EQCM Study on Electrochemical Zinc Deposition-Dissolution in Water-In-Salt Electrolyte

Abstract

The reversible deposition-dissolution reactions of zinc in several “water-in-salt” (WIS)-based electrolytes were monitored and compared using an electrochemical quartz microbalance (EQCM). These WIS electrolytes, irrespective of composition, promote reversible zinc deposition with high Coulombic efficiency. The mass change on the electrode, the local viscosity, and the change in the water content differ among electrolyte compositions, possibly because of phase separation to a solid zinc salt-water precipitate induced by the local concentration change of electrolyte during zinc dissolution.

1. Introduction

Zinc is a versatile element for electrochemistry because of its material abundance, high solubility in both acidic and strongly basic solutions, and high reversibility of the redox reaction between zerovalent and divalent zinc species in various media because of high hydrogen overpotential on zinc metal. By providing these benefits, zinc metal has achieved wide use for negative electrodes in practical primary cells. Furthermore, efforts have been undertaken to use zinc metal in advanced secondary battery systems. As a negative electrode for secondary batteries with high energy density, zinc presents several additional benefits such as high theoretical volumetric capacity (5854 mAh L⁻¹) and low electrode potential (−0.76 V vs. NHE).^1–3

The redox reactions between zerovalent and divalent zinc, exhibited as zinc deposition and dissolution, show different electrochemical details such as formal potential, reversibility, shape of deposits, side reactions, and surface passivation by the by-products. Historically mild acidic and alkaline electrolytes have been used for such investigations. However, acidic electrolyte delivers self-discharge of the zinc metal anode and severe hydrogen reduction by-reaction.^2,4–6 Alkaline electrolyte promotes the formation of inactive zinc oxide and thereby engenders dendritic zinc deposition during charge–discharge cycles.⁷ Such undesirable side reactions are mainly responsible for solvent water. Recently hyper-concentrated aqueous solutions of metallic salt with bulky sulfonyl imide anions, which contain the least amount of water, designated as “water-in-salt”⁸ or “hydrate melt”⁹ solutions, have been proposed as alternative aqueous electrolyte candidates with attractive properties such as an extraordinary large electrochemical window up to 2.8 V. Such concentrated solutions have promoted interesting performance to aqueous lithium or sodium-ion insertion batteries,^8–13 lithium–air,¹⁴ lithium–sulfur¹⁵ batteries, and supercapacitors.^16–20 Furthermore, such “water-in-salt” (WIS) electrolytes containing zinc sulfonyl imide salt²¹ or hyper-concentrated aqueous electrolytes with a similar concept^22–24 have been used for zinc electrodes. In fact, they have shown unprecedented high reversibility of zinc deposition-dissolution with no dendrite growth.

Such WIS electrolytes are attractive not only from a practical viewpoint but also for elucidating the influence of water on the zinc deposition-dissolution process because only the stoichiometric amount of water is included in these media. Fruitful information is expected to be obtained by careful monitoring of reversible zinc deposition-dissolution in the WIS electrolyte. For this study, WIS electrolytes of two kinds have been used: one consisting of zinc sulfonyl imide and water and the other consisting of lithium sulfonyl imide and less zinc salt. Indeed, lithium salt provides the WIS electrolyte with a smaller amount of water. Consequently, the total water content of the zinc-based WIS electrolyte can be reduced by the coexisting lithium salt. In addition, an electrolyte to which a small amount of water is added to the WIS solution has been prepared. The electrochemical zinc deposition-dissolution in these electrolytes has been monitored using cyclic voltammetry method together with an electrochemical quartz microbalance (EQCM) technique. Change in the water content in the electrolyte during measurement has also been monitored to clarify the stoichiometric effects of water on the entire zinc deposition process.

