Large scale stationary energy storage, as well as the development of hybrid cars for the ground and the air, need for the use of sophisticated secondary batteries with high energy density, high specific energy, extended cycle lives, high safety standards, and low cost. However, due to their intrinsic limits, currently available secondary batteries such as intercalation-type Li ion batteries (LIBs), lead acid batteries, and nickel metal hydride batteries are unable to meet all these criteria at the same time. (Bruce et al., 2011; Yang et al., 2011; Zhang, 2015; Seh et al., 2016; Yuan et al., 2016; Chen et al., 2018; Liu et al., 2018; Yuan et al., 2020). Li-sulfur (Li-S) batteries are considered as one of the most promising alternatives for large-scale energy storage systems due to their potential high specific energy (over 900 Wh kg⁻¹) by employing low-cost sulfur and high theoretical specific capacity Li as cathode/anode materials, respectively. (Kolosnitsyn and Karaseva, 2008; Yang et al., 2014; Ulissi et al., 2018; Zhang et al., 2019). However, several fundamental issues and industrial obstacles, such as the poor electronic conductivity of S, the shuttle effect of Li polysulfide, and the huge volume expansion of S during cycling limits its wide applications. One solution to address the above problems is to replace the liquid electrolytes with non-flammability inorganic solid electrolytes to constructure all-solid-state Li-S batteries with enhanced safety. Moreover, the new work mechanism in solid-state battery avoids the formation of Li polysulfide and the shuttle effect. Solid electrolytes exhibit more stable Li stripping and plating due to the larger Young’s modulus, enabling to utilize lithium metal anode to fabricate solid-state lithium metal batteries with higher energy densities. (Wang et al., 2010; Son et al., 2014; Wang et al., 2014; Zhao et al., 2014; Gallardo et al., 2016; Zhu et al., 2016; Descostes et al., 2020).

Solid electrolytes, as the crucial component of solid-state battery, exert great impact on battery performance, sulfide solid-state electrolytes have advantageous properties for application in Li-S all-solid-state batteries, including high Li⁺ conductivity (10⁻³∼10⁻² S cm⁻¹) and low interfacial resistance between the electrode and electrolyte. (Pang et al., 2021; Wu et al., 2021; Liao et al., 2022a; Peng et al., 2022). Many sulfide electrolytes have been successfully developed for all-solid-state batteries. (Guo et al., 2022; Lin et al., 2022; Liu et al., 2022; Zeng et al., 2022). Compared with most of other SSEs, Li₇P₃S₁₁ achieved a much higher experimental ion conductivity of around 1.7 × 10^–2 S cm⁻¹, a theoretical conductivity of around 7.2 × 10^–2 S cm⁻¹, and lower activation energy of around 12 kJ mol⁻¹ at room temperature. (Hayashi et al., 2010; Wan et al., 2019). It also shows more favorable kinetic stability on the interface between electrodes and SSEs, making it one of the best SSEs for ALSSBs.

High-energy-density conversional-type electrode materials such as FeS₂, Li₂S, have recently drawn attention due to the progress of Li metal and solid electrolytes. FeS₂ is an abundant and inexpensive natural material with a high theoretical capacity of up to 894 mAh g⁻¹ based on the complete conversion from FeS₂ to Li₂S and Fe. Therefore, FeS₂ has been considered a promising high-capacity cathode material for lithium batteries. The shuttle effect of polysulfides, large volume variations throughout the cycling process, and irreversible loss of active ingredients are still the significant obstacles. (Cheng et al., 2016; Zhang and Tran, 2016). The primary problem is the enormous volume fluctuation that occurs throughout the Li⁺ plating/stripping process (159.2%), which results in worse contraction between cathode materials and electrolyte. (Wang et al., 2010; Wang et al., 2019). Therefore, overcoming the volume expansion is the basis for the FeS₂ cathode materials’ practical application.

