The Liquid-Mediated Synthesis and Performance Evaluation of Li-Zr-F Composite for Ion-Conduction

Junmo Moon ¹, Satita Thiangtham ¹, Ruijie Zheng ¹, Sicheng Liu ¹, Chayanaphat Chokradjaroen ¹, Yasuyuki Sawada ^1,2,3, Nagahiro Saito ^1,2,3,4,*

1. Department of Chemical Systems Engineering, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464-8603, Japan

2. Japan Science and Technology Corporation, Open Innovation Platform with Enterprises, Research Institute and Academia, Furo-cho, Chikusa-ku, Nagoya, 464-8603, Japan

3. Japan Science and Technology Corporation (JST), Strategic International Collaborative Research Program (SICORP), Furo-cho, Chikusa-ku, Nagoya, 464-8603, Japan

4. Conjoint Research Laboratory in Nagoya University, Shinshu University, Furo-cho, Chikusa-ku, Nagoya, 464-8603, Japan

* Correspondence: Nagahiro Saito

Academic Editor: Osamu Yamamoto

Special Issue: Battery Materials for Energy Conversion

Received: December 04, 2022 | Accepted: February 12, 2023 | Published: February 20, 2023

Journal of Energy and Power Technology 2023, Volume 5, Issue 1, doi:10.21926/jept.2301010

Recommended citation: Moon J, Thiangtham S, Zheng R, Liu S, Chokradjaroen C, Sawada Y, Saito N. The Liquid-Mediated Synthesis and Performance Evaluation of Li-Zr-F Composite for Ion-Conduction. Journal of Energy and Power Technology 2023; 5(1): 010; doi:10.21926/jept.2301010.

© 2023 by the authors. This is an open access article distributed under the conditions of the Creative Commons by Attribution License, which permits unrestricted use, distribution, and reproduction in any medium or format, provided the original work is correctly cited.

1. Introduction

Regarding environmental concerns and the goal of sustainable development, solid-state batteries (SSBs) are considered a new candidate energy storage technology to be the next generation of batteries. Solid electrolytes (SEs) are essential component that adapt from a lithium metal anode to a cathode [1]. Typically, the SEs are incorporated in the battery, mainly to resolve the electrolyte leakage, flammability, and limited energy density of liquid electrolytes [2]. Compared with conventional liquid batteries, known as lithium-ion batteries, SEs show superior performance, higher safety in electric vehicles (EVs) and operational durability [3,4]. Nonetheless, the fabrication of the architecture is still challenged by interfacial issues between the electrode and SEs, generally due to the instability of the SEs against the metallic anode and cathode, obstructing their use in practical applications. Owing to the grain boundaries, pores, and even single crystals of SE, which are generated, the irregular electric-field distribution and poor interfacial contact with the electrodes leading to the introduction of the dendritic deposition cause rapid short-circuiting of the battery [5,6,7,8,9]. These side reactions generally occur during repeated charge/discharge, which increases interfacial resistance and greatly inhibits battery performance. Various strategies have been proposed to overcome these limitations of SBBs by developing new types of SE, which can be classified into solid polymer electrolytes (SPEs), inorganic ceramic electrolytes (ICEs), and solid composite electrolytes (SCEs) in order to enhance the interfacial contact stability between SE and electrode [10,11]. Exceptionally, an attractive feature of ICEs coming onto the stage such as LiPON, oxide-, sulfide-, hydride-, and halide-based SEs, exhibit good mechanical properties, nonflammability, high-temperature stability, and greater ionic conductivity [10]. Due to their high ionic conductivity of over 10^-3 S cm^-1 at ambient temperature, the oxide- and sulfide-based SEs are provided to be the candidates for utilization in SSBs [10,11,12]. Notwithstanding, the rigid oxide-based SE and thermodynamically unstable with high-voltage oxide cathode materials of sulfide-based SEs might have limited practical use, especially the manufacturing complexity on a large scale and sensitivity to the air [11,13].

