Investigation for Charge-Discharge Operations of Li₄Ti₅O₁₂-Sulfur Batteries by Suitable Choice of Materials and Cell Preparation Processes

Abstract

To achieve stable charge-discharge operations of Li-S batteries, [Li₄Ti₅O₁₂ | sulfurized polyacrylonitrile (SPAN)] cells treated with a Li pre-doping process were prepared by two different techniques. Several different materials were studied without encountering any difficult barriers, and the obtained cells exhibited suitable charge-discharge operations with high efficiency.

1. Introduction

Recently, Li-ion batteries (LIBs) have become widely used for customer use as high-voltage, high-energy density storage devices.^1,2 Next-generation battery systems are expected to use innovative materials such as high-capacity and high-voltage (electrode potential) electrode materials. Li-sulfur (Li-S) batteries^3,4 possess a high energy density owing to their high positive electrode capacity of 1,672 mAh g⁻¹ compared with conventional electrode materials such as Li_xCoO₂ (137 mAh g⁻¹, 0.5 < x < 1). Li-S batteries consist of the crown-shaped molecule S₈ and Li metal as positive and negative electrode active materials, respectively, and are produced at low cost because S₈ is a by-product of petroleum refining. However, there are many problems for practical applications of the systems, such as capacity degradation from dissolution of reactive intermediates (Li₂S_x) from the active material S₈ into electrolyte solutions^5–7 and dendrite formation on the negative electrode.^8,9 To suppress the dissolution of Li₂S_x, low-solubility electrolytes have been proposed, such as solvate ionic liquids¹⁰ and highly-concentrated sulfolane-based liquid electrolytes with low Lewis acidity or basicity;¹¹ with these electrolytes, long life cycles of over 800 cycles were achieved with high charge-discharge retentions of the Li-S cell.¹² A stable S-based positive electrode of sulfurized cyclic polyacrylonitrile (SPAN)^13,14 with high conductivity and chemical stability has been prepared by high-temperature treatment of elemental S and PAN. A SPAN positive electrode enables the application of conventional carbonate-based liquid electrolytes used for general LIBs with high stability because of the chemical bonds between the S_x molecules and cyclic PAN during the charge-discharge process without dissolution of Li₂S_x,¹⁵ even though there are a few reports concerning the incorporation of monatomic S into SPAN. Conventional carbon-based C₆ (graphite)¹⁶ and the strain-free intercalation electrode Li₄Ti₅O₁₂ (LTO)^17,18 have been investigated for the suppression of Li dendrite and suitable electrode interfacial formation as favorable negative electrode materials. In this study, we explored [LTO | SPAN] cells containing stable positive and negative electrodes during charging operations, even though there are no mobile Li⁺ sources in the electrodes. Therefore, two Li pre-doping methods were applied by (i) electrochemical insertion of Li into the LTO electrode and (ii) direct installation of an appropriate Li metal foil onto the LTO negative or the SPAN positive electrodes.¹⁹ In this study, we investigated the effects of techniques and physical conditions for ‘pre-lithiation methods’ using essential stable electrochemical cells such as [LTO | SPAN] cell system. The relationships between cell preparation methods and conditions and electrochemical charge-discharge performances were evaluated for high-performance [LTO | SPAN] cells.

2. Experimental

The Li₄Ti₅O₁₂ negative electrode sheet was prepared by mixing Li₄Ti₅O₁₂ (LTO: Ishihara Sangyo), acetylene black (AB: Denka Black Li-100, Denka), and poly(vinylidene difluoride) (PVDF: Kureha) (mass ratio of LTO : AB : PVDF = 85 : 7 : 8). The SPAN positive electrode sheet was prepared by mixing SPAN (theoretical capacity: 545 mAh g⁻¹, S : PAN = 33 : 67, ADEKA), porous carbon (Ketjenblack, KB: ECP-600JD, Lion), carboxy-methyl cellulose (CMC: Nacalai Tesque) and styrene-butadiene rubber (SBR: JSR) (mass ratio of SPAN : KB : CMC : SBR = 90 : 5 : 2.5 : 2.5). The electrode materials were dissolved in N-methylpyrrolidone (LTO negative electrode) or ultra-pure H₂O (SPAN positive electrode) and mixed in a paste mixer to form a homogeneous slurry. The obtained slurries were coated onto Cu (LTO negative electrode) or Al (SPAN positive electrode) current collectors and were dried in an oven at 353 K (LTO) or 333 K (SPAN) for 12 h. The electrode sheets were compressed to increase the packing density and improve the electrical conductivity (approximate loading masses of LTO and SPAN: 8.0–8.5 mg cm⁻² and 0.6–1.2 mg cm⁻², respectively).