In addition, temperature effect has a significant impact on electrochemical performance. Higher ionic conductivity can be achieved at high temperatures and lower temperatures result in a lower ionic conductivity of solid electrolytes. Moreover, the lithium-ion migration kinetics and side reaction of the electrode/electrolyte solid interface, are also influenced by the operating temperatures. (Peng et al., 2021). Moreover, as to the liquid-state lithium–sulfur batteries, the fluidity electrolytes can build ionic conducting connections between different parts. While the three-dimensional conducting framework in solid-state Li-S batteries is based on various types of solid-solid connections, leading to more severe volume variations of solid/solid contacts during cycling, especially under different operating temperatures. (Rodrigues et al., 2017; Choi et al., 2018). However, the evaluations of solid/solid contacts and electrochemical performances of Li₇P₃S₁₁-based solid-state Li-S batteries operated at different temperatures are unclear. Revealing those working mechanisms is vital to constructing high-performance all-climate solid-state Li-S batteries.

In this work, the Li₇P₃S₁₁ are employed as solid electrolytes to construct SSBs combined with pristine FeS₂ and Li–In anode. Systematical investigations have been performed to unravel the temperature influence of FeS₂ cathode on electrochemical performances of the assembled Li₇P₃S₁₁-based SSBs. In-situ stack pressure measurement, in situ EIS and DRT are used to investigate the stress variations and monitor the resistance evolution in solid-state batteries under various operation temperatures of solid-state batteries, the effect of temperature effect are unraveled, indicating that the phase transformation and the volume effect exert great impact to the battery performance, and highly affected by ambient temperature.

Li₇P₃S₁₁ electrolyte was prepared via mechanical milling method of the appropriate stoichiometry ratio of the Li₂S (Sigma Aldrich, 99.98%), P₂S₅ (Macklin, 99%), employing a planetary ball mill (Retsch, PM 200) for 15 h at 500 rpm within a ZrO₂-coated stainless steel jar. The prepared amorphous glass mixture was annealed at 270°C for 3 h to obtain the Li₇P₃S₁₁ electrolytes. The cathode mixture FeS₂@Li₇P₃S₁₁@C was obtained by blending FeS₂ with C (CNTs) and Li₇P₃S₁₁ with a weight ratio of 4:5:1. The mixing process was achieved by the ball milling method. The above synthesis process was carried out under an Ar atmosphere in the glove box to avoid air reacting with solid-state electrolytes.

XRD patterns of the solid-state electrolyte Li₇P₃S₁₁ were collected from a SmartLab-SE Powder instrument with a 2θ ranging from 10° to 80° using Cu Kα radiation. Morphology and EDS mappings of solid-state electrolytes were observed by SEM (Hitachi S-4800 II FESEM). Ionic conductivities were measured via pelletizing 100 mg of electrolytes into a pellet with a 10 mm diameter and using mold cells. An impedance analyzer (Solartron, 1,260) with an amplitude of 10 mV was utilized to obtain the impedance spectrum.

To fabricate FeS₂/Li₇P₃S₁₁/Li-In solid-state battery, 3 mg of cathode mixture and 80 mg of solid electrolytes were pressed into a bilayer pellet under 380 MPa. Then, the Li-In anode was placed on another side of the electrolytes and pelletized under 150 MPa to form the solid-state battery. The mass loading for the solid-state batteries was fixed at 3 mg. Galvanostatic charge-discharge measurements were performed under different densities between 0.6 and 2.4 V vs. Li-In and under different temperatures (0°C, room temperature, and 60 C) by a charge/discharge device from Neware (CT4008). Cycling voltammetry curves were obtained from an electrochemical workstation (Solartron, 1470E) at a scan rate of 0.2, 0.4, 0.6, 0.8 and 1.0 mV/s. The in situ/ex-situ EIS measurements were conducted with a Bio-Logic SP-300 in the frequency range of 0.1 Hz–6 MHz with an applied voltage of 0.02 V.