In comparison, halide-based SEs such as F-, Cl-, and Br-, etc., [14] anion compounds with higher electronegative than oxide and sulfide, exhibit the wide thermodynamic range of electrochemical stability windows with electrode materials, which have demonstrated potential application in SSBs. [11,15] Among halide-based SE materials, crystalline lithium fluoride (LiF) is altering as the new candidate with a wide electrochemical stability window (0-6.4 V vs Li/Li⁺) and good chemical stability with electrodes which could aid high energy density of batteries [16,17]. Unfortunately, due to the strong bonding between Li and F ions in the LiF structure, thus, making it has high stability leading to hampering the diffusivity of the, suffering from the low contribution of ionic conductivity 10^-13 S/cm at RT [18]. Hence, it would be desirable to design the lattice structure to prevent coupling bonding of Li-F and to trigger a fast diffusion of $Li^+$ without compromising its intrinsically high electrochemical stability [16,18]. It was widely acknowledged that the relationship between the ionic conductivity of electrolytes and their crystal orientations was related to the number of vacancies, interstitials, and partial occupancy on interstices or lattice sites [19]. Additionally, a strong correlation between the orientated $Li^+$ sublattice and the energy barrier of $Li^+$ migration might also be affected to enhance the ion conduction of LiF-based SEs with low activation energy ($E_a$) [11,20]. Thus, based on the concept of cation sublattice and ion arrangement, it could be implied that the structure of Li-M-F,$M$ as the metal-transition cation in high oxidations states ($M^{3+}$: Co, Ni;$M^{4+}$: Mn, V, Cr, Ni,$M^{5+}$: Cr) [21], resulted in the possible structure design to increase its ionic conductivity by the substitution of heterovalent cation to dilate the spatial lattice of crystal layer, known tuning-defect chemistry [19,22]. Through the vacancies mechanism, the Li⁺ hopping process was sensitive to the active $Li^+$ and vacancy concentration, which might result in order magnitude changes in conductivity. Additionally, based on their electrostatic forces, the stacking structure of LiF by layered cation ionic crystals with distinctive radii could be arranged in crystal orientation [19]. Microstructural effects increased the ion conduction values of LiF through the sub-layered of metal by space-charge effects between Li-M-F causing the enhancement of the degree of amorphization [23]. For example, in 2017, J. Xie et al. [16]. Successfully synthesized LiAlF₄-SE thin film via atomic layer deposition (ALD). The LiAlF₄-SE thin film exhibited higher performance with robust stability and satisfactory ion conductivity than LiF and AlF₃. The result demonstrated that the wide range of the predicted stable electrochemical window of approximately 2.0-5.7 V vs. Li⁺/Li for LiAlF₄ led to excellent interfacial stability with both electrodes, causing a higher ion conductivity was evaluated at 3.5 × 10^-8 S/cm. Moreover, M. Feinauer et al [24]. Reported that a high-energy ball-milling method could be utilized to synthesize β-Li₃AlF₆. After mechanical milling of LiF and AlF₃ for 20 h at 600 rpm, the crystal structure of β-Li₃AlF₆ was formed. An ionic conductivity of 3.9 × 10^-6 S/cm was observed at 100°C. Even though the ion conduction of Li-M-F formation might be improved, most Li-M-F SEs were synthesized with energy-intensive and time-consuming processes, including high-force and energy ball-milling and high-temperature annealing process led to energy waste [10,12,25].

To address these issues, the liquid-mediated synthesis routes would be suggested as a promising method to synthesize Li-M-F compounds with much more energy-friendly and time-saving [10,26]. Recently, Li et al., [12] studied a method to directly synthesize high-performance Li₃InCl₆-based SE in water, with ionic conductivity as high as 2.04 × 10^-3 S/cm, under a simple synthesis method. The reactants were dissolved in water, then dried in the air to obtain the Li₃InCl₆ product. Then, in 2020 Hu et al., [27] reported a dissolution-precipitation method based on the ionic liquid to synthesize Li₃GaF₆ fluoride-based SE by combining Li₂CO₃ and Ga(NO₃)₃·xH₂O, as raw materials, in C₁₀mimBF₄ ionic liquid. The ion conduction of Li₃GaF₆ fluoride-based SE had been improved up to 10^-4 S/cm at RT. However, there are only a few publications to study the effect of aqueous solutions such as water or ethanal media on the synthesis of halide electrolytes, a lack of knowledge that hampers the development of SEs. Herein, the novel synthesis with methanol has been studied for Li-Zr-F-based SE via facile synthesis, energy-friendly conditions and short processing consuming time and post-treatment. The methanol-mediated precursor was mainly composed of methanol-coordinated halides and organic fragments to synthesize LiZrF-based SE. The organics in the precursor would vanish in the form of gas release including H₂, CH₄, H₂O, -OH, methylene oxide (H-CHO), and some unknown ionized fragments generated post-treatment. Moreover, considering the creation of vacancy clusters in the SE, a tetravalent valence cation was proposed to develop the ion conduction of LiF material. Zirconium (Zr), a metallic element of group 4, has the highest valence state of tetravalence and three inherent vacancies which were proposed to introduce the stacking layer of stabilized metal halide structure to form Li-Zr-F (LiZrF) ionic crystals. The different radius atoms were affected by the arrangement of crystal lattice based on their electrostatic forces. Based on the radius atom of $Zr^{4+}$ (0.72 Å), it exhibited the nearest neighbor within $Li^+$ (0.76 Å) sublattice which could be implied as a good structural descriptor in certain intrinsic $Li^+$ mobility attributed to vacancy size matching [18,21]. Through substituting $Zr^{4+}$ cation and under optimal synthesized conditions, Li₂ZrF₆ formation, built by the edge-sharing ZrF₈ polyhedrons, demonstrated the stabilized structure and rigid framework as isolated lagers in the bc-plane. Between the layers of ZrF₈ polyhedrons were occupied by the sandwiched edge-sharing LiF₆ octahedrons. Thus, making each of $Li^+$ in layered Li₂ZrF₆ might diffuse along a 2D channel in bc-plane, from one octahedral site to another. The electrochemical impedance spectroscopy (EIS) test demonstrated that Li₂ZrF₆-based SEs exhibited superiors ion conduction of 2.40 × 10^-8 S/cm and 3.89 × 10^-8 S/cm than LiF-based SE (1 × 10^-8 S/cm), which were synthesized under the stoichiometric proportion of LiF and ZrF₄ 2:0.8 at a dried temperature of 50°C/150°C with, respectively. Exceptionally, via the stoichiometric proportion of LiF and ZrF₄ 2:0.8, named Li-Rich LiZrF-based SE, revealed the electronic conductivity of 3.42 × 10^-9 S/cm, which was lower than the ionic conduction, approximately 10% of the overall conductivity with the activation energy ($Ea$) of 0.21 eV. This significant relationship corresponded to the theoretical crystal formation and the preferred orientation that might directly affect the polyhedrons, through which migration might be fast with low activation energies in crystalline conductors.

