Prediction of Ground State Configurations and Electrochemical Properties\\ of Li$_{3}$InCl$_{6}$ Doped with F, Br, and Ga

Zheng-Yu Lu(鲁征宇)$^{1}$, Le-Tian Chen(陈乐添)$^{1}$, Xu Hu(胡绪)$^{1}$, Su-Ya Chen(陈素雅)$^{1}$,\\ Xu Zhang(张旭)$^{2\hyperlink{s}{*}}$, and Zhen Zhou(周震)$^{1,2}$

doi:10.1088/0256-307X/41/5/058201

⇦Chinese Physics Letters, 2024, Vol. 41, No. 5, Article code 058201⇨ Prediction of Ground State Configurations and Electrochemical Properties of Li$_{3}$InCl$_{6}$ Doped with F, Br, and Ga Zheng-Yu Lu (鲁征宇)1, Le-Tian Chen (陈乐添)1, Xu Hu (胡绪)1, Su-Ya Chen (陈素雅)1, Xu Zhang (张旭)2*, and Zhen Zhou (周震)1,2 Affiliations 1Department of Materials Science and Engineering, Nankai University, Tianjin 300350, China 2Interdisciplinary Research Center for Sustainable Energy Science and Engineering (IRC4SE2), School of Chemical Engineering, Zhengzhou University, Zhengzhou 450001, China Received 24 January 2024; accepted manuscript online 23 April 2024; published online 23 May 2024 ^*Corresponding author. Email: zzuzhangxu@zzu.edu.cn Citation Text: Lu Z Y, Chen L T, Hu X et al. 2024 Chin. Phys. Lett. 41 058201 Abstract Compared with conventional solid-state electrolytes, halide solid-state electrolytes have several advantages such as a wider electrochemical window, better compatibility with oxide cathode materials, improved air stability, and easier preparation conditions making them conductive to industrial production. We concentrate on a typical halide solid-state electrolyte, Li$_{3}$InCl$_{6}$, predict the most stable structure after doping with Br, F, and Ga by using the Alloy Theoretic Automated Toolkit based on first-principles calculations, and verify the accuracy of the prediction model. To investigate the potential of three equivalently doped ground state configurations of Li$_{3}$InCl$_{6}$ as solid-state electrolytes for all-solid-state lithium-ion batteries, their specific properties such as crystal structure, band gap, convex packing energy, electrochemical stability window, and lithium-ion conductivity are computationally analyzed using first-principles calculations. After a comprehensive evaluation, it is determined that the F-doped ground state configuration Li$_{3}$InCl$_{2.5}$F$_{3.5}$ exhibits better thermal stability, wider electrochemical stability window, and better lithium ion conductivity (1.80 mS$\cdot$cm$^{-1}$ at room temperature). Therefore, Li$_{3}$InCl$_{2.5}$F$_{3.5}$ has the potential to be used in the field of all-solid-state lithium-ion batteries as a new type of halide electrolyte.

