Solid-state batteries

Abstract

The present invention provides rechargeable batteries comprising a solid electrolyte (SSE) containing an alkali metal disposed between two electrodes. The battery is volumetrically constrained, thereby imparting increased stability under voltage cycling conditions, such as by mechanical compression of the microstructure of the solid electrolyte and the electrolyte-electrode interface. These batteries of the present invention are advantageous because they can be all-solid-state batteries, e.g., do not require liquid electrolytes, and can achieve higher voltages with minimal electrolyte degradation.

Description

Field of the Invention

[0001] The present invention relates to the field of solid-state batteries having an alkali metal sulfide solid-state electrolyte. Background Art

[0002] Solid-state lithium-ion conductors, a key component that enables all solid-state lithium-ion batteries, are one of the most sought-after research goals in the battery field. The strong interest in solid-state electrolytes and more generally in solid-state batteries stems primarily from improved safety, the ability to enable new electrode materials, and better low-temperature performance. Since the liquid electrolytes currently used are typically highly flammable organic solvents, safety improvements in solid-state battery cells are expected. Replacing these electrolytes with non-flammable solids will eliminate the most problematic aspects of battery safety. In addition, solid-state electrolytes are compatible with several high-energy density electrode materials that cannot be implemented with liquid electrolyte-based configurations. Solid-state electrolytes also maintain better low-temperature operation than liquid electrolytes, which experience significant ionic conductivity drops at low temperatures. This low-temperature performance is critical in the rapidly growing electric vehicle market.

[0003] Among the solid-state electrolytes currently being studied, sulfides remain one of the highest performing and most promising families. Sulfide glass solid-state electrolytes and glass-ceramic solid-state electrolytes (in which a crystalline phase has been precipitated within a glass matrix) have demonstrated 0.1–1 mS cm -1 Order of magnitude and above 1mS cm -1 Ceramic-sulfide electrolytes, most notably Li 10 GeP2S 12 (LGPS) and Li 10 SiP2S 12 (LSPS) are particularly promising because they maintain extremely high ionic conductivity. LGPS is one of the first solid-state electrolytes to achieve an ionic conductivity of 12 mS cm, comparable to that of liquid electrolytes. -1 , only to be replaced by LSPS, which achieved 25 mS cm -1 Despite these promising conductivities, the ceramic-sulfide family suffers from a narrow stability window. That is, both LGPS and LSPS tend to reduce at voltages below about 1.7 V or oxidize above about 2.1 V relative to lithium metal. This limited stability window has proven to be a major obstacle for battery cells that need to operate in a voltage range of about 0-4 V.

[0004] Therefore, there is a need for an improved solid-state battery that incorporates a solid electrolyte with controllable structural properties and surface chemistry. Summary of the invention

[0005] We have developed a rechargeable solid-state battery using a solid electrolyte with improved cycling performance. The rechargeable solid-state battery disclosed herein is advantageous because the solid electrolyte has superior voltage stability and excellent battery cycling performance. The battery of the present invention can be stabilized against the formation of lithium dendrites and / or can operate at high current density for extended cycles.

[0006] In one aspect, the invention features a rechargeable battery comprising a first electrode, a second electrode and a solid electrolyte disposed therebetween. The solid electrolyte comprises a sulfide comprising an alkali metal, such as lithium. In certain embodiments, the solid electrolyte is under a volume constraint sufficient to stabilize the solid electrolyte during electrochemical cycling. In a particular embodiment, the volume constraint applies a pressure of about 70 MPa to about 1,000 MPa, such as about 100-250 MPa, to the solid electrolyte, such as applying a mechanical compression of the order of 15 GPa to the microstructure of the solid electrolyte. In certain embodiments, the volume constraint provides a voltage stability window of 1 V to 10 V, such as 1-8 V, 5.0-8 V, or greater than 5.7 V or even greater than 10 V.

[0007] In some embodiments, the solid electrolyte has a core-shell morphology. In certain embodiments, the alkali metal is Li, Na, K, Rb, or Cs, such as Li. In some embodiments, the solid electrolyte comprises SiPS, GePS, SnPS, PSI, or PS. In some embodiments, the solid electrolyte is Li 10 SiP2S 12 , Li 10 GeP2S 12 or Li 9.54 Si 1.74 P 1.44 S 11.7 Cl 0.3 In some embodiments, the first electrode is a cathode, which may include LiCoO2, LiNi 0.5 Mn 1.5 O4, V Li2CoPO4F, LiNiPO4, Li2Ni(PO4)F, LiMnF4, LiFeF4 or LiCo 0.5 Mn 1.5 In certain embodiments, the second electrode is an anode and may include lithium metal, lithiated graphite, or Li4Ti5O 12 In certain embodiments, the volume confinement provides a mechanical compression of the solid electrolyte of about 1 GPa to about 100 GPa, such as about 15 GPa.

[0008] In another aspect, the invention features a rechargeable battery comprising a first electrode, a second electrode, and a solid electrolyte disposed therebetween, wherein the second electrode is an anode comprising an alkali metal and graphite. In some embodiments, the battery is at a pressure of about 70-1000 MPa, such as about 100-250 MPa. In specific embodiments, the alkali metal and graphite form a composite material. In some embodiments, the alkali metal is Li, Na, K, Rb, or Cs, such as Li. In some embodiments, the solid electrolyte comprises SiPS, GePS, SnPS, PSI, or PS. In certain embodiments, the solid electrolyte is Li 10 SiP2S 12 , Li 10 GeP2S 12 or Li 9.54 Si 1.74 P 1.44 S 11.7 Cl 0.3 In certain embodiments, the first electrode is a cathode and may include LiCoO2, LiNi 0.5 Mn 1.5 O4, VLi2CoPO4F, LiNiPO4, Li2Ni(PO4)F, LiMnF4, LiFeF4 or LiCo 0.5 Mn 1.5 In some embodiments, the cell is under an external stress that provides mechanical compression of the solid electrolyte of about 1 GPa to about 100 GPa, such as about 15 GPa.

[0009] In another aspect, the invention features a rechargeable battery comprising a first electrode, a second electrode, and a solid electrolyte disposed therebetween, wherein the solid electrolyte may include a sulfide comprising an alkali metal; and the battery is under isochoric constraint. In some embodiments, the isochoric constraint is provided by compressing the solid electrolyte at a pressure of about 3-1000 MPa, such as about 100-250 MPa. In certain embodiments, the alkali metal is Li, Na, K, Rb, or Cs, such as Li. In some embodiments, the solid electrolyte comprises SiPS, GePS, SnPS, PSI, or PS. In certain embodiments, the solid electrolyte is Li 10 SiP2S 12 , Li 10 GeP2S 12 or Li 9.54 Si 1.74 P 1.44 S 11.7 Cl 0.3 In certain embodiments, the first electrode is a cathode and may include LiCoO2, LiNi 0.5Mn 1.5 O4, V Li2CoPO4F, LiNiPO4, Li2Ni(PO4)F, LiMnF4, LiFeF4 or LiCo 0.5 Mn 1.5 In some embodiments, the isochoric confinement provides a mechanical compression of the solid electrolyte of about 1 GPa to about 100 GPa, such as about 15 GPa.

[0010] In another aspect, the invention features a rechargeable battery having a first electrode, a second electrode, and a solid electrolyte disposed therebetween. The solid electrolyte comprises a sulfide comprising an alkali metal and optionally having a core-shell morphology. The first electrode comprises an interface-stabilizing coating material. In certain embodiments, the first electrode and the second electrode independently comprise an interface-stabilizing coating material. In certain embodiments, one of the first electrode and the second electrode comprises a lithium-graphite composite material.

[0011] In some embodiments, the first electrode comprises a material as described herein, e.g., in Table 1. In some embodiments, the coating material of the first electrode is a coating material as described herein, e.g., LiNbO3, AlF3, MgF2, Al2O3, SiO2, graphite, or a coating material as described in Table 2. In certain embodiments, the alkali metal is Li, Na, K, Rb, or Cs, e.g., Li. In some embodiments, the solid electrolyte comprises SiPS, GePS, SnPS, PSI, or PS. In certain embodiments, the solid electrolyte is Li 10 SiP2S 12 , Li 10 GeP2S 12 or Li 9.54 Si 1.74 P 1.44 S 11.7 Cl 0.3 In some embodiments, the first electrode is a cathode and may include LiC o O2、LiNi 0.5 Mn 1.5 O4, V Li2CoPO4F, LiNiPO4, Li2Ni(PO4)F, LiMnF4, LiFeF4 or LiCo 0.5 Mn 1.5 In some embodiments, the battery is under an external stress that provides mechanical compression of the solid electrolyte of about 1 GPa to about 100 GPa, such as about 15 GPa. In certain embodiments, the battery is under a pressure of about 70-1000 MPa, such as about 100-250 MPa.

[0012] In another aspect, the invention features a method of storing energy by applying a voltage across a first electrode and a second electrode and charging a rechargeable battery of the invention. In another aspect, the invention provides a method of providing energy by connecting a load to a first electrode and a second electrode and allowing a rechargeable battery of the invention to discharge. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1A-1B : Cyclic voltammetry (CV) tests of LGPS in liquid (A) and solid (B) states at different pressures. LGPS / C films with a ratio of 90:10 were tested in liquid electrolyte (black curve in (A)). CV tests were also performed by replacing the liquid electrolyte with LGPS pellets at different pressures, which is an all-solid-state CV. The decomposition intensity decreases significantly with increasing applied pressure. At a reasonably low pressure of 6T (420MPa), there is no obvious decomposition peak before 5.7V (purple curve), indicating that applying external pressure or volume compression at the cell level can broaden the electrochemical window of solid electrolytes.

[0014] Figure 2A-2B :LiCoO2(LCO)-Li4Ti5O 12 (LTO) Capacity (A) and cycle performance (B) of all-solid-state full-battery. Since the chemical potential of LTO is 1.5 V (relative to Li), the working platform on the cathode side is higher than 4 V (relative to Li).

[0015] Figure 3A-3B :LiNi 0.5 Mn 1.5 Capacity (A) and cycle performance (B) of O4(LNMO)-LTO all-solid-state full battery. Since the chemical potential of LTO is 1.5 V (relative to Li), the working platform on the cathode side is higher than 4.7 V (relative to Li).

[0016] Figure 4 : High voltage cathode candidates for 6V and higher all-solid-state lithium-ion battery technology. The legend is marked as: F is fluoride, O is oxide, P, O is phosphate, and S, O is sulfate. A complete list of these high voltage fluorides, oxides, phosphates, and sulfates is provided in Table 1. Commercially available LiCoO2 (LCO) and LMNO are marked as stars.

[0017] Figure 5A-5B : (A) Graphical representation of the effect of strain on the decomposition of LGPS, where x D is the fraction of LGPS that has decomposed. The lower dashed line shows the Gibbs energy (G) of the binary combination of the original LGPS and any set of decay products (D) when negligible pressure is applied. o (x D))(isobaric decay, p≈0GPα). The solid line shows the Gibbs when mechanical constraints are applied to the LGPS. Since the LGPS tends to expand when it decomposes, the strain Gibbs (G 应变 ) increases. At some rupture point, represented by x f , the Gibbs energy of the system exceeds the energy required to break the mechanical confinement (upper dashed line). The highlighted path is the proposed ground state of the mechanically confined LGPS system. Then the area x D <x f is metastable. (B) Schematic representation of the work differential in the case of a "fluid-like" system and a "solid-like" system. For the "fluid-like" system on the top, the system undergoes internal volume expansion ("stress-free" strain) due to decomposition rather than applied stress. The bottom system represents elastic deformation away from an arbitrary reference state.

[0018] Figure 6 :LGPS and LGPSO in the mean field limit (Li 10 GeP2S 11.5 O 0.5 ) stable window. 壳 =V 核 -1 θ p V 核 Indicates the rigidity of the constraint mechanism. Limit β 壳 →0 and β 壳 →∞ represents the isochoric limit and isobaric limit. Under isobaric conditions, the inherent material stability is restored (~1.7-2.1V).

[0019] Figure 7A-7B :(A) Illustration of the nucleation decay mechanism. Radius R o The original LGPS particle has a radius R at its center i In the absence of stress, the radius of the decomposed region is now R d , which must be squeezed into R i The final nucleated particle (iv) is generated, where the strain is not zero. (B) The difference between the hydrostatic / mean field model and the nucleation model The unit is KV. For a typical Poisson's ratio, it can be seen that the strain term is similar to the ideal core-shell model (β 壳 =0) is equal to or better than.

