In Situ-Constructed Li_xMoS₂ with Highly Exposed Interface Boosting High-Loading and Long-Life Cathode for All-Solid-State Li–S Batteries

2.1 Materials Characterization

The preparation procedures of the S/MoS₂/C composite are presented in Figure 1a. The sulfur host material (MoS₂/C) has been prepared based on a hydrothermal-annealing method, using Na₂MoO₄/thiourea as MoS₂ precursors and oxidized CNTs/glucose as carbon sources.^[32] In the synthesis, CNTs act as heterogeneous nucleation sites and inhibitors to impede vertical stacking of S-Mo-S layers and induce their lateral growth, while glucose molecules can intercalate between adjacent MoS₂ monolayers and reduce surface energy.^[33] As a result, the ultrathin MoS₂ nanosheets with less than five stacked layers could be attained, which could be spontaneously crumpled into a paper ball-like shape during the hydrothermal reaction. The subsequent annealing in Ar results in the complete carbonization of oxidized CNTs and glucose, endowing the MoS₂/C composite with good electronic conductivity. With these unique structural characteristics, the MoS₂/C composite features robust mechanical elasticity and anti-aggregation ability. Sulfur has been loaded onto MoS₂/C by a vapor deposition method, and the saturated sulfur content in the resultant S/MoS₂/C composite is ~32.3 wt% determined by TG analysis (Figure 1b).

Spectroscopy characterizations clarify materials' chemical compositions. The XRD pattern of S/MoS₂/C is shown in Figure 1c, with elemental S powder and MoS₂/C as references. The MoS₂/C host exhibits a few weak diffraction peaks indexed to the few-layered 2H-MoS₂ phase, similar to the reported result.^[32] It is worth noting that the (002) diffraction peak of the 2H-MoS₂ phase is significantly broadened, related to the coexistence of non-expanded and carbon-intercalated MoS₂.^[34] Upon sulfur deposition, all those peaks become almost invisible, and strong reflection signals associated with crystalline S element also disappear. This result suggests that the sulfur deposited onto the MoS₂/C surface is highly dispersed without bulky sulfur residue in S/MoS₂/C.^[35] Besides, no detectable reflections assigned to CNTs or pyrolyzed carbon are found in the XRD patterns of MoS₂/C and S/MoS₂/C, probably attributed to their low content (9.78 and 6.62 wt% in MoS₂/C and S/MoS₂/C, see Table S1, Supporting Information). Raman spectra (Figure S1, Supporting Information) confirm the presence of carbon components in both samples as characteristic D- and G-band peaks emerge around 1336 and 1580 cm⁻¹, respectively.^{[36, 37]} Other sharp Raman peaks centered at 377 and 402 cm⁻¹ are associated with in-plane (${\mathrm{E}}_{2\mathrm{g}}^1$) and out-plane (A_1g) vibrations of S atoms in MoS₂, respectively.^[36] Further, nitrogen physisorption analyses (Figure S2, Supporting Information) reveal that MoS₂/C features highly exposed surfaces with a specific surface area of 109 m² g⁻¹, significantly reduced to 34 m² g⁻¹ for S/MoS₂/C upon sulfur loading. XPS is a surface sensitive technique employed to identify the surface chemical state and composition of MoS₂/C and S/MoS₂/C. As shown in Figure S3, Supporting Information, the XPS survey spectra demonstrate that both MoS₂/C and S/MoS₂/C consist of C, Mo, and S elements. In the high-resolution Mo 3d & S 2s spectrum of MoS₂/C (Figure 1d), the component peaks centered at 233.3 and 230.1 eV correspond to Mo 3d_3/2 and Mo 3d_5/2 of Mo⁴⁺, respectively; another broad peak located at 227.3 eV is characteristic of S 2s. The S 2p spectrum of MoS₂/C (Figure 1e) displays two dominant peaks at 164.1 and 162.9 eV relating to S 2p_1/2 and S 2p_2/3 of S²⁻, respectively. As for S/MoS₂/C, all those peaks associated with Mo 3d, S 2s, and S 2p prominently shift toward lower binding energy, which can be attributed to the interaction between MoS₂ and elemental S after sulfur loading. Meanwhile, new doublet peaks at 165.2 and 164.1 eV can be deconvoluted in the S 2p spectrum of S/MoS₂/C, which is derived from the elemental S. The above spectroscopy results confirm the sulfur deposition in S/MoS₂/C.