DOI:10.1088/0256-307X/41/5/058201 © 2024 Chinese Physics Society Article Text After decades of development, lithium-ion batteries consisting of graphite as the anode, transition metal oxides as the cathode, and liquid organic electrolyte have reached the upper limit of energy density (about 350 Wh$\cdot$kg$^{-1})$.[1] However, some issues related to the cycle life and safety of traditional lithium-ion batteries make it difficult to meet the increasing demand for future applications. The safety threat associated with the leakage and flammability of liquid organic electrolytes cannot be ignored. The improved safety, wider electrochemical window, and longer cycle life of solid-state electrolytes (SSEs) make all-solid-state lithium batteries (ASSLB) the most promising contenders for the next-generation batteries.[2,3] SSEs mainly include several types of polymers, oxides, sulfides, and halides. The ionic conductivity of polymer SSEs is greatly affected by temperature, which makes them difficult to work stably in a wide temperature range.[4] Oxide SSEs exhibit good electrochemical stability, however, their poor mechanical properties due to the rigid structure lead to insufficient contact between the electrode and electrolyte, resulting in a large interfacial impedance.[5] Sulfide SSEs possess high ionic conductivity, whereas they are less stable in the presence of air and humidity.[6] Halide SSEs offer several advantages, including a wider electrochemical window, better compatibility with oxide cathode materials, greater stability in the presence of water and air, and easier preparation conditions conducive to industrial production. The halide SSEs can be represented in the form of Li–M–X, where M is a metal element and X is a halogen. The categorization of halide SSEs is based on the classification of the metal element M, which can be grouped as follows: (1) M belonging to group IIIA, (2) M belonging to group IIIB, and (3) M being a divalent metallic element such as Ti, V, Fe, Co, and Ni. Asano et al..[7] reported novel halide SSE materials, Li$_{3}$YBr$_{6}$ and Li$_{3}$YCl$_{6}$, which exhibit high ionic conductivity while also possessing satisfactory electrochemical and thermal stability as well as mechanical properties suitable for large-scale fabrication. Doping with different elements has shown promising results in the field of discovering novel SSE materials. On the one hand, doping can improve the ionic conductivity of SSE creating vacancy defects and broadening ion transport channels. On the other hand, doping can regulate the composition of SSE, reduce interfacial resistance, and improve chemical stability, thus effectively solving the SEI problem. For example, doping with Mo,[8] Fe,[9,10] Zn,[11] and Mn[12] can effectively improve the ionic conductivity of Li$_{7}$P$_{3}$S$_{11}$ and form a metallic phase at the interface to reduce the interfacial resistance; doping with O,[11,13] Cl,[14,15] Br,[11,16] and I[17,18] can also improve the ionic conductivity and chemical stability of SSEs. However, the wide range of available doping concentrations means that researchers are constantly searching for efficient ways to determine optimal doping concentrations. While the traditional trial-and-error experimental method obviously requires a significant amount of time and may lead to inaccurate data, computational simulations based on density functional theory (DFT) can make a big difference. Softwares for ground state configuration search have been successfully developed, such as Alloy Theoretic Automated Toolkit (ATAT),[19-23] which uses the total energy calculated by the Vienna ab initio simulation package (VASP),[24] a program to determine the most stable phase for each component. The ATAT program employs the cluster expansion (CE)[25-28] method through the embedded MIT ab initio phase stability (MAPS)[23] program to generate various candidate structures with doping concentrations ranging from 0 to 100%, and the energy of each cluster and structure is then evaluated and compared by linked first-principles calculations. The CE method has a significant advantage in terms of rapid convergence, enabling the use of relatively compact clusters to determine the energies of crystalline phases through the structure inversion method while still achieving sufficient accuracy.[29] Li et al.[30] successfully synthesized a halide SSE material, Li$_{3}$InCl$_{6}$, exhibiting high ionic conductivity (up to $1.49 \times 10^{-3}$ S$\cdot$cm$^{-1}$ at 298 K), as well as stability to air and humidity. Since Br and F are congeners of Cl with the same valences, and Ga for the same reason, we have chosen F, Br, and Ga as doping elements. In this work, we predict the ground state configurations of Li$_{3}$InCl$_{6}$ after equivalent doping of F, Br, and Ga using ATAT, and evaluate the accuracy of the adopted CE model. The potential of the predicted ground state configuration to be applied to all-solid-state lithium-ion batteries is also explored through DFT calculations and modules in the Python materials genomics (Pymatgen)[31] program. Computational Methods. For DFT calculations, we use VASP with the projector augmented wave[32] method to describe the interaction between ions and electrons. The Perdew–Burke–Ernzerhof[33] functions are employed for exchange correlation (XC). In order to balance the computational accuracy and efficiency, we use the default value of the cutoff energy during the ground state structure search process. Moreover, for further structure optimization and property calculation, we apply a cutoff energy of 450 eV. The $K$-points in the Brillouin zone are selected using the Monkhorst–Pack method.