[0020] Figures 8A-8E :Voltage (φ), lithium chemical potential (μ Li +) and the Fermi level (ε f) distribution. (A) Conventional cell design. (B) Conventional cell with hybrid solid electrolyte / active material cathode. I Figure 2. Interfacial voltages formed between active materials and solid electrolytes due to different lithium ion chemical potentials. (C) Illustration of previous speculation on how insulating layers can lead to variable lithium metal chemical potentials within the cell. (D) How the voltage from part (C) is expected to relax given efficient electron conduction due to lithium hole migration. (E) Results from part (D) once the applied voltage exceeds the intrinsic stability window of the solid electrolyte. It can be seen that local lithium is formed within the insulating region, where the interfacial voltage (χ I ) is equal to the applied voltage.

[0021] Figures 9A-9D : Comparison between the microstructure and chemical composition of LGPS and super-LGPS particles. (A, C) Typical TEM bright field images of LGPS and super-LGPS particles, respectively, showing the obvious surface layer of super-LGPS particles. (B, D) STEM EDS line scans for statistical analysis of various LGPS and super-LGPS particles of different sizes, showing that the sulfur concentration of LGPS particles is uniformly distributed from the surface to the bulk, while the sulfur concentration of super-LGPS decreases in the surface layer.

[0022] Fig.10 : STEM EDS line scans across a single LGPS particle with different particle sizes ranging from 100 nm to 3 μm, showing no regular pattern in the variation of sulfur concentration from surface to bulk.

[0023] Fig.11 : STEM EDS line scans through individual LGPS particles with different particle sizes ranging from 60 nm to 4 μm sonicated in dimethyl carbonate (DMC) for 70 hours, showing that the sulfur concentration in the surface region is significantly smaller compared to the sulfur concentration in the bulk.

[0024] Fig.12 : STEM EDS line scans through individual LGPS particles with different particle sizes ranging from 120 nm to 4 μm sonicated in diethyl carbonate (DEC) for 70 hours, showing that the sulfur concentration in the surface region is significantly smaller compared to the sulfur concentration in the bulk.

[0025] Fig.13 : Quantitative STEM EDX analysis of LGPS particles before and after ultrasonic preparation showed that the surface / bulk ratio of S was significantly lower after ultrasonic treatment in organic electrolytes (DEC and DMC).

[0026] Fig.14: STEM EDS line scans across individual LGPS particles with different particle sizes ranging from 160 nm to 3 μm immersed in DMC for 70 h without ultrasonic treatment, showing no regular pattern in the variation of sulfur concentration from surface to bulk.

[0027] Figures 15A-15H : Comparison of electrochemical performance of LGPS and super-LGPS particles, and LIBs made of LGPS and super-LGPS particles. (A, B) Cyclic voltammograms (CV) of Li / LGPS / LGPS+C / Ta unit and Li / super-LGPS / super-LGPS / Ta unit, respectively, with a lithium reference electrode and a scan rate of 0.1 mVs -1 and the scan range is 0.5V to 5V. (C, D) Sensitive electrochemical impedance spectroscopy (EIS) of the LGPS cell and the super-LGPS cell in panels (A, B) before and after CV test. (E, F) Charge-discharge profiles of LGPS-LIB (LTO+LGPS+C / glass fiber separator / Li) and super-LGPS-LIB (LTO+super-LGPS+C / glass fiber separator / Li) cycled in the voltage range of 1.0-2.2V at a current rate of 0.5C. (G, H) Cycling capacity curves of LGPS LIB and super-LGPS-LIB.

[0028] Figures 16A-16B : Cycling performance of (A) LGPS-ASSLIB (LTO+LGPS+C as cathode, LGPS as solid electrolyte, and Li as anode) and (B) Super LGPS-ASSL1B (LTO+Super LGPS+C as cathode, Super LGPS as solid electrolyte, and Li as anode) at low current rate (0.02C).

[0029] Figures 17A-17B : Cycling performances of (A) LGPS-ASSLIB (LTO+LGPS+C as cathode, LGPS as solid electrolyte, and Li as anode) and (B) super-LGPS-ASSLIB (LTO+super-LGPS+C as cathode, super-LGPS as solid electrolyte, and Li as anode) at medium current rate (0.1C).

[0030] Figures 18A-18B : Cycling performance of (A) LGPS-ASSLIB (LTO+LGPS+C as cathode, LGPS as solid electrolyte, and Li as anode) and (B) super-LGPS-ASSLIB (LTO+super-LGPS+C as cathode, super-LGPS as solid electrolyte, and Li as anode) at high current rate (0.8C).

[0031] Figures 19A-19G: Microstructure and composition of the LTO / LGPS interface after cycling in LGPS ASSLIB (S)TEM study. (A) FIB sample prepared from LGPS ASSLIB after 1 charge-discharge cycle, including the cathode layer (LTO+LGPS+C) and the SE layer (LGPS). (B) TEM BF image of the LTO / LGPS primary interface, showing a transition layer with multiple dark particles. (C) HRTEM image of LTO particles and their corresponding FFT patterns. (D) STEM DF image of the LTO / LGPS primary interface showing ultrabright particles within the transition layer, indicating the accumulation of heavy elements. (E) STEM EELS line scan across the primary interface, indicating that the bright particles within the transition layer are sulfur-rich. (F) STEM DF image of the LTO / LGPS secondary interface, again showing a higher density of bright particles with similar morphology. (G) STEM EELS line scan across the secondary interface, indicating that the bright particles are sulfur-rich.

[0032] Fig. 20 : TEM bright field image and STEM dark field image of the main LTO / LGPS interface (the interface between the cathode and LGPS solid electrolyte layer) of LGPS-ASSLIB (LTO+LGPS+C as cathode, LGPS as solid electrolyte, and Li as anode), showing a clear transition layer between the cathode and solid electrolyte layer.

[0033] Figures 21A-21B Figure 3: (A) STEM dark field image and (B) EELS line scan of the main LTO / LGPS interface (interface between cathode and LGPS solid electrolyte layer) of LGPS-ASSLIB (LTO+LGPS+C as cathode, LGPS as solid electrolyte, and Li as anode), showing the presence of Li for regions both inside and outside of the bright particles within the transition layer K and Ge M4，5 peak.

[0034] Figures 22A-22B : (A) STEM dark field image and (B) EELS line scan of the main LTO / LGPS interface (the interface between the cathode and LGPS solid electrolyte layers) of LGPS-ASSLIB (LTO+LGPS+C as cathode, LGPS as solid electrolyte, and Li as anode), showing S L1 The peak intensity is stronger on those S-rich particles within the transition layer.

[0035] Figures 23A-23F: Microstructure and composition of LTO / super-LGPS interface after cycling in super-LGPS ASSLIB (S)TEM study. (A) TEM BF image of the LTO / super-LGPS main interface, showing no Figure 6 The smooth interface of the dark particles present in B. (B) corresponds to Fig.23A (C) STEM DF image of LTO / super LGPS subinterface. (D) STEM EDS line scan shows that the sulfur atomic percentage decreases continuously from the inner super LGPS particle to the sub-LTO / super LGPS interface and finally to the LTO+C composite region. (E) STEM EDS mapping shows that the large particles in Figure 22C are LGPS particles. (F) STEM EDS quantitative analysis shows that the sulfur atomic percentage inside the super LGPS particle is as high as ~38%, while the sulfur atomic percentage at the sub-LTO / super LGPS interface is as low as 8%.

[0036] Figures 24A-24B : Additional (A) STEM dark field images and (B) STEM EDX line scans show that the S concentration at the sub-LTO / super-LGPS interface is much lower than that in the inner super-LGPS particle region.

[0037] Figures 25A-25C : (A) The number of hulls required to evaluate the stability of 67k materials if the evaluation scheme is material iteration (left column) or element set iteration (right column). (B) Illustration of the pseudo-binary method to achieve interface stability between LSPS and an arbitrary material A. represents the material-level decomposition energy that exists even in the absence of an interface, and G′ hull indicates the increased instability due to the presence of the interface. m The strongest kinetically driven reaction occurs when A and D LspS are the decomposed coating material and LSPS in the absence of an interface (e.g., at x = 0, 1). (C) Correlation of element fraction with increased chemical interface instability (G′ hull (x m )). Negative values ​​make G' decrease with increasing concentration. hull and improves the stability of the interface. On the contrary, positive values ​​tend to increase G′ hull and those atomic species that deteriorate interface stability. Elements that exist in less than 50 crystal structures are grayed out due to lack of bulk data.

[0038] Figures 26A-26C :(AC) Correlation of elemental species fraction with increased electrochemical interfacial stability at 0 V, 2 V and 4 V, respectively (G′ hull(x m )). Negative values ​​indicate that increasing concentration decreases G′ hull and improves the interfacial stability. Conversely, positive values ​​tend to increase G′ hull and those species that deteriorate interface stability. Elements that exist in less than 50 crystal structures are grayed out due to lack of bulk data.

[0039] Figures 27A-27D : (A) Hull energy of LSPS versus voltage relative to lithium metal. Darker gray [medium gray] shading highlights where decomposition occurs into oxidative [reductive] states. Light gray shading indicates the region where LSPS decays to the point where it does not consume or produce lithium (e.g., lithium neutral). The oxidative [reductive] region is characterized by an increase [decrease in hull energy with increasing voltage. (B) and (C) Hull energies at boundary voltages for the anodic and cathodic ranges, respectively, for anionic species (e.g., oxygen-containing compounds versus sulfur-containing compounds, etc.). Data points above [below] the neutral decay line are net oxidative [reductive] states in the anodic range / cathodic range. Those compounds on the neutral decay line decay without reacting with the lithium ion reservoir. (D) Average hull energy of electrochemical decomposition at the material level versus voltage.

[0040] Figures 28A-28C : Comparison of the average LSPS interfacial stability of compounds sorted by anionic species. (A) Average total maximum kinetic driving energy (G) for the chemical reaction between LSPS and each of the considered anionic classes. hull (x m )) and the contribution from the interface (G′ hull (x m (B) The total electrochemical instability of each anion class at a given voltage (G hull (x m (C) Average contribution of the interface to the electrochemical instability of each anion class at a given voltage (G′ hull (x m )).

[0041] Figures 29A-29B : Functional stability results for compounds sorted by anionic species. (A) and (B) Total number (line) and percentage (bar) of each anionic class determined to be functionally stable. The bottom bar represents the percentage of functionally stable material, and the top bar represents the percentage of potentially functionally stable material based on the reversibility of lithiation / delithiation.

[0042] Figures 30A-30F: (AD) Comparison of XRD patterns showing structural decay of LCO, SnO2, LTO and SiO2 at the solid electrolyte material interface (no applied voltage). ●、■、 They represent LCO (PDF#44-0145), LSPS (ICSD#252037), SiO2 (PDF#48-0476), Li3PO4 (PDF#45-0747), cubic Co4S3 (PDF#02-1338), and monoclinic Co4S3 (PDF#02-1458). ●、■、 They represent SnO2(PDF#41-1445), LSPS(ICSD#252037), SiO2(PDF#34-1382), P2S5(PDF#50-0813) and Li2S(PDF#23-0369) respectively. They represent LTO (PDF#49-0207), LSPS (ICSD#252037) and Li 1.95 Ti 2.05 S4(PDF#40-0878). In (D), ▲, Represents SiO2 (PDF#27-0605) and LSPS (ICSD#252037), respectively. The shaded area in (AD) highlights where a significant phase transition occurs after heating to 500°C. The interfacial chemical compatibility decreases from (A) to (D), which corresponds well to the predicted interface decays of 200 meV / atom, 97 meV / atom, 75 meV / atom, and 0 meV / atom for LCO, SnO2, LTO, and SiO2, respectively. (E, F) CV results for Li2S and SiO2. The shaded area predicts whether the curves in this region will be predominantly oxidizing, reducing, or neutral.

[0043] Figures 31A-31E :Each single phase at room temperature and 500℃: (A) LiCoO2, (B) LSPS, (C) Li4Ti5O 12 Comparison of XRD patterns of (A) SnO2 and (D) SiO2. For each phase, no significant changes can be observed between room temperature and 500 °C.