2. Materials and Methods

2.1 Materials

Lithium fluoride (LiF), high purity grade ≥99.98% trace metals basis (≤100 μm) and Zirconium (IV) fluoride (ZrF₄), high purity grade 99.9% and LiNi_0.8Co_0.1Mn_0.1O₂ (NCM811) trace metals basis and carbon black were supplied from Sigma Aldrich. Methanol (CH₃OH), Guaranteed reagent grade >99.8%(GC), was purchased from KANTO CHEMICAL CO., INC.

Polyvinylidene fluoride (PVDF) (average Mw ~534,000 by GPC, Powder) and 1-Methyl-2-pyrrolidone (NMP) (MW: 99.13) were provided by Sigma Aldrich. All chemical reagents were used without any further purification.

2.2 Liquid-Mediated Synthesis of Li-Zr-F Composite

LiZrF composite was successfully synthesized with methanol as solvent. First, highly pure starting materials LiF and ZrF₄ were weighed to the stoichiometric molar ratio of LiF:ZrF₄ 2:1 and dissolved in methanol media under an ambient environment, compared with water media. The suspensions were sonicated for 10 min at room temperature (RT) in air. The reaction continuously proceeded until the entire solution was completely dissolved to form a transparent solution. Subsequently, the homogeneous transparent solutions were dried at 25°C, 50°C, and 70°C in the air until all the solvent evaporated to obtain a dry precursor powder. Eventually, the precursor powders were subjected to the vacuum oven and heated at 150°C for 2 h. Finally, the LiZrF sample was subjected to the furnace for heat treatment at 200°C for 1 h to remove all impurities of organic solvent and humidity. After that the LiZrF composites were obtained. Figure 1 illustrates the various steps in preparing LiZrF composite via methanol-mediated precursor. Besides, to further evaluate the effect of vacancy concentration on the ionic conduction, LiZrF, named Li-Rich LiZrF, was synthesized via the stoichiometric molar ratio of LiF:ZrF₄ 2:0.8 in methanol media and 10 min sonication at RT which was dried at 50°C/150°C under vacuum. The Li-Rich LiZrF product also was subjected to a furnace oven for post-treatment at 200°C.

Figure 1 Schematic diagram of LiZrF composite preparation process via methanol-mediated synthesis.