[34] To ensure the uniform accuracy of $K$-points throughout the search process, we set KPPRA ($K$-point per reciprocal atom) to 1000 in the input file vasp.wrap of ATAT, allowing the package to automatically generate a suitable $K$-point grid. During the process of searching the structure, along with the expansion of the cell, the precision of the $K$-point is sufficient. For the further calculation, we adopt the $K$-point density recommended by VASPKIT.[35] The pseudopotential recommended by VASP is chosen for the calculations. To improve the accuracy of the obtained energy, we set DOSTATIC to perform the static calculation after completing the structural relaxation. Through the materials project (MP) REST (representational state transfer) API (application programming interface),[36] we use the Pymatgen program to access the MP database and perform phase diagrams, the energy above hull ($E_{\rm hull})$ and electrochemical stability window calculations. We construct phase diagrams based on the Gibbs free energy of all components within the ground state structure. The Gibbs free energy of a substance is defined as \begin{equation} G=H-TS=E+PV-TS, \tag {1} \end{equation} where $H$ and $S$ represent enthalpy and entropy, respectively. To approximate $G$ for the condensed phase at 0 K, we assume $G$ to be equal to $E$. The total $E_{\rm hull}$ is defined as the difference between the linear combination value of the internal energies of a compound in the phase diagram and the surrounding steady-state components. A higher $E_{\rm hull}$ value reflects lower thermal stability of the structure. We construct the phase diagram based on the grand potential of all components within the ground state structure. The grand potential of lithium is defined as \begin{equation} \phi =G-\mu_{\scriptscriptstyle{\rm Li}} N_{\scriptscriptstyle{\rm Li}} =E+PV-TS-\mu_{\scriptscriptstyle{\rm Li}} N_{\scriptscriptstyle{\rm Li}}, \tag {2} \end{equation} where $N_{\scriptscriptstyle{\rm Li}}$ means the number of lithium atoms and $\mu_{\scriptscriptstyle{\rm Li}}$ represents the chemical potential of lithium in the environment. In this work, $\mu_{\scriptscriptstyle{\rm Li}}$ is defined as \begin{equation} \mu_{\scriptscriptstyle{\rm Li}} =\mu_{\scriptscriptstyle{\rm Li}}^{0} -e\varphi, \tag {3} \end{equation} as a function of voltage. The grand potential phase diagram of lithium enables us to determine the thermal stability of the doped ground state configuration at different voltages, as well as the electrochemical stability window of the structure. The construction of the phase diagram can be realized through the phase diagram module[37] of Pymatgen. The conductivity of lithium ions is a key property that determines whether a material can be used as an SSE. To calculate the conductivity of lithium ions for the three doped ground-state configurations, we use the ab initio molecular dynamics (AIMD) method, performed by using the Nosé–Hoover thermostatic conditions at 600 K, 900 K, 1200 K, and 1500 K with time steps of 2 fs for a total simulation time of 60 ps. First, we calculate the root-mean-square displacement (RMSD) of the lithium ion as \begin{align} {\rm RMSD}&=\langle{[{r_{m}({t+t_{0}})-r_{m}({t_{0}})}]^{2}}\rangle\notag\\ &=\frac{1}{N}\sum\nolimits_m{[{r_{m}({t+t_{0}})-r_{m}({t_{0}})}]}^{2}. \tag {4} \end{align} The calculation of RMSD is performed through the diffusion analyzer module of Pymatgen. The self-diffusion coefficient of lithium ions is calculated as \begin{equation} D=\lim\limits_{t\to \infty} \frac{\langle {r^{2}(t)}\rangle}{2dt}. \tag {5} \end{equation} Then we could obtain the conductivity of lithium ions by the Nernst–Einstein equation as \begin{equation} \sigma =\frac{Nq^{2}D}{k_{\rm B} T}. \tag {6} \end{equation} According to the Arrhenius equation, we could express the conductivity as a function of temperature, using several constant parameters. Finally we can reach \begin{equation} \ln ({\sigma T})=\frac{K}{T}+A, \tag {7} \end{equation} which indicates that $\ln(\sigma T)$ is proportional to the inverse of $T$. Due to the low conductivity at room temperature, the direct calculation of RMSD is not accurate. Employing the above equation, we could use the conductivity at high temperature to obtain the conductivity and the activation energy of lithium ion transport at room temperature.

