Realizing high-capacity all-solid-state lithium-sulfur batteries using a low-density inorganic solid-state electrolyte

Lithium-sulfur all-solid-state batteries using inorganic solid-state electrolytes are considered promising electrochemical energy storage technologies. However, developing positive electrodes with high sulfur content, adequate sulfur utilization, and high mass loading is challenging. Here, to address these concerns, we propose using a liquid-phase-synthesized Li3PS4-2LiBH4 glass-ceramic solid electrolyte with a low density (1.491 g cm−3), small primary particle size (~500 nm) and bulk ionic conductivity of 6.0 mS cm−1 at 25 °C for fabricating lithium-sulfur all-solid-state batteries. When tested in a Swagelok cell configuration with a Li-In negative electrode and a 60 wt% S positive electrode applying an average stack pressure of ~55 MPa, the all-solid-state battery delivered a high discharge capacity of about 1144.6 mAh g−1 at 167.5 mA g−1 and 60 °C. We further demonstrate that the use of the low-density solid electrolyte increases the electrolyte volume ratio in the cathode, reduces inactive bulky sulfur, and improves the content uniformity of the sulfur-based positive electrode, thus providing sufficient ion conduction pathways for battery performance improvement.

To illustrate how sulfur's weight ratio and SE's density affect SE's volumetric content in the cathode, we assume that the weight ratio of carbon and sulfur are 15 wt% and wt%. Moreover, the density of carbon, sulfur, and SE are 1.9, 2.07, and ρ g cm -3 , respectively. Hence, the volume ratio of SE could be calculated    The specific energy of Li-S ASSBs is estimated using the following parameters: 1. Sulfur cathode 1) The volumetric content of sulfur is 50 vol%.
2) The weight percent of sulfur in the cathode is w wt%.
3) The volumetric carbon content in the cathode is fixed to be 15 vol%.

4) The density of SSE is ρ SSE
5) The specific discharge capacity of the sulfur cathode is 1000 mAh g -1 .
6) The average voltage during discharge is 2.0 V.

Membrane
1) The thickness of the membrane is 25 μm.

Anode
1) The areal capacity ratio of the negative electrode to the positive electrode (N/P) is 2.