2.3 Structure Characterizations

Fourier transform infrared (FTIR) spectroscopy of the precursor powder was conducted by Thermo Fisher Scientific iS50 instrument with attenuated total reflectance (ATR) iD7). The experiments were performed FT-IR in the range of 600-4000 cm^-1 with a resolution of 4 cm^-1 and a total of 64 scans for each sample. The crystalline structures of Li-Zr-F composites were identified by X-ray diffractometry (XRD) using a Rigaku SmartLab (Rigaku) machine equipped with Cu Kα irradiation (λ = 1.54056 Å) at an angular incidence of 10-90° (2θ angle range) with a scan speed of 2°/min and a scan step of 0.02°. The diffractogram peaks were examined using the JCPDS database. The reference intensity ratio (RIR) method, a semi-quantitative phase analysis via the X-ray powder diffraction internal standard method, was utilized to reveal the composition of the LiZrF composite, which was equivalent to the crystal phase formation obtained from XRD analysis. The influence of vacancy concentration and activation energy of ion migration was described by following the Arrhenius equation on ionic conductivity. The Boltzmann-Einstein equation was also used to describe the influence of other factors on ionic conductivity ($\sigma$) of Li-Rich LiZrF-based SE compared with LiZrF- and LiF-based SEs. Finally, the element mapping and micromorphology of LiZrF composites were conducted by field-emission scanning electron microscopy (FESEM) with energy dispersive X-ray spectroscopy (EDX) on a JSM 7610F machine (JEOL).

2.4 Ion Conduction Performance Evaluation

The ionic conduction of Li₂ZrF₆ ($P \overline{31} m$) composites-based SSE was examined by electrochemical impedance spectroscope (EIS) under a small amplitude alternating current (AC) potential wave with different frequencies applied to the electrochemical system. 200 mg of dried sample was cold-pressed into a polyether ether ketone (PEEK) mold with an inner diameter of 10 mm. The pellet was sandwiched between two cylindrical stainless-steel current collectors with 10 mm of diameter and pressed at a pressure of 20 MPa. The EIS measurement was carried out at the frequency range of 1 Hz-7 MHz using an electrochemical workstation in Biotech.

Later, EIS measurement was conducted on the ion-blocking symmetric batteries to calculate the temperature dependence of Li₂ZrF₆ ($P \overline{31} m$) composites-based SSE at a temperature from 308 to 368 K (frequency ranged of 1 MHz to 0.1 Hz). Furthermore, the electronic conductivity of all samples was measured by direct current polarization tests with externally applied voltages of 0.1, 0.2, 0.3, 0.4, and 0.5 V for 1 h to reach a steady voltage.

3. Results and Discussion

A liquid-mediated precursor synthesized Li-Zr-F (LiZrF) composite under methanol (CH₃OH) solution operated under low temperature and short consuming reaction time. As shown in Figure 1, LiF and ZrF₄, as the reactance, were dispersed in CH₃OH to form the homogeneous solution under sonification at RT. Under the method-mediated synthesis, $Li^+$, and $Zr^{4+}$ were combined with and $F^-$ to trigger the precipitation of LiZrF product in the solution. The chemical structure of LiZrF composite and possible chemical structures in the methanol-mediated precursor was detected by FT-IR spectroscopy, as shown in Figure 2. The different characteristic peaks at the corresponding wavenumber could verify the specific fragment structure of organic materials in the precursor. Figure 2a shows the spectrum of LiZrF composite before heat treatment that exhibited the typical band patterns of chemical functional groups in the CH₃OH structure. The characteristic broad absorption bands around 3360 and 2900 cm^-1 were assigned to the stretching of the -OH groups. The characteristic peak at 2921 cm^-1 was attributed to methyl groups’ symmetric C-H stretching vibrations of [28]. The strong peak at 1641 cm^-1 and the peak at 1460 cm^-1 corresponded to the impurity of the solvent and water vapor phase due to the presence of the deformation pattern of the OH group. The peaks frequency at 1377, 1117, and 1082 cm^-1 were attributed to the strong stretching pattern of the C-O-H fragment related to the deformed methyl group and the absorptions at 1020 and 968 cm^-1 corresponded to the C-H rocking vibration of methanol [10,29]. In contrast, these absorptions disappeared in the spectrum peak of LiZrF after heat treatment, as shown in Figure 2b. The decreased intensity in the LiZrF indicated the impurity and organic fragments as the CH₃OH and humidity were completely removed by thermal treatment. As a result, it was confirmed that in the presence of a methanol-mediated precursor synergistically with heat treatment, LiF would react with ZrF₄ to establish the crystal formation of LiZrF without impurity contamination.

Figure 2 FTIR spectra of LiZrF (a) before and (b) after heat treatment.