[0044] Figures 32A-32D : Mixture powders at various temperatures (room temperature, 300°C, 400°C and 500°C): (A) LiCoO2+LSPS, (B) SnO2+LSPS, (C) Li4Ti5O 12 Comparison of XRD patterns of LiCoO2+LSPS, SnO2+LSPS and Li4Ti5O 12+LSPS, the onset reaction temperatures were observed to be 500°C, 400°C and 500°C, respectively. At up to 500°C, no reaction of SiO2 + LSPS was observed.

[0045] Figures 33A-33F .(A, B, C) XRD of different powder mixtures before and after heat treatment at 500 °C for 36 h ((A) Li+LGPS; (B) graphite+LGPS; (C) lithiated graphite+LGPS). The symbols and corresponding phases are: ※LGPS; +Li; *graphite; ×LiS2; GeS2; ☆GeLi5P3. (D) Structure of Li / graphite anode in LGPS-based all-solid-state battery; (E) SEM image of the cross section of Li / graphite anode; (F) FIB-SEM of the interface between Li and graphite.

[0046] Fig.34A -34E. (A) Comparison of cycling performance between Li / G-LGPS-G / Li symmetric battery and Li-LGPS-Li symmetric battery; (B) SEM image of symmetric battery after cycling. Li / G-LGPS-G / Li symmetric battery (B1, 2) after 300 hours of cycling and Li-LGPS-Li symmetric battery (B3, 4) after 10 hours of cycling; (C) Rate performance of Li / G-LGPS-G / Li symmetric battery at different pressures. (D) SEM images of Li / G-LGPS-G / Li symmetric battery at different pressures after rate testing. (E) Up to 10 mA / cm 2 Ultra-high rate performance. The applied pressure in (E) is 250 MPa. The inset is the cycle distribution plotted in the range of -0.3 V to 0.3 V, indicating that there is no obvious overvoltage change after high-rate cycling. More voltage distribution zoom-in is shown in Supplementary Information Figure 42.

[0047] Figures 35A-35D .Comparison of (A) initial charge / discharge curves, (B) initial Coulombic efficiency and (C) open circuit voltage at different capacity ratios of Li to graphite in the Li / G-LGPS-LCO (LiNbO3 coated) system after 1 hour rest. The Li / G capacity ratios of 0, 0.5, 0.8, 1.5, 2.5 and 4 can be converted to Li / G thickness ratios of about 0, 0.3, 0.4, 0.8, 1.3 and 2.1, respectively. The Li / graphite thickness ratio is assumed to be 1.0-1.3 in this work without specific specification. (D) Cycling performance of Li / G-LGPS-LCO (LiNbO3 coated) battery.

[0048] Figures 36A-36B .(A) In different effective modules (Keff ) The voltage distribution of LGPS decomposition under different K eff The reduction reaction pathways and the products in the different phase equilibria in each voltage range. All decomposition products are here the ground state phase in each voltage range.

[0049] Figure 37A-37F. At 0.25 mA cm -2 XPS measurements of Ge and P of the anode-LGPS-anode symmetric cell at 100 MPa after 12 h of down-cycling with an X-ray beam focused on the (A) central part of the LGPS in the Li / G-LGPS-G / Li unit away from the interface with Li / G and (B) on the interface between Li / G and LGPS; (C) at 0.25 mA cm -2 XPS measurement of Ge and P of anode-LGPS-anode symmetric cell at 100 MPa with an X-ray beam focused on the interface between Li and LGPS in Li-LGPS-Li symmetric cell after 10 h of cycling (failure); (D) XPS measurement of Ge and P of anode-LGPS-anode symmetric cell at 100 MPa with an X-ray beam focused on the interface between Li and LGPS in Li-LGPS-Li symmetric cell (failure); (D) XPS measurement of Ge and P of anode-LGPS-anode symmetric cell at 100 MPa at 2 mA cm -2 XPS measurements of Ge and P in the anode-LGPS-anode symmetric cell with an X-ray beam focused on the Li / G-LGPS interface after the low rate test; and (E) at 250 MPa at 10 mA cm -2 XPS measurement of Ge and P in the anode-LGPS-anode symmetric cell using an X-ray beam focused on the Li / G-LGPS interface after the low-rate test; (F) at 3 MPa and 2 mA cm -2 After the low-rate test, the Ge and P XPS measurements of the anode-LGPS-anode symmetric cell were performed using an X-ray beam focused on the Li / G-LGPS interface.

[0050] Fig.38 XRD of graphite and mixture of Li and graphite after heating at 500 °C for 36 h.

[0051] Figures 39A-39C .(A) SEM image of graphite particles; SEM images of the surface (B) and cross-section (C) of the graphite film after application of high pressure.

[0052] Fig.40 .Cycling performance of Li / G-LGPS-G / Li symmetric cells with relatively small overvoltage.

[0053] Figures 41A-44B .Comparison of SEM images of Li / G anode before (A) and after (B) long-term cycling in Figure 34 (A).

[0054] Figure 42A-42C. (A) Rate test of Li / G-LGPS-G / Li symmetric battery. When the pre-cycle number is 0.25 mA cm -2 When the battery is reduced to 5 cycles, the -2 or 7mA cm -2 Under “failure”, however, when the current density was set back to 0.25 mA cm -2 When the overvoltage is increased, it always returns to normal without significant overvoltage increase. (B) Enlarged view of Figure 34 (E2) at 10 mA cm -2 Cycled cells, plotted at a smaller voltage scale (B1) or time scale (B2). (C) SEM images of the Li / graphite composite after the test shown in B at different areas and magnifications. No lithium dendrites were observed. A clear 3D structure showing this is in Figure 42 (C2).

[0055] Figures 43A-43B .(A) Fig.35D (A) Cycling profile of LCO-LGPS-Li / G cells. (B) Cycling performance based on Li anode. Both cells were tested at 25 °C at a current density of 0.1C.

[0056] Figures 44A-44B Bader charge analysis from DFT simulation. (A) Phosphorus in all P-related compounds from the decomposition product list; (B) Ge in all Ge-related compounds from the decomposition product list.

[0057] Figures 45A-45D (A) Comparison of CV curves of Li / G-LGPS-LGPS / C cells tested at 3MPa and 100MPa; (B, C) Comparison of impedance changes before and after these two CV tests; (D) Model used in impedance fitting. 本体 represents the ion diffusion resistance, and R ct represents the charge transfer resistance. All EIS data were fitted using Z-view.

[0058] Figure 46A-46G. (A) CV test of Swagelok cells after pressing with 1T, 3T, 6T and the pressed cell initially pressed with 6T. 10% carbon was added to the cathode. The voltage range was set from open circuit to 9.8V. (B) CV scan in (A) plotted with enlarged voltage and current range. (C) In-situ impedance test of the cell shown in (A) during the CV scan. (D) Synchrotron XRD of the pressed cell after no electrochemical process (black), CV scan to 3.2V, 7.5V and 9.8V. After all CVs, the voltage was kept at the same high cutoff voltage for 10 hours and then discharged back to 2.5V. Green line: Synchrotron XRD of LGPS tested in liquid electrolyte after CV scan to 3.2V and hold for 10 hours. (E) Synchrotron XRD peaks of different cells at 2θ=18.5°, showing the broadening of the XRD peaks after high voltage CV scan and hold. (F) Strain versus size broadening analysis of LGPS after high voltage hold. Points are the broadening of different peaks in 7.5V SXRD measurement, the corresponding XRD peaks are shown in Figure 52. The angular dependence of size and strain broadening is indicated by the dashed lines. (G) XAS measurements of S (g1) and P (g2) after high voltage CV scan and hold. (g3) Simulation of P XAS peak shift after c-direction strain.

[0059] Figure 47A-47D. (A) In different effective modules (K eff ) LGPS decomposition energy (a1), ground state pressure (a2) and ground state capacity relative to voltage. (B) eff Figure 3. Decomposition reaction pathways under different voltage ranges, and products induced by different phase equilibria in different voltage ranges. (C, D) XPS measurements of S (c) and P (d) elements for the original LGPS (c1, d1), the cell after 3.2 V CV scan in liquid electrolyte (c2, d2), the stressed cell after 3.2 V CV scan (c3, d3), and the stressed cell after 9.8 V CV scan (c4, d4). Each CV scan was followed by a 10 h hold at high cutoff voltage.

[0060] Figures 48A-48E. Constant current charge and discharge voltage curves of all-solid-state cells using (A1) LCO, (A2) LNMO and (A3) LCMO as cathode materials relative to LTO. The cycling performance of the cells is shown in (B1), (B2) and (B3) for LCO, LNMO and LCMO, respectively. Here, LCO and LNMO were charged and discharged at 0.3C, while LCMO was charged at 0.3C and discharged at 0.1C. All cells were tested at room temperature in a pressurized cell initially pressed with 6T, and the active material was coated with LiNbO3, as shown in Figure 48A-48E. Fig.54(C, D) XPS measurements of LCO, LNMO, LCMO-LGPS before and after 5 cycles. (E) XAS measurements of element S of LCO, LNMO, LCMO-LGPS before (E1) and after (E2) 5 cycles.

[0061] Figures 49A-49G (AD) Pseudophase simulations of the interface between LGPS and (A) LNO, (B) LCO, (C) LCMO, (D) LNMO. The graph depicts the reaction energy of the interface relative to the atomic fraction of the non-LGPS phase consumed. The value of the atomic fraction with the most violent decomposition energy is defined as x m (EG) Mechanically induced metastable patterns of the LGPS-LNO interphase (determined by Fig.49A (E) Energy over hull of the intermediate phase shows a significant response to mechanical compression. (D) and (E) show similar behavior to that observed for bulk phase LGPS and capacity response to pressure (Figures 47A-47D).

[0062] Figures 50A-50C (A) Galvanostatic charge and discharge profiles of all-solid-state cells using LCO and LCMO as cathodes and graphite-coated lithium metal as anode, with cut-off voltages of 2.6-4.5 V (LCO) and 2.6-(6-9) V (LCMO). The cells were charged at 0.3C and discharged at 0.1C. Cycling performance of LCMO lithium metal cells using (B) 1 M LiPF6 in EC / DMC and (C) constrained LGPS as electrolytes, with cut-off voltages of 2.5-5.5 V, charge rate of 0.3C and discharge rate of 0.1C.

[0063] Fig.51 Change in pellet thickness in response to applied force. The initial pellet thickness is 756 μm, the pellet weight is 0.14 g, and the pellet area is 1.266 cm 2 The compressed thickness of the pellet is 250 μm. The calculated density is 2.1 g / cm 3 , which is consistent with the theoretical density of LGPS of 2g / cm 3 near.

[0064] Figures 52A-52F (A)-(F) Synchrotron radiation XRD peaks of the cell at different 2θ angles, showing the broadening of the XRD peaks after high voltage CV scan and hold. The stressed cell showed no XRD broadening after 3.2 V CV scan and hold.

[0065] Fig.53.(top) Illustration of the propagation of the decomposition front. The decomposition phases are labeled α...γ. It can be seen that such propagation requires tangential ionic conduction. (bottom) Energy landscape of the reaction coordinate. The net result is a Gibbs energy shift ΔG, which can be positive or negative based on Eq. 2. Even when ΔG is negative (the reaction is thermodynamically favored), the presence of sufficient overvoltage due to the tangential current can significantly reduce the propagation rate of the front.

[0066] Fig.54 .STEM image and EDS pattern of LiNbO3 coated LCO.

[0067] Fig.55 Rate testing of LCO-LTO batteries using LGPS film as electrolyte. The batteries were tested at 0.3C to 2.5C.

[0068] Fig.56 XAS measurements of LCO, LNMO, and LCMO-LGPS before (denoted as p) and after (denoted as 5c) 5 cycles of element P.

[0069] Figures 57A-57B (A) Charge profile and (B) discharge profile of an LCO all-solid-state cell using LGPS as the electrolyte tested with Swagelok, Al pressurized cell, and stainless steel (SS) pressurized cell, with voltage cutoffs between 3 V and 4.15 V. Swagelok exerts little pressure; the Al cell is soft and exerts low constraint compared to stainless steel, which exerts the strongest constant constraint during battery testing.

[0070] Figures 58A-58B Comparison of CV current density of LGPS+cathode and LGPS+C. CV measurements of LGPS+LCO(30+70) (A) and LGPS+LCMO(30+70) (B) in a pressurized cell and CV measurements of LGPS+C(90+10) in a pressurized cell.