Li₇P₃S₁₁ was chosen as the solid electrolyte in combination of the FeS₂ cathode and Li-In anode in this work to fabricate all-solid-state Li-S batteries. To prepare the target Li₇P₃S₁₁ glass-ceramics solid electrolyte, the typical mechanical milling followed by sintering process was applied. As shown in Figure 1A, the XRD pattern of the raw material mixture after high-rotation process exhibits a typical halo-pattern structure, indicating the formation of an amorphous phase. After a subsequent annealing process at 270°C, the major diffraction peaks are indexed to a pure Li₇P₃S₁₁ phase with a space group of P1. AC impedance was performed on both samples using the stainless steel as the blocking electrodes. The XRD patterns and SEM figures of the synthesized FeS₂@Li₇P₃S₁₁ cathode materials are shown in Supplementary Figure S1. The Li₇P₃S₁₁ glass-ceramic solid electrolyte was obtained by annealing the high-rotation speed milled processor and thus show poor crystallinity. To achieve a good solid-solid interfacial contact of the Li₇P₃S₁₁-FeS₂ cathode mixture, a high milling speeds was applied to mill for a long duration. Therefore, the diffraction peaks due to the Li₇P₃S₁₁ electrolyte are unobvious in the obtained cathode mixture. While for the FeS₂ cathode with good crystallinity, it shows strong diffraction peaks in the pattern. As shown in Figures 1B,C the total resistance of the milled mixture is 554.6 Ω, while it lowers to 60.9 Ω for the obtained Li₇P₃S₁₁ glass-ceramics after annealing at 270°C for 3 h. The corresponding room temperature Li-ion conductivities are 0.14 and 1.27 mS cm⁻¹, respectively. Besides the ultrahigh conductivity at room temperature, the obtained Li₇P₃S₁₁ electrolyte delivers conductivities of 3.16 mS cm⁻¹ and 0.50 mS cm⁻¹ at 60°C and 0°C, respectively. The activation energy can be deduced based on temperature-dependent ionic conductivities. As shown in Figure 1D the milled mixture shows an E _a of 0.30 eV, while the sintered Li₇P₃S₁₁ glass-ceramics delivers a much smaller E _a of 0.22 eV. The precursor obtained by the mechanical milling process is a glass phase. A subsequent sintering process is applied to prepare the final Li₇P₃S₁₁ glass-ceramic solid electrolyte with a higher Li-ion conductivity due to the improved crystallinity of the material. The increase in conductivity and decrease of activation energy is due to the variation of crystal structure, which widens the transporting path of lithium ions. (Peng et al., 2021). SEM image in Figure 1E shows that the particle size of the prepared Li₇P₃S₁₁ glass-ceramic is 5–8 μm with homogenous distribution of P and S in the structure based on the EDS mapping result (Figure 1F).