2. Experimental

The WIS electrolytes were prepared by adding home-made distilled water in zinc bis(trifluoromethane sulfonyl)imide (Zn(TFSI)₂, battery grade; Tokyo Chemical Industry Co., Ltd., Japan) or lithium bis(trifluoromethane sulfonyl)imide (LiTFSI, battery grade; Tokyo Chemical Industry Co., Ltd.) until the salt was dissolved completely with heating. The lithium–zinc mixed WIS electrolyte was prepared by mixing the lithium WIS and the zinc WIS with the gravimetric ratio of 1 : 2, equivalent to the Zn : Li molar ratio of 16 : 84. The diluted zinc electrolytes were prepared by the addition of distilled water in the zinc WIS electrolyte. The water content in a WIS electrolyte was ascertained using a Karl Fischer titrator (MKC-710; Kyoto Electronics Manufacturing Co., Ltd., Japan). The titration was conducted more than three times for each sample. The average value was used for analyses. The electrochemical measurements were conducted using a commercial sealed cell designed for EQCM measurements (EC Frontier Co., Ltd., Japan). An Au-sputtered AT-cut quartz crystal (Base frequency, 8.9 MHz; Seiko EG&G Co., Ltd., Japan) was used as a working electrode. A zinc plate (0.1 mm thickness; The Nilaco Corp., Japan) was used as a counter-electrode after polishing, washing, and drying. A silver wire immersed in a saturated solution of silver nitrate (ca. 0.1 mol dm⁻³) in lithium WIS electrolyte was used as a reference electrode. The reference electrode was separated from the measurement electrolyte by a porous glass. The reference electrode stability was checked using rest potential monitoring between this electrode and a commercial Ag/AgCl reference electrode for 20,000 s. The resulting rest potential profile is shown in Fig. 1. From this result, the Ag/Ag⁺ electrode in the lithium WIS electrolyte is assumed to be stable at 0.6 V vs. Ag/AgCl (reasonable difference from the standard electrode potential between Ag/Ag⁺ and Ag/AgCl) within 20 mV potential deviation. The amount of electrolyte in the cell is ca. 0.2 mL. The assembled cell was equipped with the QCM oscillator/frequency counter (QCM922A; Seiko EG&G Co., Ltd.) and a potentiostat (HSV-110; Hokuto Denko Corp., Japan). The cyclic voltammetry (CV) measurements were conducted during 10 cycles, from −1.6 V to −0.4 V vs. Ag/Ag⁺ reference electrode at the scan rate of 2 mV s⁻¹. The cell temperature was controlled at 308 K in a temperature chamber. The frequency and oscillation resistance of the quartz working electrode were monitored. The mass change of the working electrode was calculated from the frequency change using the Sauerbrey equation. After CV measurements, the water content in the electrolyte was ascertained using a similar procedure.

Figure 1.

Time-course rest potential profile of a home-made reference electrode Ag in 0.1 mol dm⁻³ AgNO₃ in lithium WIS from the Ag|AgCl|HCl reference electrode. Cell electrolyte, lithium WIS solution; temperature, 308 K.

3. Results and Discussion

The water contents of the electrolytes used for this study are presented in Table 1. The equivalent molar ratio of water for one cation is calculated and included in the same table. The water molar ratio in the lithium WIS was calculated as 2.40, approximately equal to the value reported for LiTFSI-H₂O WIS (2.5).⁸ Divalent zinc ion therefore coordinates more water (around 9) to form a liquid. It is noteworthy that the zinc WIS is easily frozen at around ambient temperature, in particular with a change of pressure (e.g. in a pipette-aspiration). Consequently, the solution is inferred to be in a supercooled state. The water ratio for the zinc–lithium mixed WIS is obtained from calculation using the average formula weight with the consideration of the mixing ratio (Zn : Li = 16 : 84). Indeed, the rigorous attribution of coordinated number of water molecule for each cation is impossible to ascertain solely from the present information.

Table 1. Water contents in the electrolyte solutions used for this study.

