Ionic Conduction and Electric Modulus in Li₂S–CaS and CaX₂ (X = F, Cl, Br, and I) Nanocomposites

Abstract

Technology for ion conduction in Li₂S nanostructure has attracted considerable interest for achieving the intrinsic performance of Li-S batteries and for obtaining highly conductive materials. Herein, a quantitative relationship between ion transport and relaxation behavior in the Li₂S–CaS nanocomposite is revealed based on complex conductivity and electric modulus formalism. Li₂S–CaS nanocomposites prepared by high-energy ball milling show higher conductivity and activation energy for conduction than pure Li₂S. The activation energy for relaxation is lower than the activation energy for conduction in 80Li₂S·20CaS (mol%). The long-range ion transport involves high activation energy compared with the hopping of carriers at localized states in 80Li₂S·20CaS. Additionally, the dual doping of Ca and halogen reduces the activation energy for conduction and improves the conductivity of Li₂S. The findings show important insight for understanding ion transport in nanocomposites containing a Li₂S nanostructure.

1. Introduction

Lithium sulfide is attractive as a cathode material for Li–S batteries owing to its interesting properties, e.g., high electrochemical capacity (1,167 mAh g⁻¹).^1,2 A Li-S battery based on a Li₂S cathode allows for coupling with a high-capacity anode such as Si and Sn; this is an appealing alternative for conventional Li ion batteries.^1,3 However, the low electronic and ionic conductivity of Li₂S limits the complete utilization of its intrinsic capacity. Achievements of high-power density in Li-S batteries require understanding the transport properties of Li₂S. Recently, multivalent cation doping to Li₂S improved its ionic conductivity and extracted a high battery performance as a cathode in all-solid-state Li-S batteries.^4–6 Both Li₂S–Al₂S₃ and Li₂S–MgS solid solution demonstrated a higher conductivity (by two orders of magnitude of 10⁻⁶ S cm⁻¹ at room temperature) than the milled Li₂S because of the nanoionics. In addition, AlI₃-doped Li₂S cathode materials offered enhanced reversible capacity.⁷ Based on solid-state chemistry, the multiple unique nature of nanostructured materials containing Li₂S has inspired several devices that use rechargeable batteries.

Lithium sulfide plays an important role as a starting material for obtaining highly conductive materials for solid-state batteries.⁸ Studies on sulfide ionic conductors were originally conducted in Li₂S-based glass conductors. In the 1980s, Li₂S-based glass, e.g., Li₂S–P₂S₅–LiI,⁹ delivered an ionic conductivity of the order of 10⁻³ S cm⁻¹ at room temperature. Extensive studies have discovered crystalline ionic conductors of the order 10⁻² S cm⁻¹ at room temperature, such as Li₁₀GeP₂S₁₂,¹⁰ Li₇P₃S₁₁,¹¹ and argyrodite.^12,13 These material classes have evolved through cation substitution on P-site, as well as anion substitution. Recently, cation substitution on the site for a charge carrier identified superionic conductors involving Li_5.4Al_0.2PS₅Br,¹⁴ Li_5.35Ca_0.1PS_4.5Cl_1.55,¹³ and Na_2.730Ca_0.135PS₄.¹⁵ Interestingly, both ionic conductivity and activation energy of the Na_2.730Ca_0.135PS₄ increased from a pristine Na₃PS₄ solid electrolyte. Ca doping in Li ion and sodium ion conductors is of considerable interest for expressing fast ionic conduction.^13,16,17

Herein, we report the effect of adding Ca ions and dual ions on ion-transport properties in a Li₂S structure. Note that (100 − x) Li₂S·xCaS (mol%) nanocomposites, prepared by high-energy ball milling, exhibit crystallite sizes ranging from 7 to 31 nm. The Li₂S–CaS nanocomposites demonstrate higher conductivity and activation energy for conduction than the milled pure Li₂S. To understand the fundamental relationship between activation energy for conduction and relaxation in (100 − x)Li₂S·xCaS, we examined the complex conductivity and electric modulus formalism. The activation energy for electric relaxation is lower than the activation energy for conduction in 80Li₂S·20CaS. The result indicates that long-range ion transport, including the contribution of a grain boundary, involves higher energy barriers than the hopping of carriers at localized states. The dual doping of Ca and halogen into Li₂S reduces the activation energy of conduction and improved the conductivity. These important insights can provide an understanding of ion transport properties in functional materials for batteries involving Li₂S.