All-solid-state Li-S batteries consisting of the FeS₂ cathode, the prepared Li₇P₃S₁₁ solid electrolyte, and the Li-In alloy anode were constructed and cycled at different charge/discharge C-rates between 0.6 V and 2.4 V (vs. Li-In). Figure 2A shows the charge/discharge profiles of the chosen 1st, 40th, and 80th cycles when the assembled battery cycled at 0.1C under room temperature and the GITT plots of the assembled batteries are shown in Supplementary Figure S2. During the initial discharge process, the battery delivers two discharge plateaus of ∼1.3 and ∼0.8 V (vs. Li-In) with a discharge capacity of 620.8 mAh g⁻¹. This lower voltage plateau agrees well with the initial discharge plateau (∼1.4 V vs Li⁺/Li^o) of the lithium batteries using organic liquid electrolytes. (Xu et al., 2018). This initial discharge process with two discharge plateaus reflects the 4-electron reactions from the starting FeS₂ active electrode material to Fe and Li₂S. (Zhou et al., 2020). During the following initial charge process, two charge plateaus located at ∼ 1.2 and ∼1.8 V (vs Li-In) are observed in the profile. The lower charge plateau represents the electrochemical reaction between Fe and Li₂S to form the FeS and Li, resulting in a theoretical voltage of 1.7 V (vs Li⁺/Li^o). (Sun et al., 2020). This value agrees well with the first charge plateau voltage of ∼1.2 V vs Li-In (∼1.8 V vs. Li⁺/Li^o). While the higher charge plateau reflects the oxidation of the remaining Li₂S to S with a theoretical voltage of ∼2.3 V (vs. Li⁺/Li^o), which is in good agreement with the other charge plateau observed in the profile at ∼ 1.8 V vs. Li-In (∼2.4 V vs. Li⁺/Li^o). (Mwizerwa et al., 2020). After 80 cycles, the assembled battery sustains a discharge capacity of 284.8 mAh g⁻¹ with a capacity retention of 52.4% based on the second discharge capacity (Figure 2B). Based on previous literatures, the conversion-type electrode materials suffer severe capacity degradation during cycling even with organic liquid electrolytes due to the large volume expansions. (Son et al., 2014). This situation becomes even worse when it combines Li₇P₃S₁₁ solid electrolytes in all-solid-state batteries, the effective solid-solid interfaces between different compounds in the electrode mixture are destroyed due to the volume expansion, yielding large interfacial resistances and poor cycling performance for the assembled all-solid-state Li-S batteries. Besides, the rate capability of the assembled battery was also validated at different charge/discharge C-rates. In Figure 2C, at a low charge/discharge rate, the platforms are caused by the two-step reaction with FeS₂ → FeS + Li₂S → Fe + Li₂S, accompanied by two plateaus at 1.5 V and 0.8 V (vs Li-In), respectively. In contrast, the plateau represents the one-step reaction (FeS₂ → Fe + Li₂S) at a high rate (such as 1C). As shown in Figure 2D, the FeS₂/Li₇P₃S₁₁/Li-In battery delivers discharge capacities of 591.5 mAh g⁻¹ at 0.05C, 587.6 mAh g⁻¹ at 0.1C, 542.3 mAh g⁻¹ at 0.2C, 417.8 mAh g⁻¹ at 0.5C, and 313.6 mAh g⁻¹ at 1°C, respectively, showing a rate capability retention of 53.0% (for rate ranging from 0.05 to 1C).

In addition, ex-situ XRD was performed on the cycled cathode mixtures of the assembled FeS₂/Li₇P₃S₁₁/Li-In batteries after different charge/discharge states to unravel the phase evaluations during the initial cycle. In Figure 3B, the diffraction peak at ∼ 21° is assigned to the protective film to isolate the direct contact between the cycled mixture and the air/moisture. As shown in Figure 3B, the diffraction peaks of point A are indexed to the pure phase of the FeS₂ phase at the very beginning of the initial discharge process. It should be mentioned that very weak diffraction peaks due to the Li₇P₃S₁₁ solid electrolyte can be detected in this pattern. That’s because the electrode mixture was prepared by mixing the crystalline FeS₂ with glass ceramic Li₇P₃S₁₁ electrolyte with a high rotation speed up to 500 rpm for long milling durations to ensure good solid-solid interfaces. The Li₇P₃S₁₁ electrolyte in the mixture transfers to an amorphous phase after this mechanical milling process, which makes it difficult to be detected by the typical powder XRD. (Prasada Rao et al., 2016). During the discharging process from A to C, the diffraction peaks assigned to the FeS₂ phase become weak because of the reaction between the FeS₂ and Li ions. During this initial discharge process, the XRD peaks indexed to the FeS₂ phase become weaker while the XRD diffraction peaks belonging to the Li₂S and Fe become stronger, which is attributed to the conversion reaction from FeS₂ to Li₂S and Fe. In the subsequent charging process of the first cycle, more diffraction peaks indexed to the Li₂S and FeS phases are detected in the XRD patterns of point C and point D. These peaks almost disappear in the subsequent processes and the FeS phases are observable when the electrode is charged to the high cut-off voltage of 2.4 V (vs Li-In) at point F in Figure 3A, suggesting the phase transformation from the Li₂S/Fe to the FeS. It should be mentioned that no signal belonging to Fe is detected in the XRD pattern. Based on previous research, a spontaneous reaction will occur between Fe and FeS₂ to form the FeS. (Zhou et al., 2020). Partial of the FeS formed during the charging/discharging processes shows an amorphous phase structure, which makes it difficult to be observed with the typical powder XRD and TEM characterization methods. (Wan et al., 2019). As shown in Figure 3A, a tiny capacity is obtained during the charging process from 1.3 V (Point D) to 1.8 V (vs. Li-In). When the charging voltage rises to point E, the oxidation reaction from Li₂S to S occurs during this process, resulting in decreased Li₂S phase and increased S in the cycled cathode mixture. Therefore, weaker diffraction peaks assigned to the Li₂S phase can be detected in the XRD pattern of point E. When the charging voltage further reaches the upper cut-off voltage (2.4 V vs. Li-In), almost all Li₂S in the cathode mixture has already been transferred to the amorphous S, S_x, and the FeS phases. No diffraction peaks belonging to the S and its analogous can be detected in the XRD pattern of the cathode mixture obtained after the initial charging process (point F) due to the amorphous phase of these yielded products. In the following cycles, the charge/discharge capacities are associated with the lithiation/delithiation reaction of those S, S_x, and FeS complex formed during the first cycle.