The structure of MoS₂/C and S/MoS₂/C samples can be observed by SEM and TEM in detail. As shown in Figure 1f, the S/MoS₂/C composite exhibits a spherical shape with a diameter ranging from 200 to 300 nm, accompanied by some surface ridges and folds (Figure 1g). Furthermore, the internal carbon nanotubes (indicated by white arrows) within the S/MoS₂/C particle are observable. These CNTs are wrapped in thin layers, providing a highly conductive pathway for electrons through the interparticle interfaces of the S/MoS₂/C, which also can contribute to avoiding their direct contact with SSEs, thus reducing the decomposition of SSEs. The high-resolution TEM image of the S/MoS₂/C composite is presented in Figure 1h, revealing nanolayers that are less than five layers in thickness. Notably, two sets of lattice fringes with interplanar distances measuring 0.626 and 1.021 nm can be discerned. The lattice fringe exhibiting interplanar distances of 0.626 nm is attributed to the (002) plane of 2H-MoS₂, while the broader interlayer spacing (1.021 nm) is a result of carbon intercalation derived from glucose, which facilitates electron/ion transfer.^{[29, 38-40]} The lattice fringes with d-spacings of 0.227 and 0.272 nm also correspond to the (103) and (100) planes of 2H-MoS₂.^[34] The aforementioned results are in accordance with the XRD findings of host material MoS₂/C. The TEM images of the MoS₂/C host are also shown in Figure S4, Supporting Information. The structural changes before and after sulfur loading are not discernible. For the S/MoS₂/C composite, the mappings reveal that C, Mo, and S elements are evenly dispersed across the whole composite. It is hard to distinguish between sulfur from MoS₂/S host because of the same morphology and chemical composition. The possible reason is that electron beam irradiation triggers the sublimation of sulfur in an ultra-high-vacuum environment during TEM measurement. Fortunately, the presence of sulfur has been confirmed by TG and XPS analysis. Furthermore, AFM has been used to measure the thickness of the monolayer of S/MoS₂/C pieces after they were deliberately broken with ultrasonic technique (Figure 1j). AFM analysis suggests that the thickness of the S/MoS₂/C layer is <2 nm. This is clear evidence to prove that the S/MoS₂/C layer is thin.

The cathode is prepared by mechanically milling the S/MoS₂/C composite and Li₆PS₅Cl powders in a 6:4 mass ratio, without any additional carbon additives. Based on the low carbon content (6.62 wt%) in the S/MoS₂/C composite, the use of the carbon additive is greatly reduced in the S/MoS₂/C/Li₆PS₅Cl cathode composite (only ~3.97 wt%). As shown in Figure 2a, cathode composite powder shows ball-like nanoscaled particles and homogeneity uniform dispersion with Li₆PS₅Cl. Further cold pressing by 360 MPa for 3 min, the layer construction was obtained (Figure 2b). The cross-sectional SEM image of the cathode layer shows a thickness of 42 μm in ASSLSBs. To get a more detailed view (Figure 2c), it can be found that the S/MoS₂/C are distributed uniformly with squashed morphology in the cathode layer. In addition, the elemental mappings of the cathode are shown, where S, Mo, and P are evenly distributed throughout the cathode, with much less content of P (Figure S5, Supporting Information), suggesting the Li₆PS₅Cl content is less common in the cathode, and the S/MoS₂/C is the main part of the cathode. For the visual description in Figure 2d, a high interfacial contact area between Li₆PS₅Cl and S/MoS₂/C would be formed after ball milling. Further cold pressing during the ASSLSBs assembling, the S/MoS₂/C composites are changed from ball-like nanoparticles to nanosheets, which is favorable to establish the closer area-to-area contact between the active material and Li₆PS₅Cl and low-tortuous carrier transport pathways for ion/electron during (dis)charging in the inner cathode layer. This construction is beneficial to the subsequent electrochemical performance.