The analysis of the crystal phase formation and components of LiZrF products synthesized via different mediator conditions has been conducted utilizing XRD analysis. As the consideration under the XRD database with the inorganic crystal structure database (ICSD) for compounds containing three elements, the LiZrF product was mainly three kinds of compounds including Li₂ZrF₆ with space group of $P \overline{31} m$ (PDF#72-1291) or $P21/c$ (PDF#28-0611), and Li₄ZrF₈ with a space group of $Pnma$ (PDF#50-1317) [30]. Furthermore, to calculate the composition of LiZrF compared with LiF, the RIR equation was typically used to investigate the stoichiometric composition of material based on crystal phase from the XRD pattern [31,32], revealed in Figure 3. Table 1 reveals the phase composition of LiZrF composites, compared among different precursors including none-mediated, H₂O-mediated, and CH₃OH-mediated precursors. Considerably, under the none-mediated pathway, the XRD pattern showed the peak of the crystal phase of reactants. It could be suggested that the reaction could not be completed within 12 hours under none-mediated synthesis leading to occurring the crystal phase of LiF and ZrF₄, as can be seen in Figure 3a. On the other hand, in underwater and CH₃OH liquid-mediated synthesis, the phase compositions of LiZrF products were significantly changed from the none mediator condition. This result implied that the mediators might improve the productive synthesis of the reaction by enhancing the reaction rate and providing the opportunity and sensitivity for the colliding of $Li^+$,$Zr^{4+}$ and $F^-$ during the propagation of the reaction. Considering water-mediated synthesis (Figure 3b), the generated LiZrF had the existence of all three-formed structures. Based on the phase diagram of the LiF-ZrF₄ system [33], the presence of such a three-phase sample was unusual because the region does not exist in the phase diagram. This phenomenon might be due to the high solubility of LiF and ZrF₄ in water, which precipitated too fast in the evaporation process of the mediator, causing the formation of a non-equilibrium phase composition.

Figure 3 XRD pattern of (a) non-solvent, (b) H₂O-mediated synthesis, and (c) CH₃OH-mediated synthesis.

Table 1 Mass percentage of product phase composition of LiZrF products synthesized under different reaction conditions.

On the other hand, both LiF and ZrF₄ were well dispersed under methanol solution at RT, resulting in the propagation of a non-equilibrium reaction that enhanced the composition of the LiZrF product. As shown in Figure 3c, under the methanol mediator, the XRD pattern of LiZrF exhibited the main structure of Li₂ZrF₆ ($P21/c$) phase formation with a small amount of Li₂ZrF₆ ($P \overline{31} m$) formation. Generally, Li₂ZrF₆ ($P \overline{31} m$) corresponded to the composition in the phase diagram at X(ZrF₄) equal to 0.33, which was preferred as the suitable crystal orientation for lithium-ion conductors [11,19]. However, it could be noted that a part of LiF still existed in the sample, which would be caused by the rapid precipitation rate of a small amount of LiF dissolved in methanol near the completion of the drying process. Due to the high drying rates, an unbalanced phase composition might form, resulting in an unstable thermodynamic structure and limiting the synthesized material under the temperature range operation.

To demonstrate the effect of the annealing evaporation rate of the methanol mediator on the crystal phase composition, the dried temperature condition to synthesize LiZrF was investigated under 70°C/150°C, 50°C/150°C, and 25°C/150°C. The respective XRD pattern and the phase composition analysis are shown in Figure 4 and Table 1, respectively. Figure 4a exhibits the crystal phase compositions of the LiZrF dried under the temperature condition of CH₃OH-70°C/150°C, the LiZrF composite was formed in a ternary mixture dominated by Li₂ZrF₆ ($P21/c$). Compared to the phase composition of LiZrF generated under CH₃OH-50°C/150°C (Figure 4b), the phase composition was significantly improved, which completely formed the crystal phase of Li₂ZrF₆ ($P \overline{31} m$) without obviously impurity phase formation. This result could be suggested that the composition of the product might be effectively regulated by controlling the evaporation temperature of the mediator, and the lower evaporation rate was conducive to the formation of the equilibrium phase, which could not only improve the purity of the product, but also improve the stability of the material. Nevertheless, the reaction was difficult to occur without the external input of extra energy under the dried condition of 25°C/150°C (Figure 4c). Despite the same reaction time, almost no reactant was obtained compared to the sample obtained at 50°C/150°C. Due to many residual reactants, the composition mass fraction of dried condition at 25°C/150°C was uncounted in CH₃OH-mediated precursor. However, the reaction rate was still significantly higher than the none- mediator precursor. Hence, the experiment result suggested that the dried temperature condition of CH₃OH-50°C/150°C was considered to be the optimal reaction condition for synthesizing Li₂ZrF₆ ($P \overline{31} m$)-SE via methanol mediated synthesis pathway. Additionally, according to the correspondence to ICSD No. PDF#72-1291, Figure 4d exhibits the crystallographic model of Li₂ZrF₆ ($P \overline{31} m$), fluorine atoms were placed in the space with hexagonal close-packed (HCP) arrangement, this kind of arrangement is very common in structures consisting of small anions and large cations [30]. LiF₆ and ZrF₆ octahedrons are connected with a common vertex. Therefore, the chemical formula of Li₂ZrF₆ would be proposed in terms of Li₂ZrV’₃F₆, where V’’ represented the intrinsic vacancies. These intrinsic vacancies might be essential in the ionic conductivity of the material [19].