Fig. 1. Schematic diagram of the crystal structure of Li$_{3}$InCl$_{6}$ (a) ball-stick (b) polyhedral. The blue, purple, and green balls represent Cl, In, and Li, respectively.

Prediction of Ground State Configurations. Figure 1 shows the crystal structure of Li$_{3}$InCl$_{6}$ with space group $c2/m$. The crystal lattice of Li$_{3}$InCl$_{6}$ is composed of Cl$^{-}$ ions arranged in six-coordinated octahedral sites, with In$^{3+}$ and Li$^{+}$ occupying these sites and sharing common edges. After structural relaxation calculations, the lattice constants of Li$_{3}$InCl$_{6}$ obtained agree well with the experimental data ($a= b=6.364$ Å, $c=6.368$ Å, $\alpha=\beta=100.197^{\circ}$, $\gamma=119.630^{\circ}$).[30] We successfully predict the ground state configurations of Li$_{3}$InCl$_{6}$ doped with F, Br, and Ga using ATAT. Figure 2 illustrates the relationship between the energy of the doped structure and the doping concentration. To verify the accuracy of the CE model, we compare the energy values obtained from first-principles calculations to those predicted by the CE model. As shown in Fig. 3, there is not much difference between the predicted and calculated energy values for the three doping strategies. To better visualize the relationship between predicted and calculated energy values for various configurations, a linear fitting has been performed, as shown in Fig. 4. We verify the accuracy of the CE model for all three doping strategies and find that it falls within the ideal range. Therefore, the ground state configurations predicted by ATAT are accurate and can be used for subsequent property studies.

Fig. 2. Relationship between configuration energy and doping concentration predicted by ATAT (a) for dopant F, (b) for dopant Br, and (c) for dopant Ga.

Fig. 3. Difference between the energies calculated by DFT and predicted by the CE model (a) for dopant F, (b) for dopant Br, and (c) for dopant Ga.

Fig. 4. Linear fitting of the energies calculated by DFT and predicted by CE model (a) for dopant F, (b) for dopant Br, and (c) for dopant Ga.

Electrochemical Properties of Ground State Configurations. The ground state configurations of Li$_{3}$InCl$_{6}$ doped with F, Br, and Ga are obtained by prediction as Li$_{3}$InCl$_{2.5}$F$_{3.5}$, Li$_{3}$InCl$_{0.33}$Br$_{5.67}$, and Li$_{3}$In$_{0.83}$Ga$_{0.17}$Cl$_{6}$, respectively. Their structures are shown in Fig. 5, which have been expanded for comparison. Comparison with the original structure of Li$_{3}$InCl$_{6}$ reveals that the lattice of all the three doped ground-state configurations is distorted and the symmetry is weakened, however the cations are still located on the six-coordinated co-sided octahedral sites composed of anions, indicating that the overall transport mechanisms and pathways of lithium ions have not been significantly altered. To investigate their potential applications in ASSLB, we calculate the electronic structures of both the original structure of Li$_{3}$InCl$_{6}$ and the three doped ground state configurations. Figure 6 shows the diagrams of band structure and density of states for every configuration. The band gaps of the four configurations are obtained by analyzing the energy band structure, as listed in Table 1. It can be found that doping with Br significantly reduces the band gap, making it unsuitable for use as an SSE in ASSLB. Conversely, doping with F and Ga does not significantly alter the band gap and improves the insulation performance, suggesting their potential suitability for use in ASSLB.

Fig. 5. Ground state structure of Li$_{3}$InCl$_{6}$ doped: (a) Li$_{3}$InCl$_{0.33}$Br$_{5.67}$, (b) Li$_{3}$InCl$_{2.5}$F$_{3.5}$, (c) Li$_{3}$In$_{0.83}$Ga$_{0.17}$Cl$_{6}$. The blue, purple, green, brown, yellow, and red balls represent Cl, In, Li, Br, F, and Ga, respectively.