In-situ EIS was performed on the FeS₂/Li₇P₃S₁₁/Li-In battery during the initial cycle when cycled at room temperature. As shown in Supplementary Figures S4, S5, the battery shows small total resistances at different charge/discharge states under 60°C than that at room temperature during the first cycle. That’s why the battery delivers higher initial charge/discharge capacities at 60°C than at room temperature. It also implies that the volume expansions in the cathode mixture at the beginning slightly affect the battery performance. When the operating temperature decreases to 0°C, the battery exhibits the largest total resistances among those different temperatures, yielding the smallest initial charge/discharge capacities. Since the spectra show small changes during cycling, the distribution of relaxation time (DRT) based on the EIS spectra were applied to unravel the contribution of resistances from different sections of the assembled battery. As depicted in these Figure 4, the peaks centered at 10^–3∼10^–1 Hz is assigned to ion transport across the negative and positive interfaces. The peaks located below 0.1 Hz is related to the solid-state diffusion of Li-ion in FeS₂ in the cathode mixture. Those peaks at 0.1 Hz show clear variations under different discharge states. The intensities first decrease at the beginning of the discharge process and then become stable, reflecting the two lithiation processes of FeS₂ to Li₂S phase in the cathode mixture during cycling. (Li et al., 2022). During the subsequent charging process, obvious peaks are also detected in the DRT figure. The intensity of those peaks centered at 10^–3∼10^–1 Hz increase during the charging process, suggesting a continuous electrochemical reaction associated with the formed Li₂S in the cathode mixture.

Due to the high ionic conductivities of the prepared Li₇P₃S₁₁ solid electrolytes at different temperatures, the corresponding electrochemical performances of the assembled FeS₂/Li₇P₃S₁₁/Li-In all-solid-state Li-S batteries under these operating temperatures between 0.6 and 2.4 V (vs Li-In) were also investigated. As shown in Figure 5A, the FeS₂/Li₇P₃S₁₁/Li-In battery delivers much higher initial discharge and charge capacities of 866.4 mAh g⁻¹ than that at room temperature when the battery cycled at 0.1C. Moreover, similar voltage plateaus are observed during the first charge and discharge processes as that at room temperature. However, it shows much faster discharge capacity decay in the subsequent 40 cycles (Figure 5B). After 40 cycles, the battery only sustains a discharge capacity of 61.8 mAh g⁻¹ with a low-capacity retention of 7.1%. The poor cyclability may be associated with the large interfacial resistances of the FeS₂/Li₇P₃S₁₁/Li-In battery during cycling when operated at elevated temperatures. One possible reason is the huge volume expansion. (Liao et al., 2022b; Wei et al., 2022b; Wu et al., 2022). As described in the previous section, the FeS₂ electrode materials suffer severe volume expansion at room temperature due to the conversion reaction occurring during cycling. The elevated operating temperatures make this situation even worse. Large volume expansions cause the loss of effective contact between the active material and solid electrolyte particles, yielding lower discharge capacities and poor cycling performances in Figure 5B.