Figure 4 XRD pattern of (a) CH₃OH-70°C/150°C, (b) CH₃OH-50°C/150°C, (c) CH₃OH-25°C/150°C, and (d) crystallographic structure model of L₂ZrF₆ ($P \overline{31} m$).

To investigate the morphology of Li₂ZrF₆ as a SE compared with LiF, FESEM analysis was indicated, as shown in Figure 5. The surface morphology of LiF (Figures 5a, b) exhibited the cubic formation of a face-centered cubic (FCC) lattice stabilized by a strong bonding between $Li^+$ and $F^-$ ions. Thus, making LiF has the widest electrochemical window [18]. Although, the strong bonding between $Li^+$ and F^- on the one hand contributes to the high stability of LiF, it restricts the kinetic transportation of $Li^+$ [18,19,34]. Through Zr substitution via methanol-mediated synthesis, the surface morphology of the Li₂ZrF₆-based SE changed significantly with $Zr^{4+}$ doping generating a disorder rock-salt LiF structure, as shown in Figures 5c, d. In particular, three $Li^+$ cations would be removed by one metal ion doping, and two vacancies would be introduced in the octahedral interstitial locations. As a result, the vacancies occupy approximately 33.3% of octahedral substitutional sites in the structure of Li₂ZrF₆. Presumably, these vacancies originated from aliovalent metal-cation doping of $Li^+$ and were crucial to increase the ionic conduction of Li₂ZrF₆. The grouping of vacancies and vacancy clusters within the grain boundary, crystal lattice, and defect chemistry served as important parameters for $Li^+$ migration via the vacancy-rich environments, which were attributed to the morphology change that significantly corresponded to the XRD pattern. Surface morphologies at different magnifications of the cathode in SSB compared between LiF and LiZrF were presented in Figures 5e, g, respectively. According to metal cation substitution in LiZrF ionic crystals with different radius atom, crystal phase might be arranged based on their electrostatic forces. As can be seen those three components in the composite cathode pellet with Li₂ZrF₆ remained in relatively closer contact than LiF without the obvious cracks in the pellet even at ultrahigh magnification. This result implied that the effective 3-dimension conduction channels for ions and electrons diffusion.

Figure 5 Surface morphology of (a,b) LiF particles, (c-d) Li₂ZrF₆ composite, composite cathode with (e-f) LiF, and (g-h) Li₂ZrF₆, respectively, at different magnifications.

Principally, EIS measurement is used to evaluate the ion conduction of the electrolyte. It is operated by applying a small amplitude AC potential wave with a different frequency to the electrochemical system and then measuring the curve of the ratio of the AC potential to the current signal with the sine wave frequency. The real part of the impedance is equal to 0 for the perfect capacitor, and then the Nyquist plot is a straight-line recombining with the imaginary axis. The ionic conductivity was calculated according to Eq. (1): [16]

\[ \sigma=\dfrac{L}{\left(R \ x \ S\right)} \tag{1} \]

where $\sigma$ is an ion conductivity (S/cm), $R$ is the resistance determined from impedance spectra (Ω cm), $L$ is sample thickness, and $S$ is the contact area between two electrodes (cm²).

The activation energy of the ion migration in the material might also be obtained according to the Arrhenius equation in Eq. (2): [16]

\[ \sigma=\frac{A}{T}e^{-\frac{E_a}{k_BT}} \tag{2} \]

where $\sigma$ is ionic conductivity of material, $A$ is the Arrhenius constant, $E_a$ is the activation energy of ion migration, $k_B$ is the Boltzmann constant, and $T$ is the Kelvin temperature. Then convert Eq. (2) into antiderivatives:

\[ ln\sigma T=-\dfrac{E_a}{k_B}\cdot\dfrac{1}{T}+ln A \tag{3} \]

According to Eq. (3), $ln\sigma T$ is linear with $1/T$ and the slope is equal to $-E_a/k_B$. Therefore, the activation energy of ion migration can be calculated from the slope of the line obtained by plotting $ln\sigma T$ against $1/T$. It is a formulation of the relationship between temperature dependence and reaction rates. With the Arrhenius equation, it could be approximately that the ion moving rate of reaction increases with the increase of temperature leading to the enhancement of ion conductivity. The term of $e^{-E_a/k_B T}$ indicates the fraction of molecules with energy greater than or equal to $E_a$.