For the first cell preparation scheme, in the first step of the assembly, the LTO electrode sheet, a glass separator (thickness: 0.21 mm, GA-55: Advantec), ethylene carbonate/diethyl carbonate = 3/7 by mass, a 1.0 mol L⁻¹ LiN(SO₂F)₂ electrolyte solution (ES, Mitsubishi chemical), and Li metal (Honjo Metal) were placed in 2032-type coin cells (Miclab) that could be disassembled. All cell-treatment processes were carried out in an Ar-filled glovebox ([O₂] < 10 ppm, dewpoint < 193 K: Miwa Mfg Co., Ltd.). In the second step, electrochemical Li pre-doping was performed in galvanostatic mode using a charge-discharge measurement system (Hokuto Denko) at 303 K and 1.0–3.0 V with a current density of C/12 (charge and discharge process, ca. 15 mA g⁻¹) based on the LTO negative electrode, repeating the activation three times. The C-rate is a measure of the rate (current) at which a prepared cell is completely charged or discharged, relative to theoretical capacity of based active material. After Li pre-doping, the Li-doped LTO (Li-LTO) electrodes were removed from the cells and rinsed in dehydrated dimethyl carbonate (Wako chemical). In the third step, [Li-LTO negative electrode | ES | SPAN positive electrode] cells were prepared using the Li-LTO electrodes and encapsulated into a 2032-type cell. The prepared cells were evaluated by galvanostatic charge-discharge at 303 K and −0.5–1.5 V at a current density of C/4 (ca. 136 mA g⁻¹) based on the SPAN positive electrode.

For the first step of the second cell preparation scheme (Fig. 1), the SPAN positive electrode, a glass separator, ES, an LTO negative electrode, and a capacity-regulated Li metal foil (Li on LTO side: Scheme 1; SPAN side: Scheme 2) were encapsulated into 2032 coin-type cells (electrode capacity: LTO > Li metal > SPAN). In Scheme 2, the prepared cells were treated pre-cycle by a galvanostatic charge process at 303 K and 1.0 V at a current density of C/24 (ca. 23 mA g⁻¹) based on the SPAN positive electrode materials. Finally, the prepared cells were evaluated by galvanostatic charge-discharge at 303 K and −0.5–2.3 V at a current density of C/24 (ca. 23 mA g⁻¹) based on the SPAN positive electrode.

Figure 1.

Scheme for preparation and operation of [L₄T₅O₁₂ negative electrode | electrolyte solution | SPAN positive electrode] cell by direct Li installation methods (electrochemical Li intercalation processes of Li₄Ti₅O₁₂ to SPAN (a), SPAN to Li₄Ti₅O₁₂ (c), and stabilization process by continuous charge-discharge operation (b, d)).