Electrolyte	State	Gravimetric water content	Water per metal atom
Zn(TFSI)2-H2O (zinc WIS)	Pristine	0.260	9.05
	After EQCM	0.346
LiTFSI-H2O (lithium WIS)	Pristine	0.151	2.40
	After EQCM	0.147
Zn(TFSI)2-LiTFSI-H2O (zinc–lithium mixed WIS)	Pristine	0.214	4.07
	After EQCM	0.180
Zn(TFSI)2-H2O (zinc WIS+water)	Pristine	0.454	15.8
	After EQCM	0.421
Zn(TFSI)2-H2O (diluted Zn(TFSI)2)	Pristine	0.811	28.2
	After EQCM	0.775

The CV curves (a, d, g), mass change profiles (b, e, h), and oscillation resistances (c, f, i) for the simultaneous measurements for the zinc WIS (a–c), the zinc–lithium mixed WIS (d–f), the zinc WIS with small amount of water (Zn WIS+water) (c–i) are portrayed in Fig. 2. A significant redox couple centered around −1.4 V is shown for all of these CV curves. The reduction current delivers the increase of electrode mass. The subsequent oxidation current delivers the decrease of the mass to nearly zero. The mass change value per electron (mpe) was calculated from the CV current and mass change data every 0.5 V (1000 s) at the first cycle for each electrolyte. The mpe value is negative when the electrode mass increases on the cathodic scan, because the cathodic charge is indicated in negative sign within this paper. On the anodic scan, it is negative when the electrode mass decreases. It is shown by potential in Fig. 3. Open and filled symbols respectively represent the mass change during the oxidation and subsequent reduction processes. For the potential region in which redox current flows, the mpe values for these electrolytes are close to −32.7, the theoretical mpe for the deposition and dissolution of zinc. From these mpe values, the reduction current from −1.2 V vs. Ag/Ag⁺ and subsequent oxidation current represents the reversible deposition and dissolution process of zinc, respectively. The redox current for these electrolytes appears to be reversible without overpotential. Such a reversible feature is similar to earlier reports describing reversible zinc deposition in super-concentrated zinc salt aqueous electrolytes of several kinds.^21,22 The current value is ranged within −10 mA cm⁻² to 10 mA cm⁻² irrespective of the concentration of zinc, suggesting that the current is rather limited by diffusion near the electrode surface. The potential at zero-current and the onset slope of redox current for the zinc WIS+water electrolyte differ from WIS (both zinc WIS and zinc–lithium WIS), suggesting that water addition to the WIS electrolyte might lower the liquid viscosity and change the zinc ion activity.

Figure 2.

CV curves for zinc deposition-dissolution, corresponding mass change profiles and oscillation resistance profiles in various WIS-based electrolytes. CV scan rate, 2 mV s⁻¹; potential range, −1.6 to −0.4 V vs. Ag/Ag⁺; temperature, 308 K. Selected cycles: first, second, third and tenth cycle (drawn by lighter gray for more developed cycle). (a) CV curve in the zinc WIS, (b) mass change in the zinc WIS, (c) oscillation resistance change in the zinc WIS, (d) CV curve in the zinc–lithium mixed WIS, (e) mass change in the zinc–lithium mixed WIS, (f) oscillation resistance change in the zinc–lithium mixed WIS, (g) CV curve in the zinc WIS+water, (h) mass change in the zinc WIS+water, (i) oscillation resistance change in the zinc WIS+water.

Figure 3.

Mass per electron plots versus potential at the first cycle for the CV-EQCM measurements in various electrolytes. Electrolytes: ◇ zinc WIS (reduction), zinc WIS (oxidation), □ zinc–lithium mixed WIS (reduction), zinc–lithium mixed WIS (oxidation), △ zinc WIS+water (reduction), zinc WIS+water (oxidation).

The reversibility of zinc deposition and dissolution is also an indicator of the status because only metallic zinc can be dissolved on the anodic process and an efficiency value based on current displays directly the ratio of zinc deposition reaction to the total electron transfer in one cycle. Coulombic efficiency of zinc dissolution to zinc deposition has been calculated by dividing the sum of positive charge during cathodic current flows by the total negative charge during anodic current flows on a cyclic voltammogram. The Coulombic efficiency in these three electrolytes are shown by the cycle numbers in Fig. 4. Coulombic efficiencies in these electrolytes are approximately 0.9, indicating the reversibility of the redox processes in current. Both the zinc WIS and the zinc–lithium mixed WIS reach similar efficiency values at 0.93. The efficiency for the zinc WIS+water electrolyte is slightly lower than those found for the WIS electrolytes.