2. Experimental

2.1 Preparation of the Li₂S–CaS and Li₂S–CaX₂ (X = F, Cl, Br, I)

Lithium sulfide (99.9 %, Mitsuwa), CaS (99.99 %, Kojundo Laboratory), CaF₂ (99.99 %, Sigma-Aldrich). CaCl₂ (99.99 %, Sigma-Aldrich), CaBr₂ (99.98 %, Sigma-Aldrich), or CaI₂ (99.999 %, Sigma-Aldrich) were mixed at appropriate molar ratios. Then, the powders obtained were loaded in 45 ml of zirconia pots with φ 10 mm zirconia balls and subjected to high-energy ball milling with planetary ball milling (Fritsch Pulverisete 7 Premium line) at a rotation speed of 510 rpm for 10 h (1 h milling; 15 min pause).

2.2 Material characterization

Powder X-ray diffraction (XRD) measurements were performed under 2θ = 10°–80° in a step interval of 0.02° at a scan rate of 0.2° min⁻¹ using a Rigaku Ultima IV diffractometer. An X-ray beam was generated using CuKα radiation (40 kV, 30 mA). For XRD measurements, we used an XRD holder with a beryllium window (Rigaku). Subsequently, structural refinements were analyzed by the Rietveld method using the RIETAN-FP computer program.¹⁸ The crystallite size, D, and the microstrain, ε, for phases in the prepared nanocomposites were calculated from the Williamson–Hall method presented in Eq. 1:   

\begin{equation} \beta \cos\theta = C\epsilon \sin\theta + \frac{K\lambda}{D}, \end{equation} (1)

where β is the integral breadth, K and C are constants (0.94 and 4), and λ is the wavenumber of X-rays. Note that instrumental broadening was corrected using crystalline Li₂S without ball milling. Field emission scanning electron microscopy (FE-SEM) characterization was carried out on a Hitachi-S4800 with an EDS (ULTIM MAX, Oxford Instrument).

2.3 Electrochemical measurements

The overall conductivities of prepared samples were determined by alternating current impedance spectroscopy (SI 1260, Solatron) in a frequency range of 1 MHz to 10 Hz under a dry Ar flow. To fabricate cells for electrochemical impedance spectroscopy (EIS) measurements, each sample (∼80 mg) was filled in a holder made of polyether ether ketone (PEEK) with two stainless steel rods as blocking electrodes. We applied a uniaxial pressure of 256 MPa to pellets having diameters of ∼10.0 mm at room temperature. The complex conductivity, σ$^{*}$, of samples was determined from the complex impedance, Z$^{*}$, obtained using EIS data and is reported as follows:   

\begin{equation} \sigma^{*} = \sigma' + j\sigma'' = \frac{l}{AZ^{*}} = \frac{l}{AZ'} + j\frac{l}{AZ''}, \end{equation} (2)

where A and l are the contact area and the thickness of the pellets, respectively.

The complex electric modulus, M$^{*}$, was determined from Eqs. 3 and 4, which are related to complex conductivity and permittivity, respectively, as follows:   

\begin{equation} \varepsilon^{*} = \varepsilon'' - j\varepsilon'' = \frac{\sigma''}{\omega \varepsilon_{0}} - j\frac{\sigma'}{\omega\varepsilon_{0}} \end{equation} (3)

  

\begin{align} M^{*} &= \frac{1}{\varepsilon^{*}} = \frac{1}{(\varepsilon' - j\varepsilon'')} \\ &= M' + jM'' = \frac{\varepsilon'}{(\varepsilon')^{2} + (\varepsilon'')^{2}} + j\frac{\varepsilon''}{(\varepsilon')^{2} + (\varepsilon'')^{2}}, \end{align} (4)

where ω (= 2πf) is a radial frequency expressed in radians/second parameter and ε₀ is the permittivity of free space.

