High‐Performance Se–S Composite Cathode Rich in Defects for Wide‐Temperature Solid‐State Lithium Batteries

All‐solid‐state lithium batteries (ASSLBs) are a research hotspot for their superior safety. The solid electrolytes (SEs) are key components in ASSLBs, and the emerging rare‐earth halide SEs (RE‐HSEs) are valued for their comprehensive performances of good ionic conductivity, electrochemical stability, and deformability. In addition, cathode materials can influence the properties of ASSLBs, and sulfur (S) attracts much attention due to the lower toxicity and much higher energy density compared with commercial oxide cathodes. However, the S possesses poor electronic conductivity, which can be improved by the introduction of selenium (Se) with much higher electronic conductivity. In this work, a series of SexS1–x composites is synthesized by a melting method. Due to the introduction of Se and the enriched defects from the melting process, the electronic and ionic conductivities of SexS1–x are improved. After application in ASSLBs based on RE‐HSE Li3YBr6, the SexS1–x materials exhibit good performances with low polarizations, good cycling stabilities, and excellent rate properties at room temperature. Moreover, the assembled solid batteries can realize stable cycling performance (100 cycles) at low temperature (−30 °C) and a normal discharge–charge process at high temperature (120 °C).

Up to now, the researches on RE-HSEs applied in Li-S batteries were limited and they exhibited large discharge-charge polarizations and poor rate performances at a high temperature of 60 °C. [44]As a member of the same family of S, selenium (Se) exhibits superior electronic conductivity (Se: 1 Â 10 À5 S cm À1 ; S: 5 Â 10 À30 S cm À1 ) and similar chemical properties to S. [48] In addition, some works have selected Se material [46,48,49] as a cathode to realize high efficiency and good cycling and excellent rate performances at room temperature (RT); however, the theoretical capacity was decreased obviously (Se, 675 mAh g À1 ).Thus, Se can be introduced into the S material to form Se x S 1-x composites, which can exhibit high electronic/ionic conductivities and maintain large battery capacities. [50]In addition, the ASSLBs with RE-HSEs possess the promising applications over a wide-temperature range for the good ionic conductivity and thermal stability of RE-HSEs.
In this work, a series of Se x S 1-x (0 ≤ x ≤ 0.9) composites were prepared by melting method (380 °C).Benefiting from the introduction of Se and the melting process, the electronic/ionic conductivities of synthesized Se x S 1-x materials can be improved.Therefore, the Se x S 1-x exhibited low discharge-charge polarizations, and good cycling and rate performances with high capacities in ASSLBs based on Li 3 YBr 6 (LYB) electrolyte at RT.In addition, the assembled solid batteries can be operated over a wide-temperature range (À30-120 °C) for the high stability and ionic conductivity of LYB and good electronic/ionic conductivity of the Se x S 1-x cathode.Therefore, this optimization method for the S cathode can promote the practical application of ASSLBs with RE-HSEs.