Figure 4.

Coulombic efficiency from CV curve in various electrolytes. Electrolytes: (open symbol) zinc WIS, (filled symbol) zinc–lithium mixed WIS, (gray symbol) zinc WIS+water.

Nevertheless, the redox reaction is not fully ‘reversible’ from the mass change during the cycle. The precise shape of the oxidation current peak changes after the second cycle to the first for the zinc–WIS electrolyte. The enlarged CV curves at this region are presented in Figs. 5a–5c. The high potential end of the oxidation current peak is enlarged; alternatively, an additional oxidation current peak or shouldered current appears at the high potential end after the second cycle. Some mpe values within the potential region where the current peak or the shoulder current appears (at the third cycle in the zinc WIS and the tenth cycle in the all electrolytes) are presented in Fig. 5d. All of these mpe values are negative, indicating weight loss during the oxidation process. Although the mpe value is not uniform, the region with mpe as large as −20 is inferred as representing the dissolution of zinc, together with the formation of zinc hydroxide by the partial chemical reaction of zinc and water, to a certain degree.   

\begin{equation} \text{Zn} + \text{2H$_{2}$O} \to \text{Zn(OH)$_{2}$} + \text{H$_{2}$} \end{equation} (1)

Nevertheless, the reason for the smaller mpe remains unclear.

Figure 5.

(a–c) enlarged CV curves in various electrolytes. Electrolytes: (a) zinc WIS, (b) zinc–lithium mixed WIS, (c) zinc WIS+water, (d) mpe at oxidation process in developed cycles versus potential. Electrolyte: ◇ zinc WIS (3rd cycle), zinc WIS (10th cycle), zinc–lithium mixed WIS, zinc WIS+water (10th cycle).

During all cycles, a large residual mass is observable. In particular for the zinc–lithium mixed WIS, a large residual mass is observed after the redox process for about half of the total mass change after the first charge–discharge cycle. One candidate for such a large amount of the residual mass might be zinc hydroxide formed by the chemical reaction of zinc and water. Furthermore, the cathodic reduction of water might occur as a side reaction.^5,25   

\begin{equation} \text{2H$_{2}$O} + \text{2e$^{-}$} \to \text{H$_{2}$} + \text{2OH$^{-}$ or 2H-ads} + \text{2OH$^{-}$} \end{equation} (2)

Zinc hydroxide Zn(OH)₂ can be formed from zinc ions in the electrolyte and the formed hydroxide ion. These formation processes of zinc hydroxide must occur at a certain extent. The resulting surface hydroxide layer might shift the oxidation potential in a higher direction, as in the case of zinc deposition in conventional neutral and mild acid electrolytes. However, the overall mass changes are impossible to explain based on this process alone. The assumption that Zn(OH)₂ is main product in the case of the zinc-lithium mixed WIS electrolyte is contradicts the larger amount of residual mass in the zinc-lithium mixed WIS which contains smaller amount of water compared with the case of such electrolytes as the zinc WIS and the zinc WIS+water. Another possible contribution for the residual mass increase for the zinc-lithium WIS electrolyte is a decomposition product of TFSI⁻ anion which has been reported to be formed in the lithium WIS electrolyte at the potential region under −0.4 V vs. Ag/AgCl and act as a protective layer to inhibit hydrogen evolution reaction described as Eq. 2.^8,16 However, no sign indicating the formation of a solid deposit by the cathodic decomposition of TFSI⁻ (formula weight is as high as 280). From these reasons, the main portion of the residual mass on the electrode in the WIS electrolytes must be a product by other processes.