3. Results and Discussion

3.1 Structural characterization

Figure 1a shows the XRD patterns of (100 − x)Li₂S·xCaS with Si powder as a standard. The XRD patterns for all samples show diffraction peaks attributed to a crystalline phase of Li₂S with the space group Fm-3m. These patterns show no change in peak position and no side reaction after adding CaS. The intensity of the peak at 27° decreases with an increasing CaS content, whereas the intensity of the most intense peak at 31° in the XRD pattern of the CaS phase increases. These experimental results indicate that Li₂S–CaS prepared by high-energy ball milling comprises a mixture of Li₂S and CaS phases without the substitution of Ca in Li ions. To examine lattice distortion and mean crystallite size of mechanically treated Li₂S–CaS, we performed Rietveld refinement based on XRD data. Figure 1b shows the Rietveld refinement pattern for 80Li₂S·20CaS. A multi-phase crystal model, including Li₂S and CaS phases, offers R factors with good fitness, which are sufficiently low to support the extracted structural model. R_wp, R_B, and R_F obtained for the Li₂S phase are 10.1 %, 2.4 %, and 2.0 %, respectively. Figure 1c shows the mean crystallite sizes and the microstrain as a function of composition x in (100 − x)Li₂S·xCaS. Both the mean crystallite size and microstrain were estimated using the Williamson–Hall method (Figs. S1a and S1b). The Williamson–Hall plot for the sample with x = 50 deviates from a linear relationship because of the anisotropic line broadening. The Li₂S–CaS prepared by high-energy ball milling forms nanocrystalline composites with crystallite sizes ranging from ∼7–31 nm. The crystallite sizes demonstrate a decreasing trend with an increasing CaS content. The microstrain of the Li₂S phase in the composites increases with the addition of CaS. Figure 2 shows FE-SEM images of (100 − x)Li₂S·xCaS particles (0 ≤ x ≤ 40). The particle sizes of all samples are identical, which are estimated to be ∼50 µm. The particle morphology is rougher with increasing CaS content. On the other hand, no calcium atoms were detected on the particle surface of the Li₂S-CaS doping from EDS mapping (Fig. S2). This experimental result indicates that the Li₂S-CaS forms nanocomposite particles covered with the Li₂S phase.

Figure 1.

(a) The XRD patterns of (100 − x)Li₂S·xCaS with added Si powder. The bottom profile shows CaS as a reference. (b) The Rietveld refinement pattern for 80Li₂S·20CaS. The observed diffraction intensities and calculated patterns are denoted by red plus signs and a solid green line, respectively. The blue trace at the bottom represents differences between calculated and experimental patterns. The short green bars below the observed and calculated profiles indicate the positions of the allowed Bragg reflections of Si, CaS, and Li₂S in order from the bottom. (c) The crystallite sizes and the microstrain of the Li₂S and CaS phases in (100 − x)Li₂S·xCaS. Circles and squares correspond to the Li₂S and CaS phases, respectively.

Figure 2.

FE-SEM images of (100 − x)Li₂S·xCaS particles (0 ≤ x ≤ 40).

3.2 Complex impedance spectroscopy

Figure 3a shows the real part σ′ of the complex conductivity of 80Li₂S·20CaS as a function of frequency. At lower frequencies, the impedance spectra of obtained nanocrystalline samples demonstrate an obvious direct current (DC) plateau, which indicates long-range ion transport.¹⁹ The electronic conductivity of Li₂S is lower than the ionic conductivity by at least two orders of magnitude in the temperature range from 30 to 450 °C; thus, the observed DC plateau is attributed to ion transport.²⁰ The frequency-dependent dispersive region is observed at a higher frequency. Frequently, this observation has been interpreted as a fingerprint of the forward and backward jumps of carriers involving a short length scale.^21,22 This suggests that the carrier repeats the polarization and its relaxation at a stable site for the carrier. A higher temperature allows for long-range ion transport in samples over a broader frequency range. In the high-frequency region, the frequency dependence of conductivity obeys Jonscher’s power law,¹⁹ i.e., σ ∝ ωⁿ where ω is the angular frequency and exponent n is the degree of interaction between mobile ions as well as the lattice around them. Most ionic conducting compounds demonstrated no deviation from a value between 0.6 and 1.0 of exponent n.¹⁹ However, the exponent n demonstrates a value of >1 in 80Li₂S·20CaS at lower temperatures than 90 °C. Furthermore, dispersive behavior in pure Li₂S yields an exponent n of 1.2 at a temperature from 130 to 170 °C (Fig. S3). The higher exponent n may have originated from the migration of carriers from the occupied site to other sites via quantum mechanical tunnelling between asymmetric double-well potentials.^23,24 This finding is consistent with the previous computational results that Li ions in the Li₂S crystal exist in isolated tetrahedral sites in the face-centered cubic sulfur framework.²⁵ In contrast, the 80Li₂S·20CaS demonstrates an exponent n of 0.63 at 110 °C. The power law exponent n is related to the effective dimension of ion diffusion pathways. The exponent of the 80Li₂S·20CaS indicates the presence of 3-dimensional Li-ion pathways in the structure.²⁶ All samples follow the Arrhenius equation shown in Eq. 5:   