Fig. 6. Diagrams of band structure and density of states of Li$_{3}$InCl$_{6}$ pristine structure and doped structure: (a) Li$_{3}$InCl$_{6}$, (b) Li$_{3}$InCl$_{0.33}$Br$_{5.67}$, (c) Li$_{3}$InCl$_{2.5}$F$_{3.5}$, (d) Li$_{3}$In$_{0.83}$Ga$_{0.17}$Cl$_{6}$.

Table 1. Band gap and $E_{\rm hull}$ of Li$_{3}$InCl$_{6}$ pristine structure and doped structure.
	VBM (eV)	CBM (eV)	Band gap (eV)	$E_{\rm hull}$ (eV/atom)
Li$_{3}$InCl$_{6}$	0.48	3.84	3.36	0.055
Li$_{3}$InCl$_{0.33}$Br$_{5.67}$	0.12	2.46	2.34	0.055
Li$_{3}$InCl$_{2.5}$F$_{3.5}$	$-0.04$	3.46	3.50	0.053
Li$_{3}$In$_{0.83}$Ga$_{0.17}$Cl$_{6}$	0.48	3.79	3.31	0.061

The $E_{\rm hull}$ values of the original structure of Li$_{3}$InCl$_{6}$ and the three doped ground state structures are calculated by the PhaseDiagram module of Pymatgen, as listed in Table 1. It can be seen that the stability of the structure doped with Br and F does not differ significantly from that of the original structure, however the thermal stability becomes worse after doping with Ga. Then, we calculate lithium grand potential phase diagrams of the above four structures by Pymatgen to explore the thermal stability of the doped ground state configurations at different voltages, and the resulting electrochemical stability windows of the materials are obtained as shown in Fig. 7 and Table 2. It can be seen that the width of the electrochemical stability window is reduced for all the three doped structures. However, it is significantly wider than many current sulfide and oxide SEs such as LGPS (1.72–2.29 V) and Li$_{3}$PS$_{4}$ (1.71–2.31 V).[38,39]

Fig. 7. Electrochemical stability windows of the four configurations: (a) Li$_{3}$InCl$_{6}$, (b) Li$_{3}$InCl$_{0.33}$Br$_{5.67}$, (c) Li$_{3}$InCl$_{2.5}$F$_{3.5}$, (d) Li$_{3}$In$_{0.83}$Ga$_{0.17}$Cl$_{6}$.

Finally, to investigate the lithium ion conductivity of the three doped configurations, we utilize the diffusion analyzer module of Pymatgen. We perform AIMD calculations at 600 K, 900 K, 1200 K, and 1500 K, and calculate the ionic conductivity at the corresponding temperatures from the obtained RMSD data. To obtain the ionic conductivity at room temperature, we employ the Arrhenius equation to plot $\ln(\sigma T)$ against $1000/T$ and fit a straight line for extrapolation, as shown in Fig. 8 and Table 3.

Table 2. Electrochemical stability window ranges and widths of the four configurations.
	Electrochemical stability range (V)	Electrochemical stability width (V)
Li$_{3}$InCl$_{6}$	2.28–4.42	2.14
Li$_{3}$InCl$_{0.33}$Br$_{5.67}$	2.17–3.14	0.97
Li$_{3}$InCl$_{2.5}$F$_{3.5}$	2.94–4.43	1.49
Li$_{3}$In$_{0.83}$Ga$_{0.17}$Cl$_{6}$	2.32–4.25	1.93

Fig. 8. Lithium ion conductivity vs. temperature for three doped configurations (black, red, and blue represent doped Br, F, and Ga, respectively).