The oscillation resistance (Figs. 2c, 2f, and 2i) provides information about the local viscosity beneath the electrode surface. Resistance during a CV cycle in the zinc WIS and the zinc–lithium mixed WIS electrolytes is influenced strongly by the state of zinc deposition and dissolution. The resistance decreases slightly with the first zinc deposition process. It significantly increases at the potential region above the current peak of zinc dissolution. Then it gradually decreases at a certain extent. The increase of the oscillation resistance may correspond to the increase of the local viscosity beneath the zinc surface, probably as the result of the rapid increase of zinc ion in the solution phase and the subsequent the lack of a coordinating water molecule for the WIS electrolyte. After the cycle develops, besides the baseline increase because of the precipitation of by-product, a significant resistance decrease can be observed soon after the resistance increase at the zinc dissolution process. The decrease of the resistance, suggesting the decrease of the local viscosity, during zinc deposition is expected to be related to the decrease of the zinc ion concentration at the interface.

In the zinc WIS, the increase of the oscillation resistance together with the zinc dissolution is to be highlighted. Such a rapid increase of zinc ion results in the precipitation of solid Zn(TFSI)₂·xH₂O from the super-cooled WIS state. Two reasons exist for this inference. When the test cell was disassembled after the CV-EQCM measurement, some salt-like precipitation on the Au-quartz electrode, together with the remaining liquid electrolyte, was visually observed, as shown in Fig. 6. The photograph shows the disassembled cell body after the Au-quartz electrode was removed. Needle-like crystals filled in the hole and the O-ring contacted with the Au-quartz electrode is clearly observable. Such a dense salt deposition must be the result of the phase separation of bulk electrolyte, even if some electrochemical process is triggered the phase separation. The water content of the liquid electrolytes in the disassembled cells is also included in Table 1. It is noteworthy that the water content in the zinc WIS increases after the measurement. Because the water content is similar before and after the similar 10 cycle voltammetry measurement for the lithium WIS electrolyte under the same conditions, the influence of water leakage from the environment can be ignored. Although it is difficult to explain such a complicated behavior, one reasonable explanation is the phase separation between a liquid phase and a solid phase having different water content. Assume that a solid salt containing a smaller amount of water, which can be denoted as Zn(TFSI)₂·xH₂O (x < 9) was crystalized from the WIS electrolyte. The residual liquid phase can include a larger amount of water than the electrolyte, which can be denoted as Zn(TFSI)₂·yH₂O (y > 9), before the measurement. From the resistance profile, one can infer that such phase separation might occur together with the zinc dissolution. Some ‘stabilization’ between these separated phases, containing the immersion of the water-rich liquid Zn(TFSI)₂·yH₂O and exfoliation of a part of the bulky solid Zn(TFSI)₂·xH₂O, might decrease the resistance. For the zinc–lithium mixed WIS and the zinc WIS+water electrolytes, the water content decreases considerably after the CV-EQCM measurement. The moles of the decreased water are several hundred times the entire flow of electrons during the measurement. Therefore, the decrease of water is not attributable to the electrolysis of water. Furthermore, in these cases, the phase separation occurs to form the solid phase Zn(TFSI)₂·xH₂O, even though no solid-like deposit is visually observed probably because of the lower content of zinc ion compared with the zinc WIS. For the zinc–lithium mixed WIS, the residual mass after cycles might be the deposited Zn(TFSI)₂·xH₂O. If the water content of the solid phase Zn(TFSI)₂·xH₂O is less than the zinc WIS but greater than the zinc–lithium mixed WIS, 4 < x < 9, then the water content in the electrolyte can decrease after the CV measurement. Based on this assumption, the marked increase of the resistance for the zinc–lithium mixed WIS is explainable by the decrease of water in the electrolyte. If the formation of Zn(TFSI)₂·xH₂O phase over the bulk electrolyte in the case of the zinc WIS would not provide the frequency change in EQCM but the relatively small amount of Zn(TFSI)₂·xH₂O in the zinc-lithium mixed WIS would deposit on the electrode surface and change the frequency in EQCM, the complicated behavior of residual mass can be explained.

Figure 6.

Photograph of deposited salt at the contact space with the Au-quartz electrode.