\begin{equation} \sigma_{\text{DC}}T = \sigma_{0}\exp\left(-\frac{E_{\text{a},\text{DC}}}{k_{\text{B}}T}\right), \end{equation} (5)

where σ_DC is the DC conductivity of the samples, T is the temperature, σ₀ is the pre-exponential, E_a,DC is the activation energy, and k_B is the Boltzmann constant. Figure 3b shows the DC conductivity of the dependence temperature in (100 − x)Li₂S·xCaS. The ionic conductivity for Li₂S is highly dependent on the content of CaS that is an insulator for Li ion. 80Li₂S·20CaS demonstrates the highest conductivity of 1.4 × 10⁻⁴ S cm⁻¹ at 150 °C among the prepared samples; this value is two orders of magnitude higher than that of the milled pure Li₂S (2.1 × 10⁻⁶ S cm⁻¹). This improvement could be attributed to the enhanced ion transport in the interface between the Li₂S and CaS phases involving nanocrystallite sizes, i.e., the nanoionics.^27–30 The conductivity of (100 − x)Li₂S·xCaS demonstrates a decreasing trend in the sample with an increasing CaS content from 20 to 50 mol%, which is assumed to have caused discontinuity in the conduction region in nanocomposites as the CaS insulator content increased. Therefore, the optimal composition for conductivity in nanocomposites is determined by the relationships between the enhancement of local conduction at the interface and the continuity of the conduction region. Figure 3c shows the activation energy and pre-exponential factor extracted from the Arrhenius plots in (100 − x)Li₂S·xCaS. Both the activation energy, E_a, and the pre-exponential factor, σ₀, increased in samples with the CaS content x in the range of 0–20. Based on the hopping theory, the pre-exponential factor is expressed in Eq. 6:   

\begin{equation} \sigma_{0} = \frac{\gamma Nq^{2}a_{0}^{2}}{6k_{\text{B}}}H_{\text{R}}f_{0}, \end{equation} (6)

where, γ is the fraction of charge carriers that are mobile, N is the total density of charge carriers, q is the charge of the mobile ion, a₀ is the jump distance, Haven ratio H_R expresses the presence of possible cross correlation in the diffusion, and f₀ corresponds to the frequency where σ′(f) crosses over from power law to DC.²⁶ H_R < 1 means the presence of correlations between the mobile ions or electronic conduction provides. The pre-exponential factor of 90Li₂S·10CaS is 26-fold that of pure Li₂S, and f₀ of 90Li₂S·10CaS is 14-fold that of pure Li₂S. This fact implies that the increase in f₀ is the primary reason for the increase in the pre-exponential factor. Therefore, the formation of the Li₂S-CaS nanocomposites increases the rate of successful jumps in mobile cations, leading to the enhancement of the ionic conductivity. The conductivity for a broad range of inorganic materials often increases with a reduction in activation energy, whereas the conductivity of Li₂S–CaS increases with increasing activation energy. This is consistent with the results observed in Na_3–2xCa_xPS₄ and Li_3−2xCa_xPS₄ as superionic conductors.^15,17 Moon et al. reported that high conductivities and activation energies in Na_3−2xCa_xPS₄ were attributed to an impediment caused by the increased energy barriers close to Ca ions in the conduction pathways for the carrier.¹⁵ Therefore, high activation energy and conductivity in Li₂S–CaS nanocomposites can be described by the notion that the ion transport properties are governed by the conduction of mobile Li ions near the interface with the CaS nanocrystalline. In the sample with the CaS content x in the range between 20 and 50, the activation energy decreases. The decreased activation energy for ion conduction might be attributed to a strong amorphization for Li₂S phase, which is favorable for ion transport. The relative densities of the pellets of Li₂S-CaS composites show a decreasing trend with the addition of CaS (Table S1). This result suggests that the CaS phase in Li₂S-CaS composites shows poor mechanical deformability.

Figure 3.

