chem. A, 2015, 3, 8332.
13. C. Jo, Y. Park, J. Jeong, K. T. Lee and J. Lee, ACS Appl. Mater. Interfaces, 2015, 7, 11748.
14. W. Luo, C. Bommier, Z. Jian, X. Li, R. Carter, S. Vail, Y. Lu, J.-J. Lee and X. Ji, ACS Appl. Mater. Interfaces, 2015, 7, 2626.
15. A. Ponrouch and M. R. Palacín, Electrochem. Commun., 2015, 54, 51.
16. G. Jeong, H. Kim, H. S. Lee, Y.-K. Han, J. H. Park, J. H. Jeon, J. Song, K. Lee, T. Yim, K. J. Kim, H. Lee, Y.-J. Kim and H.-J. Sohn, Sci. Rep., 2015, 5, 12827.
17. Y.-K. Han, G. Jeong, K.-J. Lee, T. Yim and Y.-J. Kim, Phys. Chem. Chem. Phys., 2016, 18, 24841.
18. M. J. Frisch, et al., Gaussian 09 (Revision D.01), Gaussian Inc., Wallingford, CT, 2009.
19. A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
20. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785.
21. E. Pretsch, P. Buhlmann and C. Affolter, Structure determination of organic compounds, Springer, Germany, 2000.
22. J. S. Hammond, J. W. Holubka, J. E. Devries and R. A. Duckie, Corros. Sci., 1981, 21, 239.
23. A. Shchukareva and J.-F. Boily, Surf. Interface Anal., 2008, 30, 349.
24. A. N. Dey, H. C. Kuo, P. Piliero and M. Kallianidis, J. Electrochem. Soc., 1988, 135, 2115.
25. D. L. Foster, H. C. Kuo, C. R. Schlaikjer and A. N. Dey, J. Electrochem. Soc., 1988, 135, 2682.

---