[0071] Figures 59A-59D LCMO / LGPS / Li all-solid-state batteries assembled with (A) bare lithium metal, (B) graphite, and (C) graphite-coated Li as anodes. (D) Cycling performance of LCMO solid-state batteries using different anodes. In the first cycle, all three samples can be charged to about 120 mAh / g, and it is clear that Li / graphite shows the highest discharge capacity at about 83 mAh / g. It can be clearly seen that both Li and graphite anodes have rapid decay problems within the first 5 cycles and after 20 cycles, and their capacities all drop below 20 mAh / g. In contrast, the capacity of the Li / graphite anode is maintained. DETAILED DESCRIPTION

[0072] The present invention provides a rechargeable battery comprising a solid electrolyte (SSE) containing an alkali metal and a sulfide disposed between two electrodes. The solid electrolyte may have a core-shell morphology, thereby imparting increased stability under voltage cycling conditions. These batteries of the present invention are advantageous because they may be all-solid-state batteries, e.g., do not require a liquid electrolyte, and may achieve higher voltages with minimal electrolyte degradation.

[0073] It has been shown that the core-shell morphology in which the core of a ceramic-sulfide solid electrolyte is encapsulated in a rigid amorphous shell improves the stability window. It is believed that the mechanism behind this stability is related to the tendency of ceramic-sulfides to expand by up to more than 20% during decay. Applying a volume constraint mechanism, this expansion is prevented, which in turn suppresses decay. We have generalized this theory and provided experimental evidence showing improved stability using post-synthesis production of a core-shell morphology of LGPS. Based on the decay morphology, the magnitude of stability will vary. The mean field solution of the generalized strain model has been shown to be the lower limit of strain-induced stability. The second decay morphology studied (nucleation decay) has been shown to provide greater stabilization capabilities. In addition, experimental evidence suggests that decay is actually the latter (nucleation) morphology, leading to the significant potential of ceramic-sulfide full batteries.

[0074] Further development of the theory underpinning the enhanced stability and performance of core-shell electrolytes has revealed that the strain stabilization mechanism is not limited to the material level, but can also be applied at the cell level via external stress or volume compression. The strain provided by the core-shell structure stabilizes the solid electrolyte via a local energy barrier, which prevents bulk decomposition from occurring. This stabilization effect provided by the local energy barrier can also be produced by applying external stress or volume compression from the cell, where a voltage stability window of up to 5.7 V can be obtained for LGPS, as shown in Figure 2. Figure 1A-1B In a battery cell design based on this technology, a higher voltage stability window exceeding 5.7 V can be expected under higher pressure and volume compression.

[0075] In solid-state batteries, lithium dendrites form when the applied current density is above a critical value. The critical current density is usually reported to be 1-2 mA em at an external pressure of about 10 MPa. -2 In the present invention, the decomposition pathway of the solid electrolyte (e.g., LGPS) at the anode interface is changed by mechanical compression, and the growth of lithium dendrites is suppressed, resulting in excellent rate and cycle performance. -2 After cycling at current densities of 2.5 Å and 8.5 Å, no short circuit or lithium dendrite formation was observed.

[0076] Solid electrolyte

[0077] The rechargeable battery of the present invention comprises a solid electrolyte material and alkali metal atoms incorporated into the solid electrolyte material. Specifically, the solid electrolyte used in the battery of the present invention may have a core-shell morphology, wherein the core and the shell generally have different atomic compositions.

[0078] Suitable solid electrolyte materials include sulfide solid electrolytes, such as Si x P y S z , for example, SiP2S 12 , such as Li 10 SiP2S 12 , or β / γ-PS4. Other solid electrolytes include, but are not limited to, germanium solid electrolytes, such as Ge a P b S c , such as GeP2S 12 , such as Li 10 GeP2S 12 , tin solid electrolyte, such as Sn d P e S f , such as SnP2S 12 , iodine solid electrolytes, such as P2S8I crystals, glass electrolytes, such as alkali metal sulfide-P2S5 electrolytes or alkali metal sulfide-P2S5-alkali metal halide electrolytes, or glass-ceramic electrolytes, such as alkali metal P g S h-i Electrolyte. Another material includes Li 9.54 Si 1.74 P 1.44 S 11.7 Cl 0.3 Other solid electrolyte materials are known in the art. The solid electrolyte material can be in various forms, such as powder, granules, or solid sheets. An exemplary form is a powder.

[0079] Alkali metals that can be used in the solid electrolyte used in the battery of the present invention include Li, Na, K, Rb and Cs, such as Li. Examples of Li-containing solid electrolytes include, but are not limited to, lithium glasses, such as xLi2S·(1-x)P2S5, such as 2Li2S·P2S5 and xLi2S·(1-x)P2S5-LiI, and lithium glass-ceramic electrolytes, such as Li7P3S 11-z .

[0080] Electrode Materials

[0081] The electrode material can be selected to have optimal properties for ion transport. Electrodes for solid electrolyte batteries contain metals, such as transition metals, such as Au, alkali metals, such as Li, or crystalline compounds, such as lithium titanates, such as Li4Ti5O12 (LTO). The anode may also comprise a graphite composite, such as lithiated graphite. Other materials for use as electrodes in solid electrolyte cells are known in the art. The electrode may be a solid piece of material, or may be deposited on a suitable substrate, such as a fluoropolymer or carbon. For example, liquefied polytetrafluoroethylene (PTFE) has been used as a binder in the preparation of a solution of electrode material for deposition onto a substrate. Other binders are known in the art. The electrode material may be used without any additives. Alternatively, the electrode material may have additives to enhance its physical and / or ion conducting properties. For example, the electrode material may have additives that alter the surface area exposed to the solid electrolyte, such as carbon. Other additives are known in the art.

[0082] 4V LiCoO2 (LCO, shown in Figure 2A-2B Middle) and 4.8V LiNi 0.5 Mn 1.5 O4(LNMO, shown in Figure 3A-3B High voltage cathodes such as 5.0V Li2C o PO4F, 5.2V LiNiPO4, 5.3V Li2Ni(PO4)F, and 6V LiMnF4 and LiFeF4 can also be used as electrode materials in the all-solid-state battery of the present invention. A voltage stability window exceeding 5.7V (e.g., up to 8V or 10V or even higher) can be achieved. Another cathode is LiCo 0.5 Mn 1.5 O4(LCMO). Exemplary cathode materials are listed in Table 1, and the calculated stabilities of the electrodes in Table 1 are shown in Figure 4 middle.

[0083] Table 1: High voltage (greater than 6 V) electrode candidates with individual material item identifiers.

[0084]

[0085]

[0086]

[0087] Electrode coating

[0088] In some cases, the electrode material may also include a coating that acts as an interface layer between the base electrode material and the solid electrolyte on its surface. Specifically, the coating is configured to improve the interfacial stability between the electrode (e.g., cathode) and the solid electrolyte to obtain superior cycle performance. For example, the coating material for the electrode of the present invention includes but is not limited to graphite, LiNbO3, AlF3, MgF2, Al2O3 and SiO2, particularly LiNbO3 or graphite.

[0089] Based on a new high-throughput analytical protocol that efficiently enables computational searches of very large data sets, a library of different materials was searched to find those coating materials that could best stabilize the interface between the sulfide solid-state electrolyte and typical electrode materials, using Li 10 SiP2S 12 As an example, more than 1,000 coating materials for cathodes and more than 2,000 coating materials for anodes were predicted to have both the required chemical and electrochemical stability. These are generally suitable for LGPS. Table 2 provides the predicted effective coating materials.

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100]

[0101]

[0102]

[0103]

[0104]

[0105]

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112]

[0113]

[0114]

[0115]

[0116]

[0117]

[0118]

[0119]

[0120]

[0121]

[0122] External stress

[0123] The strain stabilization mechanism for enhancing electrolyte stability is not limited to the material level, but can also be applied to the battery cell level by external stress or volume compression. In some embodiments, the external stress is a volume constraint applied to all or part of the rechargeable battery (e.g., a solid electrolyte) delivered, for example, by mechanical pressing. The external stress can be applied by a housing, for example, made of metal. In some cases, the volume constraint can be about 70MPa to about 1,000MPa, such as about 70MPa to about 150MPa, about 100MPa to about 300MPa, about 200MPa to about 400MPa, about 300MPa to about 500MPa, about 400MPa to about 600MPa, about 500MPa to about 700MPa, about 600MPa to about 800MPa, about 700MPa to about 900MPa, or about 800MPa to about 1,000MPa, such as about 70MPa, about The present invention relates to a pressure of about 100 MPa to about 200 MPa, about 400 MPa, about 500 MPa, about 550 MPa, about 600 MPa, about 650 MPa, about 700 MPa, about 750 MPa, about 80 MPa, about 85 MPa, about 90 MPa, about 95 MPa, about 100 MPa, about 150 MPa, about 200 MPa, about 250 MPa, about 300 MPa, about 350 MPa, about 400 MPa, about 450 MPa, about 500 MPa, about 550 MPa, about 600 MPa, about 650 MPa, about 700 MPa, about 750 MPa, about 800 MPa, about 850 MPa, about 900 MPa, about 950 MPa or about 1,000 MPa. In the present invention, "about" refers to ±10%.

[0124] The solid electrolyte may also be compressed prior to inclusion in the battery. For example, the solid electrolyte may be compressed with a pressure of about 70 MPa to about 1,000 MPa, such as about 70 MPa to about 150 MPa, about 100 MPa to about 300 MPa, about 200 MPa to about 400 MPa, about 300 MPa to about 500 MPa, about 400 MPa to about 600 MPa, about 500 MPa to about 700 MPa, about 600 MPa to about 800 MPa, about 700 MPa to about 900 MPa, or about 800 MPa to about 1,000 MPa, such as about 70 MPa, about 75 MPa, about 800 MPa, about 900 MPa, or about 1000 MPa. In some embodiments, the solid electrolyte is compressed by a force of about 100 MPa, about 200 MPa, about 80 MPa, about 85 MPa, about 90 MPa, about 95 MPa, about 100 MPa, about 150 MPa, about 200 MPa, about 250 MPa, about 300 MPa, about 350 MPa, about 400 MPa, about 450 MPa, about 500 MPa, about 550 MPa, about 600 MPa, about 650 MPa, about 700 MPa, about 750 MPa, about 800 MPa, about 850 MPa, about 900 MPa, about 950 MPa or about 1,000 MPa. Once compressed, the solid electrolyte can be used in a battery. Such a battery can also be subjected to external stress to apply mechanical compression (mechanical constriction) to the solid electrolyte (e.g., at the microstructural level), i.e., provide isochoric constraint. The mechanical compression on the solid electrolyte can be 1 GPa to 100 GPa, such as 5 GPa to 50 GPa, such as about 15 GPa. The external stress required to maintain the mechanical compression can be about 1 MPa to about 1,000 MPa, such as about 1 MPa to about 50 MPa, about 1 MPa to about 250 MPa, about 3 MPa to about 30 MPa, about 30 MPa to about 50 MPa, about 70 MPa to about 150 MPa, about 100 MPa to about 300 MPa, about 200 MPa to about 400 MPa, about 300 MPa to about 500 MPa, about 400 MPa to about 600 MPa, about 500 MPa to about 700 MPa, about 600 MPa to about 800 MPa, about 700 MPa to about 900 MPa, or about 150 MPa to about 100 MPa. The external stress may be about 800 MPa to about 1,000 MPa, such as about 70 MPa, about 75 MPa, about 80 MPa, about 85 MPa, about 90 MPa, about 95 MPa, about 100 MPa, about 150 MPa, about 200 MPa, about 250 MPa, about 300 MPa, about 350 MPa, about 400 MPa, about 450 MPa, about 500 MPa, about 550 MPa, about 600 MPa, about 650 MPa, about 700 MPa, about 750 MPa, about 800 MPa, about 850 MPa, about 900 MPa, about 950 MPa, or about 1,000 MPa. The external stress employed may vary depending on the voltage of the battery.For example, a cell operating at 6 V may employ an external stress of about 3 MPa to about 30 MPa, and a cell operating at 10 V may employ an external stress of about 200 MPa. The present invention also provides methods of manufacturing a battery using compression of a solid electrolyte prior to including the solid electrolyte in the battery, such as followed by application of an external stress.