(a) Conductivity isotherms of 80Li₂S·20CaS at different temperatures ranging from 50 °C to 150 °C. The solid lines represent fitting as per Jonscher’s power law, i.e., σ ∝ ωⁿ at the dispersive regime. (b) Arrhenius plots of the DC conductivities for (100 − x)Li₂S·xCaS (0 ≤ x ≤ 50). (c) Activation energies and pre-exponential factors for (100 − x)Li₂S·xCaS (0 ≤ x ≤ 50).

3.3 Electric modulus analysis

The investigation of dielectric peaks has long been an effective technique for examining microscopic motion in amorphous solids.²² Recently, an electric modulus analysis revealed ion dynamics in crystalline solid electrolytes.^31–33 To establish quantitative relationships between the activation energy for conduction and relaxation in (100 − x)Li₂S·xCaS, we investigated based on complex electric modulus formalism. Figure 4a shows the imaginary part of electric modulus, M′′, at different temperatures versus frequency in 90Li₂S·10CaS. Figure S4 shows the electric modulus dependence frequency for samples with other compositions. The relaxation peaks appear in all samples, and their positions shift toward higher frequencies with increasing temperature. The relaxation peak shows a transition from the overall averaged long-range to the short-range carrier migration, which corresponds to the relaxation frequency, f_max. The mean relaxation time, τ, is determined from the τ = 1/2πf_max relationship and follows the Arrhenius equation in Eq. 7 as follows:   

\begin{equation} \tau = \tau_{0}\exp\left(-\frac{E_{\text{a},\tau}}{k_{\text{B}}T}\right), \end{equation} (7)

where τ₀ is the relaxation time. Figure 4b shows the plot of the variation of log(τ_M⁻¹) versus inverse temperature for the (100 − x)Li₂S·xCaS. Despite an increase in activation energy, E_a,DC, for DC conductivity, the activation energy, E_a,τ, for the relaxation time is maintained in the sample with CaS content x from 0 to 20. In 80Li₂S·20CaS, the E_a,τ value is observably lower than the E_a,DC value. The relaxation process indicates the hopping energy of carriers between localized states, whereas the conduction process is affected by long-range ion transport, including the contribution of the interface between Li₂S and CaS. Hence, long-range ion transport with high activation energy in 80Li₂S·20CaS should be explained by the increase in the energy barrier near the interface between the Li₂S and CaS nanocrystallines.

Figure 4.

(a) The imaginary part of the modulus, M′′, versus the frequency of 90Li₂S·10CaS at temperatures ranging from 70 °C to 150 °C. (b) The τ_M⁻¹ rate versus the inverse temperature of (100 − x)Li₂S·xCaS. Dashed lines represent the linear fit. (c) For comparison, the activation energy for electric relaxation and conduction as a function of composition x in (100 − x)Li₂S·xCaS.

3.4 The effect of dual-doping CaX₂ (X = F, Cl, Br, I)

Dual doping is an effective strategy for improving ionic conductivity. In addition to CaS doping, we examined the effect of adding CaX₂ (X = F, Cl, Br, and I) on structures and ionic conductivities of Li₂S. Figure 5 shows the XRD patterns of (100 − x)Li₂S·xCaX₂ containing Si powder as a standard sample. For all samples, the XRD patterns of (100 − x)Li₂S·xCaX₂ (X = F, Cl, Br, and I) are observed as the peaks attributed to Bragg reflection of Li₂S crystalline phase with the space group Fm-3m. The peak intensity decreases and peak line-width broadens with an increasing CaX₂ (X = F, Cl, Br, and I) content. This shows the disordered structure arising from the substitution of the anion. Note that lithium halide is formed with an x value of 20 in (100 − x)Li₂S·xCaX₂ (X = Cl and Br), as well as an x value of 15 in (100 − x)Li₂S·xCaI₂. No change in the peak position following the addition of CaX₂ (X = F and Cl) is observed; however, the peak shifts toward a lower angle after adding CaX₂ (X = Br and I) (Fig. S5). These peak shifts indicate that the halogen had been substituted for S²⁻ because the variations of the lattice plane spacing correspond to the size of ionic radii of the added halogen. Unsubstituted excess halogen reacts with Li ions, thus forming a lithium halide in the samples. Despite the large difference between S²⁻ and F⁻ (1.84 versus 1.33 Å), the peak position in Li₂S–CaF₂ is independent of composition. Based on chemical defects, Lorger et al. reported that in Li₂S doped with LiF, fluoride ions created comparatively large interstitial sites in the antifluorite structure of Li₂S.²⁰

Figure 5.