2.2 Reaction Mechanism and Interface Evolution of ASSLSBs during Cycling

The cyclic voltammogram profiles of the as-synthesized S/MoS₂/C cathode are displayed in Figure 3. The onset potential of the S/MoS₂/C (Figure 3a) and MoS₂/C (Figure 3b) start at ~1.6 and ~1.1 V versus Li⁺/Li-In, respectively. It's possibly caused by the sulfur component in the S/MoS₂/C cathode. For the S/MoS₂/C cathode (Figure 3a), two distinct peaks are observed at ~1.3 and ~0.4 V versus Li⁺/Li-In during cathodic curves. The peak at ~1.3 V versus Li⁺/Li-In is attributed to the conversion of sulfur (S) into lithium sulfide (Li₂S), which is consistent with the position of the reduction peak for S as depicted in Figure 3b.^[31] The peak observed at ~0.4 V versus Li⁺/Li-In can be attributed to the insertion of lithium ions into the MoS₂ interlayer, leading to the formation of Li_xMoS₂. This observation is consistent with the lithiation process of the MoS₂/C cathode sample, as indicated by the peak position.^[41] In the anodic curves, an oxidation peak is observed, which may be attributed to the proximity of the potential oxidation peak for Li₂S and that of the delithiation process for Li_xMoS₂. Therefore, it is reasonable to assign the primary oxidation peak at ~2.2 V versus Li⁺/Li-In to the charging reactions involving delithiation of Li_xMoS₂ to MoS₂ and the conversion of Li₂S to S.^{[12, 14]}

Overall, the MoS₂/C exhibits an electronic conductivity, while the lithiation product (Li_xMoS₂/C) is a mixed ion/electron conductor, combining Li₆PS₅Cl (ionic conductor) and S (active material) to form the intimate nanosized triple-phase boundaries. The sufficient triple-phase interface makes a great contribution to the excellent electrochemical performance.

The GITT technique is employed to investigate the evolution of the cathode internal interface during cycling (Figure 4a). The cell is subjected to a series of current pulses, each followed by a 3-h relaxation period. The sulfur composite cathode exhibits a pair of plateaus at ~1.1 and ~2.2 V during the discharge/charge process, which is consistent with the CV curves. The electrode always undergoes an activation process during the first lithiation, resulting in a different appearance of the first sweep compared to subsequent ones (Figure 4a).^{[13, 44, 45]} Furthermore, the voltage polarization (ΔV) of each current pulse is determined by analyzing the relaxation curve presented in Figure 4b,c. The ΔV represents the combined effect of the iR drop and relaxation process. The iR drop is attributed to the ohmic resistance at the interface between the sulfur composite and Li₆PS₅Cl, while the relaxation process involves Li⁺ concentration reaching an equilibrium state.^[46] During discharging, the ΔV initially exhibits an increasing trend after reaching 0.8 V versus Li⁺/Li-In, whereas during charging, the ΔV shows an initial increase followed by a decrease. Generally, the value of ΔV in the 2nd cycle is slightly lower than that in the 1st cycle, which can be attributed to electrode activation. Ex situ EIS spectra were acquired during the cycling process. The overall impedance exhibits an increase gradually, which is consistent with the polarization plots obtained through GITT. During the discharge process, the semicircular response in the mid-frequency range expands toward lower frequencies. During the charging process, the semicircle feature expands sequentially toward lower frequencies, while the straight lines in the low-frequency region are progressively suppressed at 2.1 V versus Li⁺/Li-In or even vanish at 2.8 V versus Li⁺/Li-In. As marked by the blue arrow in Figure 4d, the phenomenon occurs in the second cycle. The impedance spectra at 0.4 V versus Li⁺/Li-In between two consecutive cycles remain almost unchanged, implying the reversibility of the electrode reaction.