3. Results and Discussion

Figure 2a shows the charge-discharge profiles of the prepared [Li-LTO negative electrode | ES | SPAN positive electrode] cell at 303 K. A relatively larger overpotential was observed during the first discharge process, in which Li⁺ moves from LTO to SPAN, compared with the charge process owing to the differences in reactivity between pre-lithiated LTO and untreated SPAN. Reactive cell voltages were shifted with continuous cycle operations (discharge processes: −0.10 to 0.16 V; charge processes: 0.56 to 0.63 V) and converged at stable charge-discharge voltages with high reversibility. Reactive voltages approached the calculated ideal reaction voltage (calculated by stable charge-discharge voltages of each electrodes) differences between the SPAN and the LTO electrodes (charge processes: 2.15 V vs. Li/Li⁺ (SPAN) − 1.55 V vs. Li/Li⁺ (LTO) = 0.60 V; discharge processes: 1.57 V vs. Li/Li⁺ (LTO) − 1.31 V vs. Li/Li⁺ (SPAN) = 0.26 V). Lowering of the overpotential was also suggested by electrochemical uniformize (Li⁺ insertion) into SPAN positive electrode with cycling. Moreover, the prepared cell did not exhibit significant changes of its charge-discharge capacities and overpotential, which provides indirect evidence for the dissolution suppression of lithium polysulfide (Li₂S_x) during the charge-discharge operation using SPAN. Figure 2b also shows the dependence of the charge-discharge capacities on the number of cycles and the calculated Coulombic efficiencies of the prepared [Li-LTO negative electrode | ES | SPAN positive electrode] cell. A stable, relatively high capacity of over 330 mAh g⁻¹ and an efficiency of 99.9 % for SPAN were confirmed after an initial conditioning of approximately 20 cycles. High cycleability was observed for the cell, with ca. 97 % capacity retention between 140 and 500 cycles, even though slight degradation of capacity was observed. Electrochemical behavior for dendrite formation on the Li metal was not observed during charge processes from the side of the negative electrode, and cells with long life cycles were expected, with no weak points at either the positive or the negative electrodes.

Figure 2.

(a) Charge-discharge profiles and (b) dependences of charge/discharge capacities based on SPAN with cycle number and calculated Coulombic efficiencies for [L₄T₅O₁₂ negative electrode | electrolyte solution | SPAN positive electrode] by electrochemical insertion of Li into L₄T₅O₁₂ electrode at 303.15 K.

Figure 3a shows the charge-discharge profiles of the prepared [Li on LTO negative electrode | ES | SPAN positive electrode] cell at 303.15 K. A relatively high discharge capacity of ca. 500 mAh g⁻¹, close to the theoretical capacity of SPAN during the first cycle, and stable reversibility were observed. In this study, both SPAN and LTO were considered to have a small, irreversible capacity owing to their higher electrochemical properties, such as SEI (solid electrolyte interphase)-free interfacial formation, compared with a conventional C₆ electrode system.¹⁹ LTO had sufficient electrochemical stability with Li metal, and should serve as a good source of Li. Therefore, the [LTO | SPAN] system enabled an easy Li activation process, in which Li metal only settles onto an LTO electrode, rather than engaging in a complicated Li doping process (e.g., electrochemical lithiation). In contrast, Fig. 3b shows the charge-discharge profiles of the prepared [LTO negative electrode | ES | Li on SPAN positive electrode] cell at 303.15 K. The initial open circuit voltage and reactive voltage had negative values during the Li doping process onto LTO, unlike the data shown in Fig. 3a. After sufficient charge (Li insertion) currents for LTO, Li metal should have been adsorbed onto the SPAN positive electrode, and the voltage should have converted from negative to positive values. The prepared cell also exhibited relatively high capacities of ca. 500 mAh g⁻¹ and stable charge-discharge operations, similar to the voltage profiles shown in Fig. 3a. Detailed difference between proposed two technical methods was not clear, and installation of reference electrode such as Li metal will appear promising for understanding each electrode potential.

Figure 3.

(a) Charge-discharge profiles of [Li-on L₄T₅O₁₂ negative electrode | electrolyte solution | SPAN positive electrode] and (b) [L₄T₅O₁₂ negative electrode | electrolyte solution | Li-on SPAN positive electrode] cells by direct Li installation methods at 303.15 K. Obtained capacity was calculated based on SPAN weight.

The proposed direct installation methods enable experimental applications for both positive and negative electrodes, and are introduce reactive ionic species such as Li⁺ during charge-discharge reactions. Moreover, various positional selectivity of the highly reactive metal electrodes are possible without decomposition, regardless of the electrode materials. Contributions are also expected during cell production for easy and effective Li doping. Careful cell design, including the choice of materials and the process for cell preparation, promises easy production of stable battery systems; investigation into methods to improve the cyclability and storage life of the batteries is also strongly desired. We will also achieve stable battery systems by using various electrode and electrolyte materials and reactive metal electrodes (except for Li).