The XRD patterns of (a) the (100 − x)Li₂S·xCaF₂, (b) (100 − x)Li₂S·xCaCl₂, (c) (100 − x)Li₂S·xCaBr₂, and (d) (100 − x)Li₂S·xCaI₂ with added Si powder as a standard sample.

Figure 6 plots the temperature dependence of conductivities for (100 − x)Li₂S·xCaX₂ (X = F, Cl, Br, and I). The ionic conductivities of all samples containing additives significantly improve compared with that of milled Li₂S. Among each material system, 90Li₂S·10CaI₂, 80Li₂S·20CaBr₂, 85Li₂S·15CaCl₂, and 95Li₂S·5CaF₂ demonstrate the highest ionic conductivity of 3.6 × 10⁻⁵, 1.3 × 10⁻⁵, 1.5 × 10⁻⁵, and 1.8 × 10⁻⁷ S cm⁻¹ at 50 °C, respectively. This improvement is attributed to the creation of Li⁺ vacancies following the substitution of halogen and the nanoionics. All of the samples follow the Arrhenius equation in Eq. 5 and demonstrate lower activation energy than pure Li₂S (Fig. 7). The intrinsic high electron-attracting ability of F⁻ and Cl⁻ could lead to a higher activation energy for ion transport because conduction at the applied temperature is determined by the electrostatic association that causes ionic charge carriers to be partially trapped at dopant sites.²⁰ However, the activation energy involves no correlation with the ionic radius of the introduced halogen in this study. The considerable difference in activation energy among material systems could not be explained only by the slightly different trapping energy based on the dopant halogen species. A high electronegativity for F and Cl ions may cause a disordered local framework structure in Li₂S crystals or a strong amorphization, in addition to the expansion of conduction pathways.³⁴ A systematic investigation reported that 90Li₂S·10CaI₂ demonstrates the highest ionic conductivity at 50 °C among the prepared samples owing to the presence of Li ion vacancies and expansion of Li ion pathway when introducing I⁻ with a larger ionic radius. Moreover, the ionic conductivity of 90Li₂S·10CaI₂ is higher than 80Li₂S·20LiI with identical I⁻ doping levels, which may be described by the enhanced ion transport in the interface between the I-doped Li₂S and CaS phases (Fig. S6). The activation energy of 90Li₂S·10CaI₂ reflects 0.53 eV; this is higher than that of 80Li₂S·20LiI (0.48 eV). Based on the electric modulus analysis for Li₂S·CaS composites, the increase in ionic conductivity and activation energy is anticipated to have originated from increased energy barriers near the interface of the CaS nanocrystalline.

Figure 6.

Arrhenius plots of the DC conductivities for (a) the (100 − x)Li₂S·xCaF₂, (b) (100 − x)Li₂S·xCaCl₂, (c) (100 − x)Li₂S·xCaBr₂, and (d) (100 − x)Li₂S·xCaI₂.

Figure 7.

Activation energies for conduction in (100 − x) Li₂S·xCaX₂ (X = F, Cl, Br, and I).

4. Conclusion

Our study reports the influence of CaS and CaX₂ (X = F, Cl, Br, and I) doping on the ion transport properties of Li₂S. The Li₂S-CaS composites involve crystallite sizes ranging from 7 to 31 nm, and demonstrated higher conductivity and activation energy for ion transport than the milled pure Li₂S. We demonstrate the quantitative relationship between the activation energy for conduction and relaxation in (100 − x)Li₂S·xCaS, based on complex conductivity and electric modulus formalism. In 80Li₂S·20CaS, E_a,τ values are lower than E_a,DC values. This suggests that long-range ion transport in the sample involves higher activation energy compared with the hopping of carriers at localized states, which should be influenced by the increase in energy barriers near the interface between the Li₂S and CaS nanocrystallines. The dual doping of Ca and halogen allows for decreasing the activation energy for conduction and increasing the conductivity of Li₂S. A systematic investigation reports that 90Li₂S·10CaI₂ demonstrates the highest ionic conductivity of 3.6 × 10⁻⁵ S cm⁻¹ at 50 °C among the dual doped samples; this conductivity is higher than 80Li₂S·20LiI with identical I⁻ doping levels. Moreover, the activation energy of 90Li₂S·10CaI₂ (0.53 eV) is higher than that of 80Li₂S·20LiI (0.48 eV). These findings support the existence of increased energy barriers near the interface of the CaS nanocrystalline. These fundamental insights can help gain an understanding of ion transport properties in functional materials for batteries involving Li₂S.