2.3 Theoretical Calculations

To understand the distinctive electrochemical performance of the MoS₂/C host in the sulfur cathode, density functional theory (DFT) calculation is conducted to elucidate its unique mechanisms. The host models of MoS₂/MoS₂ and MoS₂/C are depicted in Figures S6 and S7, Supporting Information. Due to the presence of a sulfur coating on the surface of MoS₂/C, we exclusively consider the hollow site of MoS₂ as the preferred formation site for Li_xMoS₂. The MoS₂/C and MoS₂/MoS₂ bilayer system serves as a simplified model for the subsequent calculations.^[32] As depicted in Figure 5a, the MoS₂/MoS₂ bilayer system has a Li_xMoS₂ formation energy of 2.75 eV. In comparison, the MoS₂/C bilayer system has a lower Li_xMoS₂ formation energy of 2.26 eV, which renders a faster reaction kinetics between lithium and MoS₂/C. The migration of Li⁺ is a crucial factor in evaluating the performance of batteries, as it accompanies both the charging and discharging processes. The migration pathway of lithium in MoS₂/MoS₂ and MoS₂/C bilayer systems is subsequently investigated (Figure 5b). The migration of lithium is simulated as it moves from one favorable adsorption site to another, while the barrier is calculated (Figure 5c). The lithium migration barrier in MoS₂/MoS₂ is 0.286 eV, which is significantly higher than that in the MoS₂/C bilayer, 0.0875 eV. This indicates that the migration rate of Li atoms is faster in the latter. For ultrathin S/MoS₂/C, the maximized exposed surface facilitates direct areal-to-areal contacting with Li₆PS₅Cl and low-tortuous carrier transport pathways for ion/electron in the inner cathode layer. Additionally, it is also interesting to study the lithium process occurring on the exterior surface of the sulfur coating. For the conversion process from sulfur to Li₂S in the ASSLSBs, the electrochemical redox reactions are conversions between the sulfur and Li₂S with an intermediate phase of Li₂S₂ formation.^[47] In fact, the conversion from Li₂S₂ to Li₂S is a rate-limiting step.^[40] Therefore, the transition from Li₂S₂ to Li₂S is utilized as a simplified model for calculating sulfur dissociation. The dissociation pathway of Li₂S₂ on the Li_xMoS₂ and Li_xMoS₂/C surfaces with lithium interaction and energy barrier is shown in Figure 5d,e, respectively. The dissociation energy barrier of Li₂S₂ calculated for MoS₂/C with lithium interaction is 1.77 eV, which is lower than that of MoS₂/MoS₂ (1.91 eV) with lithium interaction, indicating superior performance in activating sulfur during the discharging process.