[0125] method

[0126] The battery of the present invention can be charged and discharged for a desired number of cycles, such as 1 to 10,000 times or more. For example, the battery can be cycled 10 to 750 times or at least 50, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1,000, 1,500, 2,000, 3,000, 4,000 or 5,000 times. In an embodiment, the voltage range of the battery is from about 1V to about 20V, such as about 1-10V, about 5-10V or about 5-8V. The battery of the present invention can also be charged and discharged at, for example, 1mA cm -2 Up to 20mA cm -2 , for example about 1-10 mA cm -2 , about 3-10mAcm -2 or about 5-10mA cm -2 The invention can be cycled at any suitable current density.

[0127] Example

[0128] Example 1

[0129] The ion exchange rate was 0.1 mVs-1 at 400 Hz between open circuit voltage (OCV) and 6 V on a Solartron electrochemical potentiostat (1470E) using lithium (coated by Li2HPO4) as the reference electrode. -1 The cyclic voltammograms (CV) of Li / LGPS / LGPS+C were measured at different pressures at a scan rate of 1.5 Å. A liquid cell using LGPS / C film as cathode, lithium as anode and 1 M LiPF6 in EC / DMC as electrolyte was also assembled for comparison. The ratio of LGPS to C was 10:1 in both solid-state CV tests and liquid-state CV tests.

[0130] Cathode and anode films for all-solid-state batteries were prepared by mixing LTO / LCO / LNMO, LGPS, polytetrafluoroethylene (PTFE) and carbon black in different weight ratios. For LTO, LCO and LNMO thin film electrodes, the ratios of active material / LGPS / C were 30 / 60 / 10, 70 / 27 / 3, and 70 / 30 / 0, respectively. The powder mixture was then hand-ground in a mortar for 30 minutes and rolled into a film in an argon-filled glove box with the addition of 3% PTFE. Solid electrolytes for all-solid-state lithium-ion batteries were prepared by mixing LGPS and PTFE in a weight ratio of 97:3, then hand-grinding the mixed powder in a mortar for 30 minutes and finally rolling it into a film in an argon-filled glove box. In order to assemble an all-solid-state lithium-ion battery cell, the prepared composite cathode (LCO or LNMO) film, LGPS film (<100 μm) and anode (LTO) film were used as cathode, solid electrolyte and anode, respectively. The three films of cathode, electrolyte and anode were cold pressed together at 420 MPa, and the pressure was maintained at 210 MPa by using a pressurizing unit during the battery cycle test. The charging and discharging behavior was tested at room temperature using an Arbin BT2000 workstation (Arbin Instruments, TX, USA). The specific capacity was calculated based on the amount of LTO.

[0131] Example 2 - Strain-Stable LGPS Core-Shell Electrolyte Battery

[0132] Theoretical-Physical Picture

[0133] Figure 4 The mechanism by which strain can extend the stability window of LGPS is depicted in A. Consider that LGPS decomposes into an arbitrary set of decomposition products at standard temperature and pressure, denoted as “D” (LGPS→D). As the fraction of LGPS that has decomposed (x D The Gibbs energy of the system of functions of Figure 4 The orange dashed line in A is given and analytically given in Eq.

[0134] G 0 (x D )=(1-x D )G LGPS +x D G D (1)

[0135] The lowest Gibbs energy state is x D =1 (complete decomposition) and the initial state is x D = 0 (original LGPS). Therefore, the reaction energy is ΔG 0 =G 0 (1)-G 0(0) = G D -G LGPS This system is inherently unstable. That is, for all x D value, are all negative. Therefore, for x D For any initial value of , the system will move by increasing x D And reduce G 0 , and finally in the final state x D =1 to end.

[0136] Next, consider the application of a mechanical system to constrain the LGPS particles. Given that LGPS tends to expand during decay, any mechanical constraint will need to be resolved to induce strain in the surrounding neighborhood. Such a constraint system can be at the material level (i.e., core-shell microstructure) or system level (i.e., pressurized battery cells) or a combination of both. Ultimately, such a mechanical system can only induce a limited strain before rupture. The energy required to rupture the system is denoted as G 破裂 .

[0137] Any decomposition of LGPS must lead to an increase in strain energy before the confinement mechanism breaks down. Figure 5A-5B The green line in the figure is about the unconstrained Gibbs (G 0 ) and constraint-induced strain (G 应变 ) Plot the constrained Gibbs energy (G′). The highlighted curve indicates the decomposition pathway of LGPS.

[0138] 1. Particles as original LGPS (x D = 0) with an unbroken restraint mechanism

[0139] 2. As the particles begin to decompose (x D :0→δx D ), the constraint mechanism requires G 应变 Assume that the strain Gibbs is x D function, which increases with x D Towards zero and toward zero

[0140] 3. Once the Gibbs energy of the strained system (G′(x D )) exceeds the Gibbs energy of the ruptured system (G 0 (x D )+G 破裂 ), the constraint mechanism will fail. This is at the breaking point x D =x f Next

[0141] 4. Once x D >x f , the system will continue to decompose completely,

[0142] If the constraint-induced strain Gibbs (G 应变 ) is steep enough, then the total Gibbs D <x f The slope under will be positive (e.g. Figure 5A In this case, LGPS will be in the original state (x D = 0). This work focuses on the quantization of the constraint system so that D ≈0, Metastable ceramic sulfide electrolytes are allowed.

[0143] Two work differentials

[0144] As x D The function of G 应变 The existence of derives from the property of LGPS to expand upon decomposition. Depending on the set of decomposition products as determined by the applied voltage, this volume expansion can exceed 20-50%. Thus, the process of LGPS decomposition is one that can include significant "stress-free" strain, i.e., strain that is a result of decomposition rather than a result of applied stress. A proper thermodynamic analysis of such a decay pathway requires careful consideration of multiple work differentials that are reasonably negligible for other systems.

[0145] Figure 5B Schematic representation of two commonly used sources of work, the "fluid-like" form and the "solid-like" form. In a fluid-like system, the change in work under isobaric conditions is proportional to the change in volume of the system δW = -pδV. For a solid-like system, work is defined with respect to a reference / undeformed state and has the differential form δW = V ref σ ij δe ij , where V ref is the undeformed volume, ∈ is the strain tensor relative to the undeformed state, and σ is the stress tensor corresponding to ∈.

[0146] The general method of showing the equality of these two differential work expressions is as follows. The solid-like stress and strain tensors are separated into compression and deformation terms by using the deviator tensor as defined in Eq. 2. The pressure is related to the stress matrix and volumetric strain∈≡(VV ref ) / V ref And summarize.

[0147]

[0148] Using these definitions, the solid-like work can be separated into a term that includes only compression and a term that includes only deformation.

[0149]

[0150] In the fluid limit, where there is no shape change, assuming that δV ref =0, equation 3 simplifies to δW = -V ref pδ∈=pδV, recovering the fluid-like work differential. In most mechanical systems, this assumption is valid because the undeformed reference volume does not change. However, since the undeformed volume is D and thus δV ref ≠0, so it fails to describe the decomposition of LGPS.

[0151] V ref (x D )=(1-x D )V LGPS +x D V D (4)

[0152] Instead, a proper thermodynamic analysis of the LGPS decomposition requires consideration of two work terms. The fluid term - pδV ref indicates that the reference volume is compressed (i.e., changes x) in the presence of the stress tensor σ D ), and the solid term represents the work required to deform the new reference state V ref σ ij δ∈ ij Taking this into account, the full energy differential is given by Eq. 5.

[0153] δE=TδS+μ α N α -pδV ref +V ref σ ij δ∈ ij (5)

[0154] Converted into Gibbs energy G = E-TS + pV ref -V ref σ ij ∈ ij =μ α N α , yielding the differential form:

[0155] δG=-SdT+μ α δN α +Vδp-V ref ∈ ij δσ ij (6)

[0156] Note that as long as V ref constant, then the conversion G = E-TS-V often used in solid mechanics ref σ ij ∈ij =μ α N α -pV ref is sufficient, and therefore -pV ref Can be set to zero.

[0157] At constant temperature, Equation 6 has the following chemical terms (δG 0 =μ α δN α ) and strain terms Given Figure 5A-5B G′(x D ) in differential form.

[0158]

[0159] In the following discussion, for x D The function of G 应变 , G 应变 To provide a range of values ​​for which LGPS can be stable, we consider two limiting cases. The first case is that the LGPS particle is hydrostatically decomposed and the mean field approximation is used. The fraction of decomposed LGPS is assumed to be uniform throughout the particle (for all ). The second limiting case is spherically symmetric nucleation, where LGPS is i is completely decomposed in a spherical region ( ), and outside this area is the original LGPS As shown below, the hydrostatic situation arises G 应变 The nucleation model shows how this value can be much higher in practice.

[0160] Fluid statics limit / mean field theory

[0161] Local stress experienced by the partition of LGPS particles Directly follow the decomposition distribution The local stresses are said to be compressive and equal everywhere within the particle. In the mean field approximation, this is also true for the decomposed fraction Considering and If there is a one-to-one correspondence between , then the two approximations are equivalent.

[0162] We restrict the focus to x D →0 to evaluate the metastability of LGPS with respect to the original state. Then the particle is known to be at least metastable, where the overall stability is given by G 破裂 In this limit, the relationship between pressure and decomposition fraction is given in Ref. 22 It is shown as p(x D )=x D K eff ∈ RxN . Where K eff is the effective bulk modulus of the system, taking into account both the compressibility of the material and the imposed mechanical constraints. K eff Indicates how much pressure will be required to compress the system sufficiently to allow the volume expansion of the LGPS (∈ RXN ), the volume expansion is accompanied by a decomposition. As shown in Equation 8, assuming no bias strain (reasonable for the fluid model), the differential strain Gibbs can be solved from this.

[0163]

[0164]

[0165] The reference volume is the volume of the unconstrained system, V ref =(1-x D )V LGPS +x D V D Combining Equation 7 and Equation 9 with the metastable condition It is found that as long as Eq. 10 is satisfied, the fluid-like LGPS will be stable.

[0166]

[0167] In Figure 5, Eq. 9 is solved for the case of a core-shell compression mechanism, where the core consists of LGPS or oxygen-doped LGPSO (Li 10 GeP2S 11.5 O 0.5 ) and the shell is any rigid material. The effective bulk modulus is given by K eff =(β LGPS +β 壳 ) -1 Given, where β LGPS is the compression ratio of the LGPS material, and is a parameter that represents the ability of the shell to constrain the particle 22 .

[0168] Sphere-forming core limit

[0169] The maximum local (i.e., highest local pressure) decomposition mechanism is as follows Figure 6 The mechanism of sphere nucleation is shown in Figure 1. In this model, the outer diameter R oThe LGPS particle undergoes decomposition at its center. The decomposition region corresponds to the initial radius R i Since the material has been decomposed into The new reference state has a higher volume than the original state because the decomposition fraction is no longer a constant in the particle as in the hydrostatic case. Instead, for the initial (before decomposition) region r < R i All materials within And for the initial outside this region r>R i All materials,

[0170] In order to D To fit the decomposed reference state of into the gap of radius Ri, both the decomposed sphere and the remaining LGPS must be as Fig. 7A .iii and 7A.iv become strained. Therefore, with respect to the decomposition fraction x D Solving for the stress becomes the problem of a thick-walled spherical pressure vessel that compresses a solid sphere. The pressure vessel has a pressure of R i and R o The reference state inner and outer diameters are given, and the spherical particle has an equilibrium radius R D =(1+∈ RXN ) 1 / 3 R i .

[0171] Displacement vectors for decomposed and original materials and and the radial stress component and The boundary conditions are:

[0172] 1. Continuity between decomposition products and original products: R D +u D (R D )=R i +u P (R i ). The vector notation has been dropped to reflect the radial symmetry of the system.

[0173] 2. Continuity between the radial components of stress of those materials at the interface between the decomposition products and the original products:

[0174] For spherically symmetric stresses in isotropic materials, it is known that the displacement vector has the form u(r) = Ar + Br -2 , where the vector sign has been removed since the displacement is only a function of the distance from the center. The Gibbs strain of a sphere under compression gives a compression term with no deviatoric component. Likewise, at the onset of decay The hollow pressurized sphere has both a compressive component and a deflective component, which combine to form Where S p is the shear modulus of the original material. Combining these terms yields the nucleation equivalent of Eq. 8.