Results and Discussions
In this work, the reported halide material LYB was selected as the SE.The LYB sample was obtained by a simple vacuum evaporation method reported in our previous works. [45]From the X-ray diffraction (XRD) pattern (Figure 1a), the obtained diffraction peaks of the synthesized LYB sample are well indexed with the PDF card (PDF#04-009-8861), indicating that LYB has been synthesized successfully.According to the X-ray photoelectron spectroscopy (XPS) results, the valence state of Li þ , Y 3þ , and inorganic Br À in LYB can be confirmed (Figure 1b).In addition, the LYB sample possesses micro-sized particles with uniform distribution and appropriate proportion of Y/Br elements from the scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS) results (Figure 1c-f, Figure S1, Supporting Information).To study the electrochemical properties of the synthesized LYB sample, electrochemical tests were performed on assembled LYB based cells.From the electrochemical impedance spectroscopy (EIS) test, the Nyquist plot of indium (In)/LYB/In cell can be obtained (Figure S2, Supporting Information), and the calculated ionic conductivity of LYB is 0.9 mS cm À1 .In addition, the carrier of the LYB sample can be confirmed as lithium ion according to the instantaneous current tests performed on Li/Li 7 P 3 S 11 (LPS)/LYB/LPS/Li and SG/LYB/SG (SG, stainlesssteel gasket) cells (Figure S3, Supporting Information).Thus, the synthesized LYB sample is suitable for the application of solid-state battery.Furthermore, the compatibility between LYB and anode was evaluated by discharge-charge tests on Li/LYB/Li and Li-In/LYB/Li-In cells (Figure S4, Supporting Information).The Li metal is unstable with the LYB sample, which induces large polarization in the Li/LYB/Li cell.Fortunately, the LYB has good stability with the Li-In alloy for the stable and small voltage polarization (%11 mV).Therefore, the Li-In alloy anode was adopted in the LYB-based ASSLB.
The large theoretical capacity and suitable operating voltage of the S cathode are beneficial to the performances of LYB-based solid Li-S batteries.However, the poor electronic conductivity of S hinders the practical application of solid Li-S batteries.To overcome this shortcoming of S, Se with high electronic conductivity was introduced into the S material by the melting method (380 °C), and a series of Se-S composite materials, Se x S 1-x (0 ≤ x ≤ 0.9), were prepared.According to the photos of the S-based materials (Figure 2a, Figure S5-S6, Supporting Information), the yellow commercial S powder turned black after the melting process.Along with the increase in the Se ratio, the color of Se x S 1-x gradually turned to brick red with increased viscidity.When x = 0.3, the Se x S 1-x sample was black and viscoelastic, which became hard after standing process and could be ground to brick red powder.When x = 0.5, the black viscoelastic sample with increased hardness can maintain its viscidity after standing for a long time.When x > 0.7, the obtained Se x S 1-x sample lost the viscidity and could be ground to black powder.After that, the structure of the Se x S 1-x sample was analyzed by XRD (Figure 2b, Figure S7-S8, Supporting Information).The collected XRD patterns of S and Se x S 1-x (x = 0) (named as Se x S 1-x -0) were similar and could be indexed to the PDF card (S, PDF#8-247) (Figure 1e).Thus, the melting process almost has no effect on the crystal phase of S sample.When x = 0.05, the XRD pattern of Se x S 1-x sample was similar with S and Se x S 1-x -0 with decreased intensity.Along with the continued increase in the x value, the structure of Se x S 1-x sample changed from the orthorhombic phase to the monoclinic phase (x = 0.1, PDF#41-1317 [S 27.64 Se 4.36 ]; x = 0.2 and 0.3, PDF#73-2267 [Se 2.57 S 5.43 ]) with further decreased crystallinity.When x = 0.3, the XRD signal almost disappeared, and could be recovered partially after standing process (Figure S2, Supporting Information).When x = 0.5 and 0.7, the XRD signal of the Se x S 1-x sample disappeared, indicating that the sample possessed an amorphous structure.When x = 0.9, the XRD peak of the material appeared and the structure of the Se x S 1-x sample turned to the orthorhombic phase (Se 5.1 S 1.9 , PDF#78-266).In addition, the morphology and constituent of Se x S 1-x samples were characterized (Figure 1c,d, Figure S9, Supporting Information).From the SEM/EDS results, the Se x S 1-x materials possess abnormal micro-sized particles with uniformly distributed elements of Se and S.Moreover, the molar ratios of Se/S were close to the theoretical value, and a small amount of S may be lost during the synthetic processes.
To evaluate the battery performances of Se x S 1-x composite materials, LYB-based ASSLBs were assembled by a simple cold-pressing process, containing Li-In anode, LYB electrolyte, and S-based cathode.The LYB-based electrolyte layer (thickness: %270 μm) could realize compact morphology due to the good deformability of LYB sample (Figure S10, Supporting Information).The cathode composite sample was obtained by mixing Se x S 1-x , LYB, and Ketjenblack (KB) through the ballmilling method.Then, the discharge-charge and rate performances of commercial S and prepared Se x S 1-x (0 ≤ x ≤ 0.9) materials were tested at RT. From the summarized results (Figure 3, Figure S11, Supporting Information), the prepared Se x S 1-x sample exhibited the better battery performances compared with commercial S cathode.In addition, along with the increase in the x value, the cycling and rate performances of Se x S 1-x cathodes were improved with the decreased capacities.It may be due to the decreased theoretical specific capacity and the improved electronic conductivity along with the addition of Se proportion in Se x S 1-x .In view of the comprehensive discharge-charge and rate performances of cathode materials, the representative commercial S, Se x S 1-x -0, and Se x S 1-x -0.3 (Se x S 1-x , x = 0.3) were selected for the further study.According to the cycling (Figure 4a), cyclic voltammetry (CV) (Figure 4b, Figure S12, Supporting Information), dischargecharge (Figure S11a,b,f, Supporting Information), and rate (0.1C-2.0C) (Figure 3a,b,f ) results of these assembled batteries, the trend of battery performance was S < Se x S 1-x -0 < Se x S 1-x -0.3, and the Se x S 1-x -0.3 cathode exhibited the smallest polarization, the most stable cycling performance, and the best rate performance.After that, the current density in the rate performance test of the Se x S 1-x -0.3 cathode was further improved   (Figure 4c).When cycled at 0.1C, 0.2C, 0.5C, 1.0C, 2.0C, 5.0C, and 10C, the reversible capacities of the Se x S 1-x -0.3 cathode were 922, 859, 788, 698, 495, 226, and 84 mAh g À1 , respectively.After the current density came back to 0.1C, the Se x S 1-x -0.3 cathode could maintain a capacity of 855 mAh g À1 .Meanwhile, the Se x S 1-x -0.3 cathode exhibited stable cycling performance (0.1C) over 250 cycles after rate performance (0.1C-2.0C) test at RT (Figure 4d).After cycling process, the close contact of electrolyte/cathode layer in ASSLB with Se x S 1-x -0.3 cathode could be maintained for the excellent deformability of LYB; the cathode surface became uneven for the volume change of Se x S 1-x -0.3 (Figure S13-S14, Supporting Information); the Li-In anode can maintain the flat surface after cycling due to the uniform deposition of lithium; the Li-In/LYB layer can keep intimate contact due to the good deformability of LYB and Li-In anode (Figure S15, Supporting Information).In addition, the Se x S 1-x -0.3 cathode can realize good cycling performances and normal discharge-charge curves at À30 and 120 °C (Figure 4e,f, Figure S16, Supporting Information) due to the good electronic conductivity of Se x S 1-x -0.3 and the high thermal stability of LYB with good ionic conductivity.The ionic conductivities of LYB at À30 and 120 °C are consistent with the corresponding discharge-charge polarizations of ASSLB with Se x S 1-x -0.3 cathode (Figure S17, Supporting Information).
Galvanostatic intermittent titration technique (GITT) tests were conducted on the LYB-based ASSLBs with different cathodes (S, Se x S 1-x -0, Se x S 1-x -0.3) at the same conditions (Figure 5a).According to the calculated Li þ -diffusion coefficient (D Li ) values (Figure 5b) and the aggregate voltage-polarization values (Figure 5c) during discharge-charge processes, the trend of D Li is S < Se x S 1-x -0 < Se x S 1-x -0.3, which is consistent with the trend of voltage polarization (Se x S 1-x -0.3 < Se x S 1-x -0 < S).Thus, the trend of electronic/ionic conductivity of these cathodes may be S < Se x S 1-x -0 < Se x S 1-x -0.3.In addition, EIS tests were conducted on these cells, and the discharge-charge impedances of the ASSLBs with S, Se x S 1-x -0, and Se x S 1-x -0.3 cathodes decreased successively (Figure S18, Supporting Information), corresponding to the GITT results.From the EIS spectra, the interfacial impedance is not obvious, indicating that LYB possesses good compatibility with the S-based cathode and that the electrolyte/cathode layer possesses close contact for the good deformability of LYB.
To evaluate the electronic/ionic conductivity of these S-based cathode materials, S, Se x S 1-x -0, Se x S 1-x -0.3, and Se were mixed with LYB by ball-milling method, and these composites were characterized by chronoamperometry and EIS tests.According to the obtained direct current polarization curves and Nyquist plots of these composites (Figure 5d-e), the Se/LYB sample exhibits the best electronic/ionic conductivity, indicating that the introduction of Se is beneficial for improving the electronic/ionic conductivities of Se x S 1-x materials.The Se x S 1-x -0/LYB sample possesses higher ionic conductivity than the S/LYB sample, and they have similar electronic conductivities; the Se x S 1-x -0.3/LYB sample exhibits higher electronic conductivity than the S/LYB and Se x S 1-x -0/LYB samples, and the ionic conductivity of the Se x S 1-x -0.3/LYB sample is close to that of the Se x S 1-x -0/LYB sample.Thus, for the Se x S 1-x materials, the improvement in battery performance mainly comes from the higher electronic/ionic conductivity of Se x S 1-x compared with the S sample.
To study the mechanism for the enhancement of ionic conductivity of the Se x S 1-x -0 sample, the structures of S and Se x S 1-x -0 were further analyzed by XRD Rietveld refinement according to the corresponding XRD results (Figure 6a,b).The obtained cell parameters of Se x S 1-x -0 are larger than those of the S sample (Table S1, Supporting Information), thus, the Se x S 1-x -0 sample may have more vacancy defects from the melting process, which is beneficial to electronic/ionic transport and can explain the color change from S (yellow) to Se x S 1-x -0 (black).The S-based samples (S, Se x S 1-x -0, Se x S 1-x -0.3, and Se) were further studied by XPS and Raman.According to the XPS results, the Se x S 1-x -0.3 can be confirmed as the composite of Se and S with zero valence, and the shifts to the lower binding energy of XPS peaks (S 2s and S 2p) of S sample after melting process (Se x S 1-x -0) may be attributed to the addition of defects (Figure 6c,d). [50]From the Raman results, the new Raman peak (354.6 cm À1 ) of Se x S 1-x -0.3 sample may be assigned to the stretching vibration of Se-S bond (Figure 6e). [51]In addition, the S and Se x S 1-x -0 samples possess the similar Raman spectra, and some Raman peaks of Se x S 1-x -0 exhibit slight deviation to the left, which can demonstrate the more abundant defects of Se x S 1-x -0 sample (Figure 6f ). [52]In short, the Se x S 1-x materials possess better battery performances compared with the S sample, which may be due to the improvement in electronic/ionic conductivity after the introduction of Se and the increase in vacancy defects after the melting process.