Acknowledgments

We acknowledge ADEKA corporation for supplying us with valuable chemical SPAN for our research. This work was partially supported by the supported by the Advanced Low Carbon Technology Research and Development Program (ALCA), SPRING Grant JPMJAL1301 from Japan Science and Technology Agency (JST).

CRediT Authorship Contribution Statement

Kazuki Machida: Data curation (Lead), Investigation (Lead), Writing – original draft (Lead)

Hibiki Miyauchi: Conceptualization (Supporting), Data curation (Supporting), Formal analysis (Supporting), Writing – original draft (Supporting)

Yusuke Ushioda: Conceptualization (Supporting), Data curation (Supporting), Formal analysis (Supporting), Investigation (Equal), Writing – review & editing (Supporting)

Keitaro Takahashi: Conceptualization (Supporting), Data curation (Supporting), Formal analysis (Supporting), Investigation (Supporting), Writing – original draft (Supporting), Writing – review & editing (Supporting)

Shiro Seki: Funding acquisition (Lead), Project administration (Lead), Supervision (Lead), Writing – original draft (Equal), Writing – review & editing (Lead)

Conflicts of Interest

There are no conflicts to declare.

Funding

Japan Science and Technology Agency: JPMJAL1301

Footnotes

S. Seki: ECSJ Active Member

References

• 1)  D. Larcher and J. M. Tarascon, Nat. Chem., 7, 19 (2015).
• 2)  N. Nitta, F. Wu, J. T. Lee, and G. Yushin, Mater. Today, 18, 252 (2015).
• 3)  B. Scrosati, J. Hassoum, and Y. K. Sun, Energy Environ. Sci., 4, 3287 (2011).
• 4)  P. G. Bruce, S. A. Freunberger, L. J. Hardwick, and J. M. Tarascon, Nat. Mater., 11, 19 (2012).
• 5)  R. D. Rauh, F. S. Shuker, J. M. Marston, and S. B. Brummer, J. Inorg. Nucl. Chem., 39, 1761 (1977).
• 6)  J. Shim, K. A. Striebel, and E. J. Carins, J. Electrochem. Soc., 149, A1321 (2002).
• 7)  Y. V. Mikhaylik and J. R. Akridge, J. Electrochem. Soc., 151, A1969 (2004).
• 8)  J. B. Goodenough, J. Solid State Electrochem., 16, 2019 (2012).
• 9)  M. Z. Mayers, J. W. Kaminski, and T. F. Miller, J. Phys. Chem. C, 116, 26214 (2012).
• 10)  K. Dokko, N. Tachikawa, K. Yamauchi, M. Tsuchiya, A. Yamazaki, E. Takashima, J. W. Park, K. Ueno, S. Seki, N. Serizawa, and M. Watanabe, J. Electrochem. Soc., 160, A1304 (2013).
• 11)  A. Nakanishi, K. Ueno, D. Watanabe, Y. Ugata, Y. Matsumae, J. Liu, M. L. Thomas, K. Dokko, and M. Watanabe, J. Phys. Chem. C, 123, 14229 (2019).
• 12)  S. Seki, N. Serizawa, K. Takei, Y. Umebayashi, S. Tsuzuki, and M. Watanabe, Electrochemistry, 85, 680 (2017).
• 13)  J. Wang, J. Yang, J. Xie, and N. Xu, Adv. Mater., 14, 963 (2002).
• 14)  S. Wei, L. Ma, K. E. Hendrickson, Z. Tu, and L. A. Archer, J. Am. Chem. Soc., 137, 12143 (2015).
• 15)  S. S. Zhang, Energies, 7, 4588 (2014).
• 16)  Z. Ogumi and M. Inaba, Bull. Chem. Soc. Jpn., 71, 521 (1998).
• 17)  K. M. Colbow, J. E. Dahn, and R. R. Haering, J. Power Sources, 26, 397 (1989).
• 18)  T. Ohzuku, A. Ueda, and N. Yamamoto, J. Electrochem. Soc., 142, 1431 (1995).
• 19)  Y. Ushioda, K. Takahashi, M. Watanabe, and S. Seki, Chem. Lett., 50, 1217 (2021).