The ionic conductivity of the Li₂ZrF₆ was studied by electrochemical impedance spectroscopy (EIS). Because of an appealing feature of solid-state structures, through Zr substitution Li₂ZrF₆ by defeat chemistry could indicate the effectiveness of $Li^+$ hopping in the lattice. The $Li^+$ hopping process was sensitive to activated $Li^+$ and vacancy concentrations, so it could enhance conductivity by changing orders of magnitude [15] (Figure 6).

Figure 6 (a) Nyquist plots at different temperatures and (b) Arrhenius conductivity plot of Li₂ZrF₆.

The SSB system’s resistance is mainly composed of three parts: bulk resistance, grain boundary resistance and electrode/electrolyte interface resistance. The equivalent circuits of these three parts can be regarded as a resistor and a capacitor in parallel, and connected in series. The resistance of each part may be different according to the material’s properties, but the capacitance order is similar. The theoretical value of bulk capacitance, grain boundary capacitance and electrode/electrolyte interface capacitances are 10^-12 F, 10^-9 F and 10^-6 F, and the resistance of each part can be designated. In general, the ionic conductivity of SSE is defined as the reciprocal of the sum of bulk resistance and grain boundary resistance [35,36]. The optimally crystallized Li₂ZrF₆ synthesized under moderate conditions with a CH₃OH-mediated precursor exhibited the highest conductivity of 2.40 × 10^-8 S/cm at RT with $E_a$ evaluated at 0.15 eV. (Figure 6b). The substitution of $Zr^{4+}$ in the optimized crystal structure of Li₂ZrF₆ ($P \overline{31} m$) could induce the vacancies clusters. At the same time, the lattice structure would be affected by the radius of different metal cations in the metal halide electrolyte. The stacking crystal structure of metal- ion by a smaller radius $Zr^{4+}$ ion was encouraged by an hcp-T to hcp-O transition leading to the reduction of $E_a$ in Li₂ZrF₆ composite-based SE, compared with $E_a$ of LiF which is equal to 0.73 eV [37]. It could be implied that the ion conduction value of Li₂ZrF₆ based SE was significantly related to the preferred orientation and the (131) plane/(001) plane peak intensity ratio in XRD patterns generated via methanol-mediated precursor.

The Li₂ZrF₆ composite material exhibited excellent defect density compared with bivalent and trivalent metal elements due to the introduction of a tetravalent element of Zr. Moreover, to demonstrate the effect of $Li^+$ concentration on the ion-conduction, the Li₂ZrF₆ ($P \overline{31} m$) was synthesized from the different stoichiometric molar ratio of LiF:ZrF₄ of 2:0.8, named as Li-Rich LiZrF, which was studied and compared with LiZrF. XRD analysis demonstrated that the LiZrF and Li-Rich LiZrF products were stabilized structure by the formation of Li₂ZrF₆ ($P \overline{31} m$).

The XRD result of the two samples was further refined with the Rietveld refinement method, a simple method widely used to demonstrate the space group, atomic coordinates and cell size [38]. The results are shown in Figure 7, Table 2 and Table 3. Due to the low atomic scattering factor of Li and F, the site occupancy information could not be accurately obtained, so the occupancy rate of Li and F had not been refined in this part. Comparing the initial reactants amount and site occupancy, the occupancy of the Zr site was different, with the decreasing molar ratio of ZrF₄, the occupancy of Zr sites decreased. The Zr content in the Li-Rich LiZrF structure was nearly 20% lower than that in LiZrF. Since the introduction of each tervalent Zr atom in Li₂ZrF₆ ($P \overline{31} m$) would replace the surrounding four $Li^+$, 36.23% of the octahedral vacancy is occupied by $Li^+$ in LiZrF, while 39.68% was occupied by $Li^+$ in Li Rich-LiZrF, increasing the content of $Li^+$ by about 10%. Therefore, the strategy of regulating the structural characteristics by controlling the molar ratio of reactants was succussed, the number of vacancies in the material could be adjusted effectively, and the concentration of lithium ions in the material was significantly increased.

Figure 7 (a) XRD pattern and (b) the corresponding Rietveld refinements of LirZF compared with Li-Rich LiZrF, respectively.

Table 2 Refined parameters and crystallographic data of LiZF and Li-Rich LiZF.

Table 3 Rietveld refinement results of LiZrF and Li-Rich LiZrF.