2.4 Electrochemical Performance

It is essential to investigate the electrochemical performance of ASSLSB with dual active components (S + MoS₂/C) at room temperature. Figure 6a illustrates the discharge–charge profiles of S/MoS₂/C and S/B-MoS₂ cathodes during the 2nd cycle at 100 mA g⁻¹ with active loading of 1.3 mg cm⁻². In the second cycle, a discharge capacity of 968.3 mAh g⁻¹ is observed from the S/MoS₂/C cathode. The capacity is gradually increased and then stabilized during cycling, which is 980.3 mAh g⁻¹ at the 100th cycle (Figure 6b). This corresponds to a high reversible discharge capacity of 588.2 mAh g⁻¹ based on the whole cathode weight. In comparison, the discharge capacity in the second cycle for the S/B-MoS₂ cathode is only ~35 mAh g⁻¹. It is evident that the nanostructure of the sulfur host plays a crucial role in determining its capacity. The SEM image, as shown in Figure S8, Supporting Information, reveals the structure of B-MoS₂ with a 100 nm-thick solid morphology. Furthermore, the analysis of nitrogen adsorption–desorption isotherm is presented in Figure S9, Supporting Information. The B-MoS₂ material exhibits a low specific surface area of 10.55 m² g⁻¹, indicating unfavorable conditions for sulfur loading and insufficient triple-phase contact. After the introduction of sulfur, the TG test (Figure S10, Supporting Information) confirms a sulfur content of 4.8% in the S/B-MoS₂ composite. However, this low level of sulfur is insufficient to achieve high capacity for the S/B-MoS₂ cathode. To aid the understanding of the concept, the schematic diagram is presented in Figure 6c. The point-to-point contact between the active material and solid electrolyte in ASSLSB is not as efficient as a liquid electrolyte which has superior mobility.^{[14, 48]} The challenge is to design an interface between the active material and solid electrolyte to ensure sufficient ion and electron supply.^[49] Many studies have already concluded that the sufficient triple-phase interface is heavily related to the particle size and morphology of the electrode materials.^[50-54] Thus, the nano-engineered electrode material is vital for better electrochemical performance. The structure of the S/MoS₂/C composite transfers from a ball-like few-layer structure to a nanosheet under pressure after the solid-state battery assembly. The nanosheet structure with interpolated distribution between S/MoS₂/C and solid electrolyte possesses the bigger total interface contact area between the active material and solid electrolyte acting as the fast “highways” for ions and electrons thus a better electronic/ionic conductivity. On the contrary, the thick massive structure of S/B-MoS₂ composite with low sulfur content is not conductive because of the long transport distance in the inner structure of active material, the more tortuous ionic/electronic pathways, and the limited interfacial contact area between the active material and solid electrolyte. Overall, the thin layer structure of S/MoS₂/C supports a sufficient triple-phase interface for faster kinetics and full solid chemical reaction, which provides a firm foundation for better electrochemical performances. Therefore, it is reasonable to choose the few-layered structure as a sulfur host. The rate performance of the S/MoS₂/C cathode with active loading of 1.3 mg cm⁻² is displayed in Figure 6d. It shows the discharge capacity of 984.3, 801.9, 720.3, 623.6, 612.8, 424.7, 367.4314.8, 280.1, and 242.0 mA h g⁻¹ at the current density of 100, 200, 300, 400, 500, 600, 700, 800, 900, and 1000 mA g⁻¹, respectively. When the rate returns to 200 mA g⁻¹, the reversible capacity recovers to 619.7 mA h g⁻¹, demonstrating good tolerance to reduplicative rate change. The discharge–charge profiles of the S/MoS₂/C cathode at higher current density (600 mA g⁻¹) are shown in Figure 6e. It is a normal phenomenon that the polarization decreases with the increasing cycle numbers indicating an activation process occurred.^[42] The cycling performance and Coulombic efficiency are recorded in Figure 6f. The S/MoS₂/C cathode displays stable cycling performance with high Coulombic efficiency approaching 98.5% and retains a discharge capacity of 448.5 mA h g⁻¹ after 3500 cycles. The discharge–charge profiles of the S/MoS₂/C cathode at a current density of 100 mA g⁻¹ with high active mass loading of 3.2 mg cm⁻² are also shown in Figure 6g. The S/MoS₂/C cathode shows discharge capacities of 733.1, 560.4, 574.0, 632.8, 696.0, and 734.1 mA h g⁻¹ at the 1st, 10th, 20th, 40th, 60th, 80th, and 100th cycle, respectively. The discharge capacity is also decreased first and then increased after 10 cycles (Figure 6h). It is because of the activation progress with high active loading at room temperature. It is reasonable to believe that the excellent behavior of the S/MoS₂/C cathode at room temperature is aided by the few-layered nanostructure with highly exposed surfaces and in situ-formed Li_xMoS₂ as mixed ion-electron conductor during discharging. The highly exposed few-layered surfaces and conductive Li_xMoS₂ provide a necessary condition for low-tortuous carrier transport pathways to activate S/Li₂S redox.