The prepared Li₇P₃S₁₁ electrolyte shows a high Li-ion conductivity of 0.50 mS cm⁻¹ at 0°C. Considering the huge volume expansions may be mitigated under low operating temperatures, providing a potential application field for this FeS₂/Li₇P₃S₁₁/Li-In solid-state battery. As demonstrated in Figure 5C, this battery can reversibly cycle at 0.05C under 0 C for long cycles. It shows almost the same initial discharge and charge curves at this temperature compared to room and elevated temperatures, indicating a similar electrochemical reaction in the cathode mixture during cycling. Due to the lower Li-ion conductivity of the Li₇P₃S₁₁ electrolyte at 0°C, the FeS₂/Li₇P₃S₁₁/Li-In battery exhibits a much lower discharge capacity of 364.8 mAh g⁻¹ under the same test conditions. However, it shows much better cycling performances as depicted in Figure 5D. It sustains a discharge capacity of 256.6 mAh g⁻¹ with a capacity retention of 87.9% from 2nd to 100th cycle. Both the capacity and retention values are much higher than that at higher temperatures. As depicted in Figure 5E, it delivers discharge capacities of 458.6 mAh g⁻¹ at 0.05C, 390.5 mAh g⁻¹ at 0.1C, 335.4 mAh g⁻¹ at 0.2C, 265.4 mAh g⁻¹ at 0.5C, and 197.2 mAh g⁻¹ at 1.0°C, respectively. It shows superior rate capability at this low operating temperature. Finally, this battery was also cycled at a larger C-rate of 0.5C under the same condition. As shown in Figure 5F, it delivers an initial discharge capacity of 132.5 mAh g⁻¹ and maintains a discharge capacity of 88.0 mAh g⁻¹ with a capacity retention of 69.1% after 100 cycles.

The FeS₂ electrode suffers large volume expansions in lithium batteries with the typical organic liquid electrolytes. (Whiteley et al., 2016). Similarly, these volume changes in all-solid-state lithium batteries play an even more crucial role in battery performances. The good solid-solid interface contact ensures the effective Li-ion transport between different particles in the cathode mixture. (Xi et al., 2019). On the contrary, huge volume variations yield poor solid-solid contacts and large interfacial resistances, resulting in lower capacities and fast capacity decay during cycling. To study the resistance changes of the FeS₂/Li₇P₃S₁₁/Li-In battery before and after cycling under different temperatures, EIS was performed. All batteries exhibit an obvious increase in total resistances after cycles. The resistance of the solid electrolyte layer measured in the high frequencies shows minor changes before and after cycling at different operating temperatures based on the EIS spectra in Figures 6A–C, suggesting that the resistances of the solid electrolyte layers of these solid-state lithium batteries stay constant during cycling at various operating temperatures. The major variations of resistance for these batteries come from the interfacial section measured in the middle frequencies. These interfacial resistances increase largely after cycling in Supplementary Figure S3, which may be associated with the huge volume changes from different layers in the solid-state batteries and the conversion reaction of FeS₂ in the cathode mixture. (Wei et al., 2022a). The battery cycled at 60°C (Figure 6B) shows the largest increase of interfacial resistance after cycling among those different operating temperatures, which agrees well with the fast degradation of discharge capacities in Figure 5B. While the battery cycled at 0°C (Figure 6C) exhibits the smallest increase of interfacial resistances after 100 cycles, indicating small volume expansions under this temperature. Therefore, the assembled battery shows superior cycling performance at this temperature. At elevated temperatures, the FeS₂ active materials suffer much larger volume expansions during cycling, which destroys the effective solid-solid contact between FeS₂ and Li₇P₃S₁₁ electrolyte, resulting in poor cycling performance. When the operating temperature lowers to 0°C, small volume changes are expected due to the shrink effect of materials under cold temperatures. This can maintain effective Li-ion transport across the FeS₂ and Li₇P₃S₁₁ solid-solid interfaces, enabling an excellent cycling performance.