The CV curve (a) and corresponding mass change profile (b) for the aqueous solution of Zn(TFSI)₂ (diluted zinc WIS) are depicted in Fig. 7. The approximate molar concentration of this solution is 0.4 mol dm⁻³. The voltammogram shows a rather irreversible feature, for which only the reduction current is observable at the potential range of −0.6–−1.6 V. The electrode mass correspondingly increases. Also, the mpe value at the first cycle is calculated as −54.5, which is approximately equal to the theoretical value for Zn(OH)₂ −49.7. It can be inferred that the water reduction together with the Zn(OH)₂ formation occurs preferentially, and that the deposited Zn(OH)₂ or zinc oxide ZnO passivates the electrode to prevent the reversible deposition of zinc. The difference between the WIS electrolytes and this case might be the amount and activity of free water. Excess free water might promote surface passivation in such a neutral electrolyte. Although the irreversible behavior in this neutral electrolyte differs from results reported for reversible zinc deposition in neutral ZnCl₂ aqueous solution,²⁶ the reversible zinc deposition can be achieved even in these media by the scan to lower potential. It is noteworthy that the present measurement conditions are optimized for measurement in the WIS-based electrolyte system. However, by comparison using the present condition, the intrinsic differences between the WIS-based electrolyte and conventional aqueous electrolyte are clearly visible.

Figure 7.

CV curves and corresponding mass change profiles in diluted Zn(TFSI)₂ electrolyte. CV scan rate, 2 mV s⁻¹; Potential range, −1.6 to −0.4 V vs. Ag/Ag⁺; Temperature, 308 K. Selected cycles: first, second, third and tenth cycle (shown as lighter gray for more developed cycles).

Despite several important pieces of evidence, the phase separation mechanism during the reversible deposition-dissolution of zinc in the WIS-based electrolyte remains a subject of speculation. Further systematic study of the phase behavior of Zn(TFSI)₂-H₂O concentrated solution is necessary to elucidate the phenomena discussed above. It is noteworthy that the reversible redox couple observed in the WIS-based electrolyte is fundamentally a stable toward cycle. Therefore, if phase separation were to occur on the WIS-based electrolyte surface, then the Zn(TFSI)₂·xH₂O solid phase on the electrode would not passivate the reversible reaction of zinc. Phase separation, or at least some solid precipitation phenomenon, is regarded as a key factor affecting electrode processes in WIS-based electrolytes, and is worth being further examination and discussion.

4. Summary

Being different from dilute neutral aqueous electrolyte, the water-in-salt (WIS) electrolyte containing Zn(TFSI)₂, irrespective of the addition of a small amount of water, promotes the reversible deposition-dissolution of zinc metal without overpotential. The Coulombic efficiency for the zinc deposition in the WIS-based electrolyte becomes as high as 0.93. The corresponding mass change behavior differs according to the composition of the electrolyte. The change in the water content during the CV cycles and the change in the oscillation resistance of quartz electrode suggest phase separation to a solid precipitation and liquid together with the dissolution of zinc. One can assume that the Zn(TFSI)₂·xH₂O solid phase precipitated on the electrode surface is a potential component of a surface layer which stabilize the electrode surface with promotion of the reversible electrode process.

CRediT Authorship Contribution Statement

Minato Egashira: Conceptualization (Lead), Formal analysis (Lead), Investigation (Lead), Visualization (Lead), Writing – original draft (Lead)

Conflict of Interest

The authors declare no conflict of interest in the manuscript.