It is very useful to investigate the electrochemical performance of S/MoS₂/C cathode with higher active mass loading at 60 °C. The representative discharge–charge profiles and cycling performance curves of the S/MoS₂/C cathode with active loading of 6.2 mg cm⁻² (Figure 7a,b), 8.0 mg cm⁻² (Figure 7c,d), and 12.8 mg cm⁻² (Figure 7e,f) are presented, respectively. The cathode with active material loading of 6.2 mg cm⁻² delivers a similar first three reversible discharge capacities of ~1153 mAh g⁻¹. The cathode with active loading of 8.0 mg cm⁻² delivers a reversible discharge capacity of 887.9 mAh g⁻¹ at the first cycle and a similar reversible discharge capacity of 1005.8 mAh g⁻¹ at the second and third cycles. The cathode with active loading of 12.8 mg cm⁻² delivers a reversible discharge capacity of 426.1 mAh g⁻¹ at the first cycle, 754.5 mAh g⁻¹ at the second cycle, and 870.0 mAh g⁻¹ at the third cycle, respectively. The activation stage gradually evolved with the increased active material loading. As we all know, the polarization between discharge and charge plateaus is associated with the reaction's kinetics and electrochemical activity. The polarization increases with the increased active loading on the first cycle. In general, it is a normal phenomenon because of the increased transmission resistance. This increased active loading can be attributed to the lengthened electronic/ionic pathways in thick electrodes.^[14] The cathodes with active loading of 6.2, 8.0, and 12.8 mg cm⁻² display cycling performances with 98.9% Coulombic efficiency, and retain a discharge capacity of 1042.7 mAh g⁻¹ after 65 cycles, 896.9 mAh g⁻¹ after 65 cycles, and 839.4 mAh g⁻¹ after 16 cycles, respectively. Generally, the S/MoS₂/C cathode in ASSLSBs has achieved a high active loading of 12.8 mg cm⁻² and an areal capacity as high as 10.4 mAh cm⁻². The thickness of the cathode is 30, 51.8, 58.2, 89.0, and 123.0 μm, respectively (Figure S11, Supporting Information). Energy density calculations (gravimetric and volumetric) based on these measurements are provided in Table S2, Supporting Information. The cathode is measured in a mold cell with a stack pressure of 0.5 tons, so the volume is calculated from the thickness and cell area. The calculation equations are given in SI. The area capacity of the cathode and the minimizing electrolyte layer thickness are the key determinants of energy density. The SEM characterization of ASSLSBs with 12.8 mg cm⁻² active loading after failure is performed (Figure S12, Supporting Information). At the initial state, the surface and cross-section of the cathode show a smooth state, and there is no obvious delamination between it and the electrolyte layer. After failure, the surface of the cathode shows obvious cracking and delamination. Volume changes during the cycling of the cathode at high active loading lead to cracking and delamination, resulting in poor ion/electron transport and ASSLSBs failure. Finally, the cycle performance of the ASSLSBs that utilize Li as the anode is shown in Figure 7g. Such ASSLSB delivers an average reversible discharge of 924.8 mAh g⁻¹ and can retain 23 cycles.

With no additional carbon and a high percentage of active materials in the cathode, this work shows an excellent electrochemical performance in comparison with recently published works. Figure 7 summarizes the capacities based on the total weight of the cathode, maximum areal capacities, and the related cycle numbers in the recently published papers.^{[13, 15, 36, 42, 46, 47, 56, 57]} The areal capacity is the highest in our work. The achieved capacity based on the total weight of the cathode with the corresponding cycling numbers also matches some of the recent results in ASSLSBs studies.^[36] The impressive performance of the S/MoS₂/C cathode in ASSLSBs can be attributed to (i) the highly exposed MoS₂ surfaces with intimate area-to-area contact for better electronic/ionic conducting, (ii) the conversion & intercalation cathode for encompassing both lithium intercalation and metallic electronic conductivity, (iii) the in situ-formed Li_xMoS₂ during discharging for establishing low-tortuous carrier transport pathways.