Table 3. Lithium ion conductivity of the three doped configurations at different temperatures.
	600 K (mS$\cdot$cm$^{-1})$	900 K (mS$\cdot$cm$^{-1})$	1200 K (mS$\cdot$cm$^{-1})$	1500 K (mS$\cdot$cm$^{-1})$	300 K (mS$\cdot$cm$^{-1})$
Li$_{3}$InCl$_{0.33}$Br$_{5.67}$	–	297.63	549.00	720.19	0.79
Li$_{3}$InCl$_{2.5}$F$_{3.5}$	237.91	1153.56	1351.92	3140.43	2.01
Li$_{3}$In$_{0.83}$Ga$_{0.17}$Cl$_{6}$	127.10	693.06	1063.61	2120.74	0.64

The ionic conductivities of Li$_{3}$InCl$_{6}$ synthesized by Li et al.[30] using mechanical and annealing methods are 0.84 and 1.49 mS$\cdot$cm$^{-1}$, respectively, which show that Li$_{3}$InCl$_{2.5}$F$_{3.5}$ obtained by F-doping exhibits good potential as a new halide SSE using in ASSLB. Interestingly, a report has demonstrated that F-doped Li$_{3}$InCl$_{6}$ (Li$_{3}$InCl$_{5.8}$F$_{0.2})$ solid electrolyte shows significant potential in ASSLB.[40] In summary, we have utilized DFT calculations and software programs, such as ATAT and Pymatgen, to investigate potential applications of the ground state configuration of the ternary metal halide electrolyte Li$_{3}$InCl$_{6}$, after equivalently doping with Br, F, and Ga, as an SSE in ASSLBs. Our calculations successfully predict the ground state configuration of the doped structures, and the accuracy of the predicted results is scientifically evaluated using the adopted CE model, demonstrating good accuracy. Three ground state configurations after doping are obtained, which are identified as Li$_{3}$InCl$_{0.33}$Br$_{5.67}$, Li$_{3}$InCl$_{2.5}$F$_{3.5}$, and Li$_{3}$In$_{0.83}$Ga$_{0.17}$Cl$_{6}$. Comprehensive analyses of the crystal structure, band gap, energy above hull, electrochemical stability window, and lithium ion conductivity are performed to evaluate their capability as SSEs for ASSLB. The calculation results show that Li$_{3}$InCl$_{2.5}$F$_{3.5}$ possesses superior thermal stability, wider electrochemical stability window, and better lithium ion conductivity (up to 2.01 mS$\cdot$cm$^{-1}$ at room temperature). Our findings not only provide a prediction of a novel halide SSE for use in ASSLBs, but also offer a systematic approach to designs of SSE materials and a novel research model that could contribute to the advancement of the field. Acknowledgment. This work was supported by the National Key Research and Development Program of China (Grant No. 2021YFF0500600). References Erratum: Li–O2 and Li–S batteries with high energy storageInorganic Solid-State Electrolytes for Lithium Batteries: Mechanisms and Properties Governing Ion ConductionFundamentals of inorganic solid-state electrolytes for batteriesNanocomposite polymer electrolytes for lithium batteriesFast Lithium Ion Conduction in Garnet‐Type Li₇ La₃ Zr₂ O₁₂A lithium superionic conductorSolid Halide Electrolytes with High Lithium‐Ion Conductivity for Application in 4 V Class Bulk‐Type All‐Solid‐State BatteriesTailored Li₂ S–P₂ S₅ glass-ceramic electrolyte