Acknowledgments

Part of this study was supported by the Advanced Low Carbon Technology Specially Promoted Research for Innovative Next Generation Batteries (JST-ALCA-SPRING: JPMJAL 1301) program of the Japan Science and Technology Agency. The authors would like to thank Enago (www.enago.jp) for the English language review.

Data Availability Statement

The data that support the findings of this study are openly available under the terms of the designated Creative Commons License in J-STAGE Data at https://doi.org/10.50892/data.electrochemistry.19793662.

CRediT Authorship Contribution Statement

Hirotada Gamo: Conceptualization (Equal), Data curation (Lead), Formal analysis (Lead), Investigation (Lead), Methodology (Lead), Visualization (Lead), Writing – original draft (Lead)

Nguyen Huu Huy Phuc: Conceptualization (Equal), Supervision (Equal), Writing – review & editing (Supporting)

Mika Ikari: Investigation (Supporting), Writing – review & editing (Supporting)

Kazuhiro Hikima: Project administration (Equal), Writing – review & editing (Lead)

Hiroyuki Muto: Funding acquisition (Supporting)

Atsunori Matsuda: Funding acquisition (Lead), Project administration (Equal), Resources (Lead), Supervision (Equal), Writing – review & editing (Supporting)

Conflict of Interest

The authors declare no competing financial interest.

Funding

Advanced Low Carbon Technology Research and Development Program: JPMJAL 1301

Footnotes

N. H. H. Phuc: Present address: Faculty of Materials Technology, Ho Chi Minh City University of Technology (HCMUT), 268 Ly Thuong Kiet Str., Dist. 10, Ho Chi Minh City, Vietnam

N. H. H. Phuc: Present address: Vietnam National University Ho Chi Minh City, Linh Trung Ward, Thu Duc Dist., Ho Chi Minh City, Vietnam