Current collectors
1) The areal weight of current collectors (Cu and Al) is 5.83 mg cm -2 , which is estimated based on the double-sided electrode cell design.
Next, the cell-level specific energy of Li-S ASSBs could be estimated using the following equation: Specific energy= 1000 mAh g -1 ×m g sulfur cm -2 ×2.0 V 0.0025 cm×ρ SSE g cm -3 + 2×1000 mAh g -1 ×m g sulfur cm -2 3860 mAh g -1 + m g sulfur cm -2 w % +0.00583 g cm -2 The weight percent of sulfur in the cathode could be calculated using the following equation: w %= 2.07 g cm -3 ×50 vol% 2.07 g cm -3 ×50 vol%+1.9 g cm -3 ×15 vol%+ρ SSE ×(100-50-15) vol% With different SE densities and areal sulfur loading, the specific energy of Li-S ASSBs and sulfur content (wt%) in the cathode can be calculated, and the results are depicted in Figure 1b  Compared with solid-phase synthesized sulfide SEs, liquid-phase synthesized sulfide SEs usually have more impurities due to the side reactions of precursors with organic solvents, making the determination of the detailed composition of SE crystals challenging. In previous literature, the molar ratios of added precursors are typically used to determine the composition of the formed crystals synthesized via the liquid-phase method 9,18,27 . According to the molar ratios of the precursors used for synthesizing LPB (Li2S/P2S5/LiBH4=3/1/4), the composition of the formed crystal should be Li5PS4(BH4)2. However, considering the presence of impurities, the actual composition may differ from the calculated formula.
Although the XRD patterns LPB share similar peak position and relative peak intensity with previously reported XRD patterns of Li6PS5BH4 32 , we believe it is inaccurate to assign the liquid phase synthesized crystal to be Li6PS5BH4 argyrodite. Therefore, we use Li6-xPS5-x(BH4)1+x (-1≤x≤1) to represent the crystal composition of LPB.
Supplementary Note 3. Electrochemical stability of LPB against lithium metal anodes. Furthermore, we also examined the electrochemical stability of LPB against lithium metal using a Li|LPB|Li symmetric cell. The cell was assembled using a hot-pressed LPB pellet and cycled at different current densities from 0.1 mA cm -2 to 1.0 mA cm -2 under 6 ~ 8 MPa at room temperature. The voltage profile as a function of time is shown in Supplementary Figure 14a. The cell cycled stably without short circuit failure until the current density reached above 1 mA cm -2 , demonstrating that the critical current density is 1 mA cm -2 . To gain an in-depth understanding of the evolution of the SEI layer, both areal total resistance (R t ) and areal interfacial resistance (R int ) were calculated using the following equations: where U is the voltage polarization, J is the current density, R pellet is the areal bulk resistance contributed by the electrolyte pellet, l is the thickness of the pellet, and σ 25℃ is the measured ionic conductivity of LPB at 25 o C. The calculated resistance and the voltage polarization are illustrated in Supplementary Figure 14b. During the first few cycles at 0.1 mA cm -2 , the voltage polarization, total resistance, and interfacial resistance gradually increased due to the growth of the SEI layer. After the current density increased to 0.25 mA cm -2 , the evolution of the SEI layer stabilized with a low interfacial resistance of 6 ~ 7 Ω cm 2 , which remains almost constant even at 1 mA cm -2 . This result demonstrates the formation of a low-resistance metastable SEI layer at the Li/LPB interface.
Further analysis of the SEI layer by XPS reveals that this low-resistance SEI layer is mainly composed of Li2S. High-resolution XPS spectra of Li 1s, P 2p, and S 2p were collected for the original LPB SE 1. The estimation of sulfur cathodes' specific energy is based on the following parameters: 1) The weight percent of sulfur in the cathode is w wt%.
2) The specific discharge capacity of the sulfur cathode is a mAh g -1 .
3) For the S-C-LPB cathode, we used the average discharge voltage of 1.935 V (vs. Li/Li + ) for estimation. 4) Areal cathode loading is m mg cm -2 .
Next, based on the above parameters, the specific energy of sulfur cathodes could be estimated using the 2) The specific discharge capacity of the sulfur cathode is a mAh g -1 .
3) For the S-C-LPB cathode, we used the average discharge voltage of 1.935 V (vs. Li/Li + ) for estimation.
4) The porosity of the cathode is 95%.
Energy density= a mAh g -1 ×m S mg cm -2 ×1.935 V LPB, like other sulfide-based SEs, is electrochemical redox active and can contribute to capacity during charge/discharge cycles in the S-C-LPB cathodes (Supplementary Figure 17). Here, we further discuss the capacity contribution from the LPB redox reaction in the S-C-LPB cathode (KB/S/SE=10/50/24, w/w/w) based on the data in Figure 3e. If we assume the LPB delivers the same discharge capacity as was measured in Supplementary 17, the discharge capacity contribution of LPB is ~ 5.8 mAh g sulfur -1 (~ 12.1 mAh g LPB -1 ) at the first cycle and < 187.7 mAh g sulfur -1 in the following cycles. It shows that sulfur contributed to ~ 99.5 % discharge capacity at the first cycle and > 82% discharge capacity (based on the discharge capacity of the S-C-LPB at the last cycle in Figure 3e and LPB's maximum capacity contribution of 187.7 mAh g sulfur -1 in the following cycles at 167.5 mA g -1 . Note that it is just a rough estimation. The capacity contribution of LPB may vary depending on the internal resistance of the sulfur cathode, the contacting area between carbon and LPB, and the current density applied. Considering the S-C-LPB cathode with a large amount of sulfur and a small amount of LPB & carbon, the actual capacity contribution from LPB might be smaller than the abovecalculated values. Collectively, the results show that the good discharge capacity of S-C-LPB is mainly attributed to the high sulfur utilization.
Supplementary Note 6. Causes for the capacity decay of Li-S ASSBs during long-term cycling.
Two main factors contribute to the capacity fading of Li-S ASSBs using S-C-LPB cathode, namely, electrochemical and chemical degradation of LPB SE.

Electrochemical degradation of LPB SE.
Sulfide SEs typically possess a narrow electrochemical stability window between 2 ~ 2.5 V (vs. Li/Li + ). 33 ) and capacity decay. After 1000 cycles, even though no Li2S was observed in the cathode at the charging state, the intensity of the bridging sulfur (SB 0 ) peak significantly decreased. Instead, the peak corresponding to terminal sulfur (ST 1-)/P-Sx-P shows the highest intensity, possibly suggesting that lithium polysulfide (Li2Sx) reaction intermediates were not fully oxidized to S8 due to the large resistance of the cathode. Supplementary Note 7. Influence of SE particle size on cathode performance. The influence of SE particle size on LiNi0.5Mn0.3Co0.2O2 (NMC) cathode performance in all-solid-state lithium batteries has been reported in the literature, revealing that smaller SE particle size is beneficial for attaining sufficient ionic percolation pathways in the cathode. 36 Here, we additionally show that SE particle size will also influence sulfur cathode performance by regulating ionic transport pathway sufficiency. indicating poor content uniformity and insufficient Li-ion transport pathway. Intriguingly, in the meantime, we also observed phase separation between SE and carbon in the SEM-EDS images, possibly originating from the large particle size of LPSC. The phase separation, detrimental to ion/electron transport, shall further jeopardize the electrochemical performance of the cathode. Collectively, the results show that, besides the volume ratio of SE in the cathode, the large particle size of SE shall also compromise cathode content uniformity and induce the formation of inactive bulky sulfur, rendering insufficient Li-ion transport pathways and poor sulfur utilization. These results also highlight the uniqueness of the LPB SE developed via a novel liquid-phase method with low density and small particle size.