EIS also studied the ion conduction of Li-Rich LiZrF compared to LiZrF. The Li-Rich LiZrF and LiZrF pellets were sandwiched between two cylindrical stainless-steel current collector measurements. From the electrochemical impedance data, the diameter of the semicircle is characteristic of the bulk interfacial resistance of electrolytes between electrodes. Following Eq. (1) and (2), Figure 8a shows the respective ionic conductivity of Li-Rich LiZrF with the reciprocal of temperature. The ionic conductivity of Li-Rich LiZrF at room temperature was nearly twice increased to 3.89 × 10^-8 S/cm with $E_a$ of 0.21 eV (Figure 8b) while the LiZrF exhibited only 2.40 × 10^-8 S/cm which were considered by the Boltzmann-Einstein equation [39,40]. Following the relation between these equations, it could be implied that the addition of the sited-number of ions per unit volume (vacancy, Z) by increasing the $Li^+$ concertation led to enhancement of the available ion movement. The electronic conductivity of the material was also measured by DC polarization under applying constant potential. Due to a blocking electrode, the electrode cannot provide ions, which are polarized at the electrode. After the ions in the material were exhausted, it was difficult for the ions to continue moving, so the movement of electrons only generated the current at this time. The electrical conductivity can be evaluated by using Eq. (4):

\[ \sigma=\begin{bmatrix}\frac{\Delta V}{I}\times\frac{w\cdot t}{s}\end{bmatrix}^{-1} \tag{4} \]

where $\sigma$ is an electron conductivity (S/cm),$\Delta V$ is voltage is difference,$I$ is the applied current, w is the diameter of SE pellet, t is the SE pellet thickness, and s is the electrode separation.

The test results are shown in Figure 8c, d, and the calculated electronic conductivity of Li-Rich LiZrF was 3.42 × 10^-9 S/cm, which was lower than the ionic conduction of Li-Rich LiZrF, approximately 10% of the overall conductivity [41]. The addition of functional lithium-fluoride contents could appear to be a facile synthesis with the simple method to suppress Li dendrites by reducing the compact conformal SEI layer [42]. In addition, Li-Rich LiZrF also proposed the from these experiment results, it could be expected that Li-Rich LiZrF was employed as an alternative material for solid electrolyte promising for the future development of all-solid-state lithium batteries.

Figure 8 (a) Nyquist plots at different temperatures, (b) Arrhenius conductivity plot, (c) DC polarization curves, and (d) equilibrium current response at different voltages of Li-Rich LiZrF, respectively.

4. Conclusion

In this work, a novel method route with methanol precursor has been proposed for synthesizing LiZrF electrolytes. Li₂ZrF₆ was successfully synthesized by a simple one-step method, and the effects of different mediators, the stoichiometric molar ratio of LiF:ZrF₄ and temperature conditions on the synthesis were discussed. The annealing dried condition of CH₃OH-50°C/150°C was considered the optimal reaction condition for synthesizing Li₂ZrF₆ ($P \overline{31} m$)-SE via methanol mediated synthesis pathway. Moreover, the ionic conduction of the material was further improved by adjusting the proportion of the initial reactants generated in terms of Li-Rich LiZrF Li₂ZrF₆ ($P \overline{31} m$)-SE. The obtained ionic conduction was increased to 3.89 × 10^-8 S/cm, and the $E_a$ of ion migration was equal to 0.21 eV. The lower $E_a$ of Li₂ZrF₆ than the LiF through the substitution of $Zr^{4+}$ would be suggested that the framework of LiZrF-based SE was formed by the sublattice stacked layer of anion and influenced by the volume and polarity of the cation species leading to enhancement of ion conduction, which was demonstrated outstanding electrochemical performance. Besides, the substitution of $Zr^{4+}$ in the crystallographic layer also affected the air stabilization of LiZrF-based SE electrolytes greater than sulfide-based SEs. Experimental results suggested that our development of the synthesized LiZrF-based SE via a methanol-mediated precursor provided a simple method at mild conditions that could improve the ion conductivity of LiF.

Acknowledgments

The authors are grateful for the financial support of this research from Japan Science and Technology Agency, Strategic International Collaborative Research Program (Grant No.: JPMJSC18H1) and Japan Science and Technology Agency, Open Innovation Platform with Enterprises, Research Institute and Academia (Grant No.: JPMJOP1843).

Author Contributions

The manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript. Conceptualization: N. Saito; Methodology: J. Moon and R. Zheng; Investigation (experimental work): J. Moon, R. Zheng, and S. Liu; Writing original draft: J. Moon, R. Zheng, and S. Thiangtham; Writing review and editing: S. Thiangtham and C. Chokradjaroen; Supervision: N. Saito; Funding acquisition: N. Saito and Y. Sawada.

Competing Interests

The authors declare no competing financial interest.