In-situ stack pressure tests were performed on the FeS₂/Li₇P₃S₁₁/Li-In battery when cycled under the above operating temperatures at 0.5C between 0.6 and 2.4 V (vs Li-In) to monitor the pressure evolutions in the assembled battery during cycling. As shown in Figure 6, the stack pressure first decreases during the initial discharge processes and increases in the subsequent charge processes. The variation of pressure is associated with the volume expansion and shrinkage of FeS₂ in the cathode mixture during cycling. (Xu et al., 2017). As depicted in Figure D, F, FeS₂/Li₇P₃S₁₁/Li-In battery shows the highest (1.68 MPa) and lowest (0.25 MPa) pressure changes when cycled at 60°C and 0°C, respectively, indicating the largest and smallest volume changes under the corresponding temperatures.

These results agree well with our above analysis.

Finally, Cyclic voltammetry (CV) tests under different scan rates were applied to investigate Li-ion mobilities in the cathode mixture of the FeS₂/Li₇P₃S₁₁/Li-In battery when worked at different operating temperatures. As shown in Figures 7A–C, similar oxidation/reduction peaks are observed in the CV curves measured at room temperature and 60°C, while no clear peaks can be detected in the CV plots when scanned at 0°C. These results suggest that the battery delivers much higher capacities at elevated temperatures (RT and 60°C) than at low temperatures (0°C). As presented in Figures 7D, E, the FeS₂/Li₇P₃S₁₁/Li-In battery shows much higher ln (i_p) values at 60°C than that at RT, suggesting fast Li-ion diffusion rates in the cathode mixture under 60°C. Therefore, the battery delivers much higher charge/discharge capacities at higher temperatures (Figure 5B).

In summary, the pure Li₇P₃S₁₁ glass-ceramic electrolyte is successfully synthesized using high-rotation milling followed by a sintering route with a high Li-ion conductivity of 1.27 mS cm⁻¹ at room temperature. All-solid-state lithium battery using this prepared Li₇P₃S₁₁ electrolyte combined with FeS₂ cathode and Li-In anode delivers a high initial discharge capacity of 620.8 mAh g⁻¹ at room temperature when cycled at 0.1C. Ex-situ XRD results show that the initial discharge process is associated with the formation of Li₂S from the FeS₂ process, while the following charging process is assigned to the electrochemical reaction of Li₂S. Due to the huge volume expansions that occur at elevated operating temperatures, the assembled FeS₂/Li₇P₃S₁₁/Li-In battery delivers a higher initial discharge capacity of 866.4 mAh g⁻¹ at 60°C with fast degradation of capacity in the subsequent cycles. Interestingly, although the Li₇P₃S₁₁ electrolyte shows a decreased conductivity at a lower temperature, the battery can deliver a high discharge capacity of 364.8 mAh g⁻¹ at 0.1C when cycled under 0°C. Moreover, this battery can also show reversible charge/discharge capacity at a higher rate of 0.5C over 100 cycles with excellent cycling performances under the same operating temperature. EIS and in situ stack-pressure test results confirm that the superior cycling performance under 0°C and the fast degradation of capacity at 60°C are attributed to the volume changes of FeS₂ in the cathode mixture. This work reveals the temperature effects on the volume changes of solid-state batteries, providing the design principle for constructing high-performance solid-state batteries in a wide temperature range.