Footnotes

M. Egashira: ECSJ Active Member

References

• 1)  Y. Li, J. Fu, C. Zhong, T. Wu, Z. Chen, W. Hu, K. Amine, and J. Lu, Adv. Energy Mater., 9, 1802605 (2019).
• 2)  J. Shin, J. Lee, Y. Park, and W. Choi, Chem. Sci., 11, 2028 (2020).
• 3)  Z. Cao, P. Zhuang, X. Zhang, M. Ye, J. Shen, and P. M. Ajayan, Adv. Energy Matter., 10, 2001599 (2020).
• 4)  T. Yamamoto and T. Shoji, Inorg. Chim. Acta, 117, L27 (1986).
• 5)  X. G. Zhang, Corrosion and Electrochemistry of Zinc, Springer (1996).
• 6)  K. Wippermann, J. W. Schulze, R. Kessel, and J. Penninger, Corros. Sci., 32, 205 (1991).
• 7)  Z. Liu, S. Z. El Abadin, and F. Endres, Electrochem. Commun., 58, 46 (2015).
• 8)  L. Suo, O. Borodin, T. Gao, M. Olguin, J. Ho, X. Fan, C. Luo, C. Wang, and K. Xu, Science, 350, 938 (2015).
• 9)  Y. Yamada, K. Usui, K. Sodeyama, S. Ko, Y. Tateyama, and A. Yamada, Nat. Energy, 1, 16129 (2016).
• 10)  F. Wang, L. Suo, Y. Liang, C. Yang, F. Han, T. Gao, W. Sun, and C. Wang, Adv. Energy Mater., 7, 1600922 (2017).
• 11)  Z. Yu, L. A. Curtiss, R. E. Winans, Y. Zhang, T. Li, and L. Cheng, J. Phys. Chem. Lett., 11, 1276 (2020).
• 12)  X. He, B. Yan, X. Zhang, Z. Liu, D. Bresser, J. Wang, R. Wang, X. Cao, Y. Su, H. Jia, C. P. Grey, H. Frielinghaus, D. G. Truhlar, M. Winter, J. Li, and E. Paillard, Nat. Commun., 9, 5320 (2018).
• 13)  L. Suo, O. Borodin, Y. Wang, X. Rong, W. Sun, X. Fan, S. Xu, M. A. Schroeder, A. V. Cresce, F. Wang, C. Yang, Y.-S. Hu, K. Xu, and C. Wang, Adv. Energy Mater., 7, 1701189 (2017).
• 14)  Q. Dong, X. Yao, Y. Zhao, M. Qi, X. Zhang, H. Sun, Y. He, and D. Wang, Chem, 4, 1345 (2018).
• 15)  C. Yang, L. Suo, O. Borodin, F. Wang, W. Sun, T. Gao, X. Fan, S. Hou, Z. Ma, K. Amine, K. Xu, and C. Wang, Proc. Natl. Acad. Sci. U.S.A., 114, 6197 (2017).
• 16)  G. Hasegawa, K. Kanamori, T. Kiyomura, H. Kurata, T. Abe, and K. Nakanishi, Chem. Mater., 28, 3944 (2016).
• 17)  W. S. V. Lee, T. Xiong, G. C. Loh, T. L. Tan, and J. Xue, ACS Appl. Energy Mater., 1, 3070 (2018).
• 18)  Z. Li, G. Jeanmairet, T. Méndez-Morales, B. Rotenberg, and M. Salanne, J. Phys. Chem. C, 122, 23917 (2018).
• 19)  S. Thareja and A. Kumar, ACS Sustain. Chem. Eng., 9, 2338 (2021).
• 20)  J. Park, J. Lee, and W. Kim, ACS Energy Lett., 6, 769 (2021).
• 21)  F. Wang, O. Borodin, T. Gao, X. Fan, W. Sun, F. Han, A. Faraone, J. A. Dura, K. Xu, and C. Wang, Nat. Mater., 17, 543 (2018).
• 22)  N. Zhang, F. Cheng, Y. Liu, Q. Zhao, K. Lei, C. Chen, X. Liu, and J. Chen, J. Am. Chem. Soc., 138, 12894 (2016).
• 23)  L. Li, S. Liu, W. Liu, D. Ba, W. Liu, Q. Gui, Y. Chen, Z. Hu, Y. Li, and J. Liu, Nano-Micro Lett., 13, 34 (2021).
• 24)  J. Han, A. Mariani, A. Varzi, and S. Passerini, J. Power Sources, 485, 229329 (2021).
• 25)  K.-D. Song, K.-B. Kim, S.-H. Han, and H. Lee, Electrochem. Solid-State Lett., 7, C20 (2004).
• 26)  S. Cora, S. Ahmad, and N. Sa, ACS Appl. Mater. Interfaces, 13, 10131 (2021).