[0175]

[0176] Figure 7B Shows that the original material and the decomposed material have the same elastic modulus (E p =E d ) and Poisson's ratio (v p =v d ). The grey and purple lines reflect the shellless and perfect shell limits of the hydrostatic model, while the blue and red lines represent equation 10 for typical Poisson's values. It can be seen that, in general, the nucleation model provides a steeper strain Gibbs than the hydrostatic model due to the higher pressures involved. Intuitively, a smaller Poisson's ratio (more difficult to compress) improves the stability of the nucleation limit.

[0177] Passivation layer theory

[0178] When the electrode potential is outside the electrolyte stability window, the electrolyte, either liquid or solid, may react with the electrode. To address this problem, it is recommended to select electrolytes so that they form a passivating solid electrolyte interface (SEI) that is at least kinetically stable at the electrode potential. Much of the work on the subject of improving sulfide electrolytes has speculated that such a passivating layer may be formed by the formation of an electronically insulating layer on the surface of the sulfide electrolyte. In this section, we discuss the role of such a passivating layer and provide a quantitative analysis of the mechanism by which we believe the electronically insulating surface layer improves stability.

[0179] exist Fig. 8A In the paper, the thermodynamic equilibrium state is given for the most basic battery half-cell model. The cathode is separated from the lithium metal by an electrically insulating and ion-conducting material (σ=0,κ≠0, where σ,κ are the electronic conductivity and ionic conductivity), and a voltage φ is applied to the cathode relative to the lithium metal. The voltage of the lithium metal is defined as zero point. f ) and lithium ion chemical potential (μ Li +), the differential Gibbs energy can be written as Equation 12 (superscripts a and c distinguish the anode and cathode).

[0180]

[0181] Apply conservation δN a =-δN c ,δna =-δn c This gives the well-known equilibrium condition:

[0182]

[0183] Or, in other words, the electrochemical potential of both electrons and lithium ions (η = μ + zeφ) must be constant everywhere in the cell. As a result, the lithium metal potential (μ Li =η Li ++η e -) remains constant throughout the unit. Fig. 7A The band diagram seen in shows the chemical potential of each species and how the voltage varies across the cell while the electrochemical potential remains constant.

[0184] Figure 8B Depicted is the expected equilibrium state in the case of a solid electrolyte cathode, where the cathode material is embedded in the matrix of the solid electrolyte. In this case, the lower (i.e. more negative) chemical potential of the cathode material relative to the electrolyte causes charge separation, which produces an interfacial voltage χ i . Similar to the procedure following Equation 12, it can be shown that the equilibrium point now includes the anode (a), cathode (c) and solid electrolyte (SE):

[0185]

[0186] Similar to Equation 13, Equation 14 yields the condition that the lithium metal potential remains constant throughout the cell.

[0187] The proposed mechanism for the stabilization of the passivation layer in sulfide electrolytes is depicted in Figure 8C In this case, the solid electrolyte is coated with an electronically insulating material. Since the external circuit does not directly contact the solid electrolyte and there is no electron conduction pathway, the number of electrons within the solid electrolyte is fixed. Therefore, the Fermi energy cannot be balanced by electron flow. It is speculated that this effect can be used to allow deviations in the lithium metal potential within the solid electrolyte relative to the electrode, resulting in a wider operating voltage window. Figure 8C The energy band diagram of shows how the electronic electrochemical potential can experience a local maximum (or minimum) in a solid electrolyte due to the lack of electronic conduction. This local maximum (or minimum) is transferred to the lithium metal potential.

[0188] The authors believe that while the electronically insulating passivation layer is a critical design parameter, the above theory lacks the critical role of effective electronic conduction, which occurs due to the "lithium holes" created when lithium ions migrate out of the insulating region, leaving behind corresponding electrons. The differential Gibbs energy of this system can be expressed by adding the solid electrolyte term to Equation 12 (denoted by the superscript SE).

[0189]

[0190] The electron and lithium conservation constraints are now:

[0191] 1.δn SE =-δN SE : The effect of removing lithium ions from a SE is to place the corresponding electrons at the Fermi level of the remaining material.

[0192] 2.δn a =-δn c +δN SE : Lithium ions are obtained at the anode instead of corresponding electrons, which reduces the number of electrons at the Fermi level.

[0193] 3. δN a =-δN c -δN SE : Total lithium conservation.

[0194] Constraints 1 and 2 represent the tethering of electron density and lithium density in the case of insulating particles. Unlike the system governed by equation 12, the Fermi level of the solid electrolyte is not fixed by an external voltage. As a result, the number of atoms in the solid electrolyte is reduced by extracting lithium ions, and thus the number of electrons per atom, the number of electrons per atom, and the Fermi level increase in the insulating region are increased. In fact, this represents the conduction of electrons through lithium holes. Solving equation 15 for the equilibrium point given the above constraints leads to those of equation 14 between anode / cathode and the following relationship between anode and solid electrolyte.

[0195]

[0196] The total voltage experienced within the SE can be expressed as in is the voltage in the absence of lithium extraction from the SE (e.g. Figure 8C The original voltage drawn in S is the voltage generated by the separation of the charges extracted by the lithium. In other words, the system is The charge separation voltage is V, which is a charge-neutral solid electrolyte. However, Eq. 16 is not usually satisfied. Charge separation occurs, which reduces the voltage of the solid electrolyte relative to the anode. With respect to the geometrically determined capacitance C, the charge separation voltage is V S =C -1 E SE This effect is Fig.8D The voltage and chemical potential are given by the blue solid line before charge separation in the SE region. As lithium ions are extracted from the SE by the anode, the voltage in the SE increases from Down to

[0197] The net result of this voltage relaxation in the electronically insulating region is depicted in Fig. 8E Due to the efficient electron transport via lithium hole conduction, negatively charged lithium metal can form locally within the particle once the applied voltage exceeds the intrinsic stability of the solid electrolyte. The negative charge is due to the lithium ions leaving the insulating region to balance the lithium metal potential. Therefore, the local (i.e., within the insulating region) lithium metal is expected to have an interface voltage χ with the remaining solid electrolyte. I This voltage must be equal to the voltage between the lithium anode and the solid electrolyte I =φ SE In short, from a thermodynamic point of view, applying a voltage φ to an electronically insulating solid electrolyte particle relative to a lithium metal anode SE This is equivalent to applying charged lithium metal in direct contact with a solid electrolyte.

[0198] Essentially, this has no effect on the stability of the solid electrolyte. However, in the limit of very low capacitance, as expected, only a small fraction of the lithium ions will need to migrate to the anode, since Therefore, the electronically insulating shell locally captures most of the lithium ions, which maintains the high reaction strain required for mechanical stability.

[0199] Results and discussion

[0200] Electrochemical stability

[0201] The effect of mechanical compression on the stability of LGPS was investigated by comparing the attenuation metrics between LGPS and the same LGPS with an added core-shell morphology that provides a compression mechanism. In order to minimize chemical changes, a compressible core-shell morphology was produced using post-synthesis ultrasound treatment. This core-shell LGPS (hereinafter referred to as "super LGPS") was achieved by high-frequency ultrasound treatment, which caused the outer layer of the LGPS to transform into an amorphous material. The LGPS particles were treated with a 100% saturation vortex prior to ultrasound treatment (Figure 2A). Fig.9A ) and after ultrasonic treatment ( Fig. 9C ) showed obvious formation of an amorphous layer. Statistical analysis of energy dispersive X-ray spectroscopy (EDS) ( Fig. 9B and 9C ) shows that the amorphous shell is slightly sulfur-deficient, while the bulk regions of LGPS and super-LGPS maintain almost identical elemental distributions. EDS line scans of a single [super]LGPS particle ( Figure 10-12 ) confirmed that almost every Ultra-LGPS particle has a sulfur-deficient surface layer, whereas this phenomenon was not observed for LGPS particles. Note that this is true for LGPS sonicated in both tested solvents, dimethyl carbonate (DMC) and diethyl carbonate (DEC) ( Figure 11-13 Simply soaking LGPS in DMC without ultrasonic treatment had no significant effect ( Fig.14 ). This post-synthetic core-shell formation approach minimizes changes to the bulk structure of LGPS, allowing us to evaluate the effects of volume compression on stability without composition changes.

[0202] Use Li / LGPS / LGPS+C / Ta units ( Fig.15A ) and Li / Ultra-LGPS / Ta units ( Fig. 15B The electrochemical stabilities of uncompressed LGPS and compressed super-LGPS were evaluated by cyclic voltammetry (CV) measurements using a lithium reference electrode at a scan rate of 0.1 mVs -1 The scanning range is 0.5–5 V. Carbon is introduced here to measure the intrinsic electrochemical stability window of the electrolyte without kinetic compromise. 12 For LGPS, oxidation peaks at 2-4 V and 3.7 V were observed during charge, and multiple peaks below 1.6 V were observed during discharge. These redox peaks can be attributed to the solid-solid phase transition of Li-S and Ge-S components in LGPS. 24 , confirming that LGPS is unstable and undergoes severe decomposition during cycling.

[0203] In contrast, the decomposition of super-LGPS is greatly suppressed, manifested as only a small oxidation peak at a higher voltage (3 V) during charge and almost no reduction peak during discharge ( Fig. 15B ). In fact, sensitive electrochemical impedance spectroscopy (EIS) before and after CV test ( Fig. 15C , 15D ) also confirmed the higher stability of the super LGPS. EIS showed a typical Nyquist plot of battery-like behavior, with a charge transfer semicircle at medium frequencies and a diffusion line at low frequencies. The results showed that after 3 cycles of CV testing, the total impedance of the LGPS composite increased from 300Q to 620Q (an increase of 107%) ( Fig. 15C ), while the total impedance of the Ultra-LGPS composite material increased by only 32% (from 250Ω to 330Ω, Fig.15D ). The smaller increase in impedance after cycling indicates that the Super-LGPS is more stable, so less solid phase and grain boundaries are generated due to decomposition.

[0204] We found that these stability advantages of Super-LGPS over LGPS are even more pronounced when implemented in an all-solid-state half-cell battery. 12Cycling performance of (LTO) using Super LGPS or LGPS as cathode, Super LGPS or LGPS as separator and lithium metal as anode. Cycling performance of each configuration was obtained at low current rate (0.02C), medium current rate (0.1C) and high current rate (0.8C). Figures 16A-18B The results depicted in show that the cycling stability of the half-cell based on Super-LGPS significantly exceeds that of the half-cell based on LGPS.

[0205] To isolate the decomposition of LGPS in the LTO cathode composite, the solid electrolyte layer was replaced by a glass fiber separator. Fig.15E The charge-discharge profile of LGPS (LTO + LGPS + C / glass fiber separator / Li) cycled at 0.5C in the voltage range of 1.0-2.2V is shown. A flat voltage platform at 1.55V appears in 70 cycles, which can be attributed to the redox of titanium. However, the platform length decreases by almost 85.7% from cycle 1 to cycle 70, indicating a large attenuation of the cathode. On the other hand, the super LGPS (LTO + super LGPS + C / glass fiber separator / Li) ( Fig.15F ) shows that the same flat voltage platform remains almost unchanged after 70 cycles. Cyclic capacity curve ( Figure 15G and 15H ) further confirms this increase in cathode stability. For LGPS, after 70 cycles, the charge and discharge specific capacities decreased from ˜159 mAh / g to ˜27 mAh / g and from ˜170 mAh / g to ˜28 mAh / g, respectively. However, Super-LGPS exhibits much better cycling stability than its LGPS counterpart. After 70 cycles, the discharge capacity is still as high as 160 mAh / g, with only a rough 5% capacity loss.

[0206] In each of these results, those super-LGPS particles with core-shell morphology outperformed their LGPS counterparts in terms of stability. 22 As discussed in , a core-shell design was proposed to stabilize ceramic-sulfide solid electrolytes by volume confinement of the core by the shell. The present experimental electrochemical stability data are consistent with this theory. As seen in the case of super-LGPS, the lack of sulfur shell is expected to reduce the effective compressibility of the system and thus increase the volume confinement. 22 The performance of the solid-state half-cell (solid cathode + glass fiber / liquid electrolyte + lithium metal anode) in the voltage range of 1-2.2 V relative to lithium shows that Super-LGPS actually has improved stability over LGPS in both the oxidation and reduction of LGPS. In addition, the Coulombic efficiency of Super-LGPS is also higher than that of LGPS, indicating that the charge transfer efficiency in the system is improved and less charge is involved in unwanted side reactions.