by MoS₂ doping, possessing high ionic conductivity for all-solid-state lithium-sulfur batteriesCathode-doped sulfide electrolyte strategy for boosting all-solid-state lithium batteriesSulfur-Embedded FeS₂ as a High-Performance Cathode for Room Temperature All-Solid-State Lithium–Sulfur BatteriesArgyrodite Solid Electrolyte with a Stable Interface and Superior Dendrite Suppression Capability Realized by ZnO Co-DopingAll-solid-state lithium–sulfur batteries based on a newly designed Li₇ P_2.9 Mn_0.1 S_10.7 I_0.3 superionic conductorLi2S–Li2O–P2S5 solid electrolyte for all-solid-state lithium batteriesLi₃ Y(PS₄ )₂ and Li₅ PS₄ Cl₂ : New Lithium Superionic Conductors Predicted from Silver Thiophosphates using Efficiently Tiered Ab Initio Molecular Dynamics SimulationsConductivity of 70Li2S·30P2S5 glasses and glass–ceramics added with lithium halidesNovel Research Approach Combined with Dielectric Spectrum Testing for Dual-Doped Li₇ P₃ S₁₁ Glass-Ceramic ElectrolytesLiI-Doped Sulfide Solid Electrolyte: Enabling a High-Capacity Slurry-Cast Electrode by Low-Temperature Post-Sintering for Practical All-Solid-State Lithium BatteriesAn Iodide-Based Li₇ P₂ S₈ I Superionic ConductorAlgorithm for generating derivative structuresThe alloy theoretic automated toolkit: A user guideAutomating first-principles phase diagram calculationsTheoretical formulation of Li_3a+b N_a X_b (X = halogen) as a potential artificial solid electrolyte interphase (ASEI) to protect the Li anodeFirst-principles formulation of spinel-like structured Li_(4−3x) Y_x Cl₄ as promising solid-state electrolytes to enable superb lithium ion conductivity and matching oxidation potentials to high-voltage cathodesEfficient iterative schemes for ab initio total-energy calculations using a plane-wave basis setSolid State PhysicsSpringer Series in Solid-State SciencesGeneralized cluster description of multicomponent systemsNATO ASI SeriesDensity-functional theory applied to phase transformations in transition-metal alloysAir-stable Li₃ InCl₆ electrolyte with high voltage compatibility for all-solid-state batteriesPython Materials Genomics (pymatgen): A robust, open-source python library for materials analysisFrom ultrasoft pseudopotentials to the projector augmented-wave methodGeneralized Gradient Approximation Made SimpleSpecial points for Brillouin-zone integrationsVASPKIT: A user-friendly interface facilitating high-throughput computing and analysis using VASP codeThe Materials Application Programming Interface (API): A simple, flexible and efficient API for materials data based on REpresentational State Transfer (REST) principlesLi−Fe−P−O₂ Phase Diagram from First Principles CalculationsInterface Stability in Solid-State BatteriesOrigin of Outstanding Stability in the Lithium Solid Electrolyte Materials: Insights from Thermodynamic Analyses Based on First-Principles CalculationsFluorine-doped Li3InCl6 to enhance ionic conductivity and air stability