H. Gamo: ECSJ Student Member

K. Hikima and A. Matsuda: ECSJ Active Members

References

• 1)  J. Jiang, Q. Fan, S. Chou, Z. Guo, K. Konstantinov, H. Liu, and J. Wang, Small, 17, 1903934 (2021).
• 2)  M. R. Kaiser, Z. Han, J. Liang, S. X. Dou, and J. Wang, Energy Storage Materials, 19, 1 (2019).
• 3)  H. Nagata and J. Akimoto, ACS Appl. Mater. Interfaces, 13, 35785 (2021).
• 4)  N. H. H. Phuc, M. Takaki, H. Kazuhiro, M. Hiroyuki, and M. Atsunori, SN Appl. Sci., 2, 1803 (2020).
• 5)  T. Takeuchi, H. Kageyama, K. Nakanishi, M. Ogawa, T. Ohta, A. Sakuda, H. Sakaebe, H. Kobayashi, and Z. Ogumi, J. Electrochem. Soc., 162, A1745 (2015).
• 6)  N. H. H. Phuc, M. Takaki, M. Hiroyuki, and M. Atsunori, Front. Energy Res., 8, 606023 (2021).
• 7)  H. Gamo, T. Maeda, K. Hikima, M. Deguchi, Y. Fujita, Y. Kawasaki, A. Sakuda, H. Muto, N. H. H. Phuc, A. Hayashi, M. Tatsumisago, and A. Matsuda, Mater. Adv., 3, 2488 (2022).
• 8)  K. Takada, Acta Mater., 61, 759 (2013).
• 9)  R. Mercier, J.-P. Malugani, B. Fahys, and G. Robert, Solid State Ionics, 5, 663 (1981).
• 10)  N. Kamaya, K. Homma, Y. Yamakawa, M. Hirayama, R. Kanno, M. Yonemura, T. Kamiyama, Y. Kato, S. Hama, K. Kawamoto, and A. Mitsui, Nat. Mater., 10, 682 (2011).
• 11)  Y. Seino, T. Ota, K. Takada, A. Hayashi, and M. Tatsumisago, Energy Environ. Sci., 7, 627 (2014).
• 12)  P. Wang, H. Liu, S. Patel, X. Feng, P. H. Chien, Y. Wang, and Y. Y. Hu, Chem. Mater., 32, 3833 (2020).
• 13)  P. Adeli, J. D. Bazak, A. Huq, G. R. Goward, and L. F. Nazar, Chem. Mater., 33, 146 (2021).
• 14)  Z. Zhang, J. Zhang, H. Jia, L. Peng, T. An, and J. Xie, J. Power Sources, 450, 227601 (2020).
• 15)  C. K. Moon, H.-J. Lee, H. Kern, H. Park, J. W. Kwak, K. Heo, H. Choi, M.-S. Yang, S.-T. Kim, H. Hong, Y. Lee, and J. Seok, ACS Energy Lett., 3, 2504 (2018).
• 16)  K. Hikima, N. H. Huy Phuc, H. Tsukasaki, S. Mori, H. Muto, and A. Matsuda, RSC Adv., 10, 22304 (2020).
• 17)  N. H. H. Phuc, H. Kazuhiro, M. Hiroyuki, and M. Atsunori, Solid State Ionics, 351, 115324 (2020).
• 18)  F. Izumi and K. Momma, Solid State Phenom., 130, 15 (2007).
• 19)  A. K. Jonscher, Nature, 267, 673 (1977).
• 20)  S. Lorger, R. E. Usiskin, and J. Maier, Adv. Funct. Mater., 29, 1807688 (2019).
• 21)  F. Preishuber-Pflügl, P. Bottke, V. Pregartner, B. Bitschnau, and M. Wilkening, Phys. Chem. Chem. Phys., 16, 9580 (2014).
• 22)  K. Funke, Prog. Solid State Chem., 22, 111 (1993).
• 23)  K. S. Gilroy and W. A. Phillips, Philos. Mag. B, 43, 735 (1981).
• 24)  S. R. Elliott, Solid State Ionics, 70–71, 27 (1994).
• 25)  Y. Wang, W. D. Richards, S. P. Ong, L. J. Miara, J. C. Kim, Y. Mo, and G. Ceder, Nat. Mater., 14, 1026 (2015).
• 26)  D. L. Sidebottom, Phys. Rev. Lett., 83, 983 (1999).
• 27)  N. Y. Jin-Phillipp, N. Sata, J. Maier, C. Scheu, K. Hahn, M. Kelsch, and M. Rühle, J. Chem. Phys., 120, 2375 (2004).
• 28)  B. Ruprecht, M. Wilkening, S. Steuernagel, and P. Heitjans, J. Mater. Chem., 18, 5412 (2008).
• 29)  H. Tsukasaki, S. Mori, H. Morimoto, A. Hayashi, and M. Tatsumisago, Sci. Rep., 7, 4142 (2017).
• 30)  J. Maier, Solid State Ionics, 157, 327 (2003).
• 31)  M. Gombotz and H. M. R. Wilkening, ACS Sustain. Chem. Eng., 9, 743 (2021).
• 32)  D. Prutsch, B. Gadermaier, H. Brandstätter, V. Pregartner, B. Stanje, D. Wohlmuth, V. Epp, D. Rettenwander, I. Hanzu, and H. M. R. Wilkening, Chem. Mater., 30, 7575 (2018).
• 33)  I. Hanghofer, M. Brinek, S. L. Eisbacher, B. Bitschnau, M. Volck, V. Hennige, I. Hanzu, D. Rettenwander, and H. M. R. Wilkening, Phys. Chem. Chem. Phys., 21, 8489 (2019).
• 34)  Y. S. Choi and J. C. Lee, J. Power Sources, 415, 189 (2019).

Appendix A. Supplementary Data

The Williamson–Hall plots for Li₂S and CaS phase in (100 − x)Li₂S·xCaS (Fig. S1), SEM-EDS images of in (100 − x)Li₂S·xCaS (Fig. S2), densities of the pellet (Table S1), conductivity isotherms of Li₂S (Fig. S3), the imaginary part of modulus M′′ versus the frequency of Li₂S, 80Li₂S·20CaS, 70Li₂S·30CaS, and 50Li₂S·50CaS (Fig. S4), a magnified view of the XRD patterns for (100 − x)Li₂S·xCaF₂, (100 − x)Li₂S·xCaCl₂, (100 − x)Li₂S·xCaBr₂, and (100 − x)Li₂S·xCaI₂ (Fig. S5), an Arrhenius plot of the DC conductivities for 80Li₂S·20LiI (Fig. S6).