[0207] Decomposition mechanism

[0208] To better understand the mechanism of LGPS decomposition, TEM analysis was performed to investigate the microstructure of the LTO / [super]LGPS interface after cycling. FIB samples were prepared after 1 charge-discharge cycle relative to a lithium metal anode ( Fig.19A ), which includes a composite cathode (LTO+LGPS+C) and a separator layer (LGPS). A platinum layer was deposited onto the cathode layer during FIB sample preparation to protect it from ion beam milling. At the cathode / separator interface (hereinafter referred to as the "LTO / LGPS main interface"), there is a transition layer with multiple small dark particles, as shown in the TEM bright field (BF) image ( Fig.19B , Fig. 20 ) and STEM dark field (DF) images ( Fig.19D , Fig. 20 ). The particles in the transition layer of the STEM DF image show bright contrast, indicating the accumulation of heavy elements. In order to understand the chemical composition of the transition layer, STEM EELS (electron energy loss spectroscopy) line scans were performed. EELS spectra show that Li k ,Ge M4，5 ( Figures 21A-21B )、Ge M2，3 and P L2，3 ( Fig.15E ) peak exists in the entire transition layer, but the sulfur peak (S L2，3 , S L1 ) only appears inside the bright particles and does not exist in the area outside the bright particles ( Fig.15E 12-14). This observation indicates that the bright particles in the transition layer are sulfur-rich, which is confirmed by both STEM images (sulfur is the heaviest element among Li, Ge, P, and S) and EELS line scan observations ( Fig.19E , Fig.21A , Fig. 21B , Fig.22A and Fig. 22B ) and is supported by previous studies reporting that the decomposition products of LGPS would be sulfur-rich phases including S, LiS, P2S5, and GeS2 12 confirmation.

[0209] Since the composite cathode layer is composed of LTO, LGPS and C, there will be a secondary LTO / LGPS interface (hereinafter referred to as "LTO / LGPS secondary interface") prevalent within the cathode layer. Fig.19F A typical STEM DF image showing the LTO / LGPS subinterface, where bright particles with similar morphology appear again. The density of such bright particles is much higher due to the higher carbon concentration in the cathode layer and thus promotes the decomposition of LGPS. The corresponding STEM EELS line scan spectrum ( Figure 19G ) shows strong S L2，3 The peak is present in the interface region, again confirming that the bright particles are sulfur-rich.Thus, after 1 charge-discharge cycle, sulfur-rich particles are present at both the primary and secondary LTO / LGPS interfaces in the LGPS half-cell.

[0210] As a comparison, Figures 23A-23F (S)TEM study showing the microstructure and composition of the Super-LGPS half-cell. The main LTO / Super-LGPS interface after 1 charge-discharge cycle is shown by TEM BF image ( Fig.23A ) characterization. A smooth interface was observed between the ultra-LGPS isolation layer and the composite cathode layer ( Fig. 23B ). The primary LTO / super LGPS interface is clean and uniform, showing no transition layer or dark particles. The secondary LTO / super LGPS interface is also studied by STEM DF images, EDS line scans, and EDS mapping for comparison ( Figures 23C-23E ). The results show that as the STEM EDS line scan progresses from the internal super-LGPS particles to the sub-LTO / super-LGPS interface, and finally into the LTO+C composite region ( Fig.23D and Fig.24A , Fig. 24B ), the atomic percentage of sulfur continues to decrease. In other words, the super LGPS particles still maintain the sulfur-deficient shell characteristics after cycling, and no sulfur-rich transition layer is formed at the LTO / super LGPS sub-interface. STEM EDS quantitative analysis ( Fig.23F ) show that the sulfur atomic percentage inside the super-LGPS particles is as high as ~38%, while the sulfur atomic percentage at the sub-LTO / super-LGPS interface is as low as 8%.

[0211] These results suggest that the nucleation limit is a more realistic representation of the actual decay process than the hydrostatic limit. i ≈20 nm. In super-LGPS, the shell thickness is also roughly l≈20 nm. Therefore, if we consider the formation of such sulfur particles near the core-shell boundary in super-LGPS, the minimum distance from the center of the sulfur-rich particle to the outside of the shell is R o =R i +l≈40nm. In this case, It satisfies the conditions required for applying the nucleation model R i <<R o In summary, we know that LGPS decays by a mechanism that leads to the nucleation of sulfur-rich particles on the surface. We also know that applying a shell with a certain thickness such that l ≈ R iThese results suggest that the original core-shell state is at least metastable with respect to decay toward a state with nucleation decay just below the core-shell interface.

[0212] in conclusion

[0213] In summary, we have developed a generalized strain model to show how mechanical compression can lead to metastability over a significantly extended voltage range, given the nature of LGPS expansion upon decay. The exact level of the compression-expanded voltage window depends on the decay morphology. We have performed a theoretical analysis of two extremes of the decay morphology: the minimum confinement case and the maximum confinement case. The minimum confinement case consists of mean-field theory, where every part of the particle decays simultaneously, while the maximum confinement case consists of nucleation decay. It has been shown that both cases have the potential to greatly extend the stability window, although the maximum confinement case is the best. We have also developed a theory for the role of an electrically insulating passivation layer in such a strain-stabilized system. The model suggests that this passivation layer contributes to stability by keeping the lithium ions confined within the particle, maximizing the reaction strain.

[0214] The experimental results of the stability performance of LGPS before and after the increase of compressibility shell support this theory. After the shell is formed by ultrasonic treatment, LGPS shows cyclic voltammetry, solid-state battery cycle and solid-state half-cell cycle with significantly improved performance. Because the shell is applied in the post-synthesis method, the chemical difference between core-shell and pure LGPS sample is kept minimum, otherwise this chemical difference may affect stability. It is believed that core-shell is an example of LGPS with mechanical constraints, because during any decomposition period, the LGPS core will attempt to expand, and the shell will remain fixed. In other words, the shell provides quasi-isochoric constraint to the core, which depends on the biaxial modulus and particle geometry of the shell.

[0215] Analysis of the decay morphology seen in LGPS particles, but not in Super-LGPS particles, suggests that the nucleation-decay limit more accurately reflects true thermodynamics. In LGPS, the sulfur-rich decay centers that nucleate after cycling are found to be embedded in the surface of the LGPS particles. Additionally, these nucleated decay centers are not seen in the cycled Super-LGPS. Super-LGPS maintains a shell thickness comparable to the decay sites in LGPS (approximately 20 nm), which is predicted to be sufficient for the high level of stabilization provided by the nucleation model. These results, combined with the improved stability of Super-LGPS, indicate that not only does strain stabilization occur, but also that the magnitude of strain stabilization that occurs is dominated by the maximum confined decay mechanism. This is a promising result, as this nucleation decay has been shown to provide greater G 应变 values, opening the door to solid-state batteries operating at much higher voltages than reported so far.

[0216] method

[0217] Sample preparation

[0218] LGPS powder was purchased from MSE Supplies. Ultra-LGPS was synthesized by soaking LGPS powder in organic electrolytes such as dimethyl carbonate (DMC) and diethyl carbonate (DEC) and then sonicating for 70 h in a Q125 sonicator from Qsonica, a microprocessor-based programmable sonicator.

[0219] Electrochemistry

[0220] The concentration of the electrolyte was 0.1 mVs / s between 0.5 V and 5 V on a Solartron electrochemical potentiostat (1470E) using lithium as the reference electrode. -1 The cyclic voltammograms (CV) of the Li / LGPS / LGPS+C / Ta unit and the Li / super LGPS / super LGPS / Ta unit were measured at a scan rate of 1000 nm. The electrochemical impedance spectra of the Li / LGPS / LGPS+C / Ta unit and the Li / super LGPS / super LGPS / Ta unit were measured at room temperature by applying a 50 mV amplitude AC potential in the frequency range of 1 MHz to 0.1 Hz before and after the CV test. The composite cathode used was prepared by mixing LTO, (super) LGPS, polyvinylidene fluoride (PVDF) and carbon black in a weight ratio of 30:60:5:5. This powder mixture was then manually ground in a mortar for 30 minutes and rolled into a film in an argon-filled glove box. SE was prepared by mixing (super) LGPS and PVDF in a weight ratio of 95:5, then manually grinding the mixed powder in a mortar for 30 minutes and finally rolling it into a film in an argon-filled glove box. To assemble the solid-state cell, the prepared composite cathode film, (super)LGPS film and Li metal foil were used as cathode, solid electrolyte and counter electrode, respectively. Before being assembled into a battery, the composite cathode and (super)LGPS films were cold pressed together. A piece of glass fiber separator was inserted between the (super)LGPS film and the Li metal foil to avoid interfacial reactions between the two phases. Only 1 drop of 1M LiPF6 in a solution of ethylene carbonate (EC) and dimethyl carbonate (DMC) (1:1) was carefully applied to the glass fiber to allow lithium ion conduction through the separator. The Swagelok-type cell was assembled in an argon-filled glove box. The assembly process of the (super)LGPS battery was the same as that of the (super)LGPS solid-state battery, except that the (super)LGPS SE layer was removed. The charge / discharge behavior was tested at room temperature using an ArbinBT2000 workstation (Arbin Instruments, TX, USA). The specific capacity was calculated based on the amount of LTO in the cathode film (30 wt %).

[0221] Characterization

[0222] For FIB sample preparation, after 1 charge-discharge cycle in the (super)LGPS solid-state battery, the cold-pressed thin films of the composite cathode and (super)LGPS were taken out of the argon-filled glove box. They were then mounted on a SEM stub and sealed into a plastic bag in the same glove box. The FIB sample preparation was performed on a FEI Helios 660 dual beam system. The prepared FIB samples were then immediately transferred to a JOEL 2010F for TEM and STEM EDS / EELS characterization.

[0223] Density functional theory calculations

[0224] To allow comparability with the Materials Project crystallographic database, all DFT calculations were performed using the Materials Project standard. All calculations were performed in VASP using the recommended projected augmented plane wave (PAW) pseudopotential. An energy cutoff of 520 eV with a k-point grid of 1000 / atom was used. Compressibility values ​​were obtained by discretely evaluating the average compressibility of the material between 0 GPa and 1 GPa. Enthalpies were calculated at various pressures by applying external stresses to the stress tensor during relaxation and self-consistent field calculations.

[0225] Example 3 - Computational Method for Selecting the Optimal Interface Coating

[0226] As with their liquid counterparts, the key performance metrics for solid electrolytes are stability and ionic conductivity. For lithium systems, two very promising families of solid electrolytes are garnet-type oxides and ceramic sulfides. These families are represented by the high-performance electrolytes of LLZO oxides and LSPS sulfides, respectively. Oxides tend to maintain good stability over a wide voltage range, but generally have low ionic conductivity (<1 mS cm -1 ) 1 In contrast, sulfides can achieve excellent ionic conductivity (25 mS cm -1 ) 6，20 , but tend to decompose when exposed to the conditions required for battery operation.

[0227] The instability of solid electrolytes can be caused by the bulk decomposition of the inherent material level or the surface / interface reaction when in contact with other materials. At the material level, solid electrolytes tend to be chemically stable (i.e., spontaneous decomposition is very rare), but are sensitive to electrochemical reactions with the lithium ion reservoir formed by the battery cell. The voltage stability window defines the range of lithium chemical potential, within which the solid electrolyte will not be electrochemically decomposed. The lower limit of the voltage window indicates the beginning of reduction, or the consumption of lithium ions and corresponding electrons, while the upper limit indicates the beginning of oxidation, or the generation of lithium ions and electrons. When always subjected to an applied voltage, the voltage window affects the body of any solid electrolyte particle. Although interfacial reactions occur between the solid electrolyte and the second 'coating' material at the contact point, these reactions can be two-body chemical reactions in which only the solid electrolyte and the coating material are reactants, or three-body electrochemical reactions in which the solid electrolyte, the coating material, and the lithium ion reservoir are all involved. These two types of reactions are independent of and dependent on the state of charge or voltage, respectively, which is determined by the participation of the lithium ion reservoir.