Conclusion
High performance of LYB-based solid batteries can be realized by the design of S-based cathode.To improve the electronic/ionic conductivity of S, a series of Se x S 1-x (0 ≤ x ≤ 0.9) composites were prepared, and the sample Se x S 1-x -0.3 was selected for the comprehensive advantages of good cycling performance (250 cycles), high capacity, and excellent rate performance (0.1-10 C) at RT.The high performance of Se x S 1-x -0.3 cathode

Figure 4 .
Figure 4.The performances of LYB-based solid batteries (Li-In/LYB/Se x S 1-x ).a) Cycling stability of ASSLBs with S, Se x S 1-x -0 and Se x S 1-x -0.3 cathodes at 0.1 C (RT).b) CV curves of ASSLB with Se x S 1-x -0.3 cathode.c) Rate performance (RT) of ASSLB with Se x S 1-x -0.3 cathode at different current densities (0.1-10C).d) Rate and cycling performance of ASSLB with Se x S 1-x -0.3 cathode at RT. e,f ) Cycling performance of ASSLB with Se x S 1-x -0.3 cathode at À30 °C (e) and 120 °C (f ).

Figure 5 .
Figure 5. a) Galvanostatic intermittent titration technique (GITT) curves, b) D Li values, and c) the summarized voltage polarizations of ASSLBs (Li-In/LYB/Se x S 1-x ) with S, Se x S 1-x -0 and Se x S 1-x -0.3 cathodes.d) Response currents at an applied bias (1.0 V) and e) electrochemical impedance spectroscopy results of LYB-based composites mixed with S, Se x S 1-x -0, Se x S 1-x -0.3, and Se, respectively.

Figure 6 .
Figure 6.a,b) Rietveld refinements of XRD patterns of S (a) and Se x S 1-x -0 (b).c,d) The XPS spectra and e,f ) Raman spectra of S, Se x S 1-x -0, Se x S 1-x -0.3, and Se.