[1]	Bruce P G, Freunberger S A, Hardwick L J, and Tarascon J M 2012 Nat. Mater. 11 172
[2]	Bachman J C, Muy S, Grimaud A, Chang H H, Pour N, Lux S F, Paschos O, Maglia F, Lupart S, Lamp P, Giordano L, and Shao-Horn Y 2016 Chem. Rev. 116 140
[3]	Famprikis T, Canepa P, Dawson J A, Islam M S, and Masquelier C 2019 Nat. Mater. 18 1278
[4]	Croce F, Appetecchi G B, Persi L, and Scrosati B 1998 Nature 394 456
[5]	Murugan R, Thangadurai V, and Weppner W 2007 Angew. Chem. Int. Ed. 46 7778
[6]	Kamaya N, Homma K, Yamakawa Y, Hirayama M, Kanno R, Yonemura M, Kamiyama T, Kato Y, Hama S, Kawamoto K, and Mitsui A 2011 Nat. Mater. 10 682
[7]	Asano T, Sakai A, Ouchi S, Sakaida M, Miyazaki A, and Hasegawa S 2018 Adv. Mater. 30 1803075
[8]	Xu R C, Xia X H, Wang X L, Xia Y, and Tu J P 2017 J. Mater. Chem. A 5 2829
[9]	Zhou L, Tufail M K, Yang L, Ahmad N, Chen R J, and Yang W 2020 Chem. Eng. J. 391 123529
[10]	Mwizerwa J P, Zhang Q, Han F, Wan H, Cai L, Wang C, and Yao X 2020 ACS Appl. Mater. & Interfaces 12 18519
[11]	Chen T, Zhang L, Zhang Z X, Li P, Wang H Q, Yu C, Yan X L, Wang L M, and Xu B 2019 ACS Appl. Mater. & Interfaces 11 40808
[12]	Xu R C, Xia X H, Li S H, Zhang S Z, Wang X L, and Tu J P 2017 J. Mater. Chem. A 5 6310
[13]	Trevey J E, Gilsdorf J R, Miller S W, and Lee S H 2012 Solid State Ionics 214 25
[14]	Zhu Z Y, Chu I H, and Ong S P 2017 Chem. Mater. 29 2474
[15]	Ujiie S, Inagaki T, Hayashi A, and Tatsumisago M 2014 Solid State Ionics 263 57
[16]	Zhang N, Ding F, Yu S, Zhu K, Li H, Zhang W, Liu X, and Xu Q 2019 ACS Appl. Mater. & Interfaces 11 27897
[17]	Choi S J, Choi S H, Bui A D, Lee Y J, Lee S M, Shin H C, and Ha Y C 2018 ACS Appl. Mater. & Interfaces 10 31404
[18]	Rangasamy E, Liu Z, Gobet M, Pilar K, Sahu G, Zhou W, Wu H, Greenbaum S, and Liang C 2015 J. Am. Chem. Soc. 137 1384
[19]	Hart G L W and Forcade R W 2008 Phys. Rev. B 77 224115
[20]	van de Walle A, Asta M, and Ceder G 2002 Calphad 26 539
[21]	Walle A and Ceder G 2002 J. Phase Equilib. 23 348
[22]	Sang J W, Yu Y R, Wang Z, and Shao G S 2020 Phys. Chem. Chem. Phys. 22 12918
[23]	Huang Y Y, Yu Y R, Xu H J, Zhang X D, Wang Z, and Shao G S 2021 J. Mater. Chem. A 9 14969
[24]	Kresse G and Furthmüller J 1996 Phys. Rev. B 54 11169
[25]	Fontaine D D 1994 Solid State Phys. 47 33
[26]	Ducastelle F 1993 Order and Phase Stability in Alloys. In: Terakura K and Akai H (eds) Interatomic Potential and Structural Stability. Springer Series in Solid-State Sciences (Berlin: Springer) vol 114 p 133
[27]	Sanchez J M, Ducastelle F, and Gratias D 1984 Physica A 128 334
[28]	Zunger A 1994 First-Principles Statistical Mechanics of Semiconductor Alloys and Intermetallic Compounds. In: Turchi P E A and Gonis A (eds) Statics and Dynamics of Alloy Phase Transformations. NATO ASI Series (Boston: Springer) vol 319 p 361
[29]	Connolly J W D and Williams A R 1983 Phys. Rev. B 27 5169
[30]	Li X N, Liang J W, Luo J, Norouzi Banis M, Wang C H, Li W H, Deng S X, Yu C, Zhao F P, Hu Y F, Sham T K, Zhang L, Zhao S Q, Lu S G, Huang H, Li R Y, Adair K R, and Sun X L 2019 Energy & Environ. Sci. 12 2665
[31]	Ong S P, Richards W D, Jain A, Hautier G, Kocher M, Cholia S, Gunter D, Chevrier V L, Persson K A, and Ceder G 2013 Comput. Mater. Sci. 68 314
[32]	Kresse G and Joubert D 1999 Phys. Rev. B 59 1758
[33]	Perdew J P, Burke K, and Ernzerhof M 1996 Phys. Rev. Lett. 77 3865
[34]	Monkhorst H J and Pack J D 1976 Phys. Rev. B 13 5188
[35]	Wang V, Xu N, Liu J C, Tang G, and Geng W T 2021 Comput. Phys. Commun. 267 108033
[36]	Ong S P, Cholia S, Jain A, Brafman M, Gunter D, Ceder G, and Persson K A 2015 Comput. Mater. Sci. 97 209
[37]	Ong S P, Wang L, Kang B, and Ceder G 2008 Chem. Mater. 20 1798
[38]	Richards W D, Miara L J, Wang Y, Kim J C, and Ceder G 2016 Chem. Mater. 28 266
[39]	Zhu Y Z, He X F, and Mo Y F 2015 ACS Appl. Mater. & Interfaces 7 23685
[40]	Wang Q T, Ma X F, Liu Q, Sun D F, and Zhou X Z 2023 J. Alloys Compd. 969 172479