[0228] Previous studies have revealed that the most common lithium-ion electrode materials, such as LiCoO2 (LCO) and LiFePO4 (LFPO), form unstable interfaces with most solid electrolytes, especially high-performance ceramic sulfides. Successful application of ceramic sulfides in solid-state batteries can employ suitable coating materials that can mitigate these interfacial instabilities. These coating materials can be inherently electrochemically stable and form an electrochemically stable interface with the ceramic sulfide over the full voltage range of operation. In addition, if different solid electrolytes are used in different cell assemblies for maximum material-level stability, the coating material can also be varied to maintain a chemically stable interface.

[0229] In short, the choice of coating material depends on both the type of solid electrolyte and the intended use of the operating voltage (anodic film, diaphragm, cathode film, etc.). Pseudo-binary computational methods can approximately solve the stability of a given interface, but the computational cost is high and has not been developed on a very large scale. The main performance bottleneck of high-throughput analysis of interface stability is the cost of constructing and evaluating many high-dimensional convex hulls. In the case of material phase stability, the dimensionality of the problem is dominated by the number of elements. For example, calculating the interfacial chemical stability of LSPS and LCO will require a 6-dimensional hull corresponding to the element set {Li, Si, P, S, Co, O}. In the case where the system is open to lithium, the electrochemical stability of this interface is calculated, thereby removing lithium from the set and the required hull becomes 5-dimensional ({Si, P, S, Co, O}).

[0230] Here, we introduce new computational protocols to perform interface analysis more efficiently and thus enable efficient high-throughput searches for suitable coating materials taking into account both solid-state electrolytes and operating voltage ranges. We demonstrate these protocols by applying them to search over 67,000 material entries from the Materials Project (MP) for suitable coating materials for LSPS that have shown lithium conductivity up to about 25 mS cm for both anodic and cathodic operation. -1 Coating material candidates that are inherently stable at the material level and form a stable interface with the LSPS over a specified voltage range are referred to as "functionally stable".

[0231] To establish the criteria, we focused on finding anode coating materials that were functionally stable in the window of 0-1.5 volts relative to lithium metal and cathode coating materials that were functionally stable in the window of 2-4 volts relative to lithium metal. These voltage ranges are based on the cycle ranges typically seen in today's lithium-ion batteries. In the anode range, we were particularly interested in finding materials that were stable at 0 volts relative to lithium metal because it would enable lithium to be used as a commercial anode material.

[0232] Due to remaining computational limitations, this work focused only on those materials that required an LSPS interface hull dimension less than or equal to 8. In other words, materials were considered only if the elements present in the material consisted of {Li, Si, P, S} plus at most four additional elements. A total of 69,640 crystal structures in the MP database were evaluated for material-level voltage windows. Of these, 67,062 materials met the requirement of less than 8 dimensions and were therefore evaluated for functional stability using LSPS. In summary, for LSPS, more than 1,000 MP entries were found to be functionally stable in the anode range, and more than 2,000 MP entries were found to be functionally stable in the cathode range. Experimental probing of interface stability was used to select materials to confirm these predictions.

[0233] Results and discussion

[0234] Data acquisition and computational efficiency

[0235] In order to efficiently evaluate the stability of the interface between each of these 67,062 potential coating materials and LSPS, two new calculation schemes were developed. In order to minimize the number of hulls that must be calculated, the coating materials are binned based on elemental composition. Each unique element set requires a different hull, but subsets of elements can be solved simultaneously. For example, calculation of the interface stability between LSPS and iron sulfate (Fe2(SO4)3) requires solving the convex hull of the 6-dimensional element set {Li, Si, P, S, Fe, O}. This hull is the same hull that must be calculated for the interface with LFPO, and includes the 5-dimensional hull required to evaluate iron sulfide (FeS) as a subset. To take advantage of this, instead of iterating through each of the 67,062 materials and calculating the hull required for that material, the minimum number of element sets that span all materials is determined ( Fig.25A ). Then, for each element set, only one hull is needed to evaluate all materials that can be constructed using those elements. This approach reduces the total number of hulls required from 67,062 (one for each material) to 11,935 (one for each element set). Fig.25A As can be seen in , only a few hulls of dimension less than 7 are needed. Compounds that would otherwise require a low-dimensional hull are solved as subsets of a larger set of elements. In addition, since multiple phases of the same composition space require the same hull, the number of required 7- and 8-dimensional hulls is greatly reduced.

[0236] A second approach for minimizing computational cost is a binary search algorithm for determining a pseudo-binary once the hull is computed. Fig.25B Since decomposition at the interface between two materials can consume any amount of each material, the fraction of one of the two materials consumed (x in Equation 1) can vary between 0-1.

[0237] (1-x)LSPS+xA→∑d i D i (1)

[0238] Pseudobinarization is a computational method to determine which x-values ​​of the decomposition described by Equation 1 are the most kinetically driven (e.g., when the decomposition energy is most severe). The RHS of Equation 1 represents the fraction of each thermodynamically favorable decay product ({d i}), and the Gibbs function of the product defines the convex hull of a given x (Hull(x) = ∑d i (x)G i ). The total decomposition energy accompanying equation 1 is:

[0239] G hull (x) = ∑d i(x)G i -(1-x)G LGPS -xG A (2)

[0240] The most strongly kinetically driven reaction between the LSPS and the coating material is the reaction that maximizes the magnitude of Equation 2 (i.e., is most negative), which defines the parameter x m .

[0241] max|G hull (x)|≡|G hull (x m )| (3)

[0242] This maximum decomposition energy is the result of two factors. The first factor, expressed as The decomposition energy is partly due to the inherent instability of these two materials. LSPS ) and decomposition products of coating materials (D A ), It corresponds to the reaction (1-x)LSPS+xA→(1-x)D LsPS +xD A By subtracting this material-level instability from the total hull energy, the effects of the interface (G′ hull ), as defined in Equation 4.

[0243]

[0244] Physically, represents the instability of the material when separated, and G′ hull (x) indicates that once the materials are brought into contact, the instability caused by the interface increases.

[0245] In this work, in order to determine the increased instability of each interface under the strongest kinetically driven fraction (G′(x m )), we implemented concavity using hull to find x within 0.01% error m The binary search algorithm (see Methods) is used to find x in 14 steps of hull evaluation. m A more traditional linear evaluation of the hull with an accuracy of 0.01% would require 10,000 equally spaced evaluations from x = 0 to x = 1. This speed increase was used to efficiently search the functional stability of 67,062 material entries.

[0246] Functional stability

[0247] The functional stability of each of the 67,062 materials at a given voltage was determined by requiring that: (i) the intrinsic electrochemical stability per atom of the material at that voltage is less than the thermal energy (|G hull (x=1)|≤k B T), and (ii) the increase in interfacial instability at a given voltage is less than the thermal energy (|G′ hull (x m )|≤k B T). Under these conditions, the only instabilities in the system are the intrinsic material-level instabilities of the LSPS, which can be overcome by strain-induced methods. 22 To stabilize. Of the 67k materials, 1,053 materials were found to be functionally stable in the anode range (0-1.5 V relative to lithium metal), and 2,669 materials were found to be functionally stable in the cathode range (2-4 V relative to lithium metal). In addition, 152 materials were determined to violate condition (i) in the anode range and 142 materials in the cathode range, but were only decomposed by lithiation / delithiation. The practical use of such materials as LSPS coating materials depends on the reversibility of this lithiation / delithiation process, so these materials are referred to as potentially functionally stable. All functionally stable and potentially functionally stable materials are catalogued in the Supplementary Information and indexed by the corresponding Material Project (MP) id.

[0248] The correlation between the atomic fraction of each element and the interface stability is plotted in Fig.25C and Figures 26A-26C . Fig.25C Describe the chemical reaction between each element and G′ hull (x m ) is related to Figures 26A-26C The electrochemical reactions of each element with G′ at 0V, 2V, and 4V relative to lithium metal are depicted respectively. hull (x m ) Correlation between elemental composition and G′ hull (x m ) means that increasing the content of this element improves the interface stability. Fig.25C The results indicate that chemical stability is best for compounds containing large anions such as sulfur, selenium, and iodine. Fig.26A and Fig.26C Indicates the element species and G′ at low voltage and high voltage, respectively. hull (x m ). This suggests that at these voltage extremes, the interface decomposition is caused by the reduction / oxidation of the intrinsic material level. Rather than interface effect (G′ hull ) dominates. At 2V relative to lithium ( Fig.26B), positive correlations (higher instability) are seen for most elements, with notable exceptions being chalcogens and halide anion groups, which are negatively correlated.

[0249] Effect of anionic species on material-level stability

[0250] Given the high correlation of anionic species with respect to interface stability, an analysis of the dataset was performed with respect to anionic composition. To eliminate overlap between data points, the only compounds considered were those with either a single anion of only one of {N, P, O, S, Se, F, I} or an oxygen-anion of one of {N, S, P}. 45,580 MP entries met one of these criteria, as summarized in Table 3. The percentage of each anionic class found to be electrochemically stable at the material level is also provided.

[0251] Table 3. Size of the single anion and oxy-anion data sets and the percentage of each that is electrochemically stable in the anodic range (0-1.5 V) and cathodic range (2-4 V). For example, F represents all compounds containing F in the chemical formula, while O+N represents all compounds containing both O and N in the chemical formula.

[0252]

[0253] Fig.27A The effect of an applied voltage on the hull energy of a material (in this case, a LSPS) is shown. When the slope of the hull energy with respect to voltage is negative, the corresponding decomposition is reduction, while if the slope is positive, the corresponding decomposition is oxidation. In the middle, there is a region where the hull slope is zero, meaning that there is no reaction with the lithium ion reservoir (i.e., the reaction is neutral with respect to lithium). With this in mind, Fig.27B and Fig.27C The characteristic redox behavior of each anion class in the anodic and cathodic ranges is plotted separately. The "neutral decay" line at 45° represents those compounds that have the same hull energy at both voltage extremes and therefore do not react with lithium ions. Data points above [below] this line are those whose hull energy increases [decreases] relative to voltage and are therefore characteristically oxidizing [reducing] in the plotted voltage range.

[0254] Fig.27B The results show that, as expected, most compounds are reduced in the anodic voltage range of 0-1.5 V relative to lithium metal. It can be seen that nitrogen-containing compounds disproportionately occupy the y-axis, indicating a higher level of stability when in direct contact with lithium metal. This is consistent with previous computational work that indicates that binary and ternary nitrides are more stable to lithium metal than sulfides or oxides. 33However, within the cathode voltage range ( Fig.27C ), a much greater change in anion class is seen. Compounds containing oxy-anions and fluorine remain primarily reducing, while compounds containing phosphorus, sulfides, and selenium are characterized as oxidizing. Oxygen-containing compounds are seen on both sides of the neutral decay line, meaning that oxides are likely lithiated / delithiated in this 2-4 V range.

[0255] exist Fig.27D In the graph, the average hull energy of each anion class is given from 0-5V in 0.5V steps. Nitrogen-containing compounds are shown to be most stable at 0V, while iodine and phosphorus compounds maintain comparable stability. For voltages above 0.5V and 1.0V, the average stability of phosphorus and iodine exceeds that of nitrogen, respectively. At high voltages (>4V), it can be seen that fluorine- and iodine-containing compounds are stable, while nitrogen-containing compounds are the most unstable.

[0256] Effect of anionic species on interfacial stability

[0257] For each anion class, the average value of the total decomposition energy (G hull (x m )) and the fraction (G′) that is a result of interface instability hull (x m ))Depicted in Figures 28A-28C middle. Fig.28A The average instability due to chemical reactions between the anionic species and the LSPS is shown. Compounds containing sulfur and selenium form, on average, the most chemically inert interfaces with the LSPS. Conversely, compounds containing fluorine and oxygen are the most reactive. As a general trend, in the total term (higher G hull (x m The more unstable classes of compounds in the Higher interface contribution (G′ hull (x m )). This means that the differences in the intrinsic chemical stability of each class play a less significant role in determining the chemical stability of the interface than their reactivity with LSPS.

[0258] Fig.28B The average total electrochemical decomposition energy of the interface from 0-5 V in 0.5 V steps is shown. Overall, each anion class follows a path that appears to be dominated by the material-level electrochemical stability of the LSPS ( Fig.27A). This is especially true in the low voltage regime (<1 V) and high voltage regime (>4 V), where the electrochemical effects will be most significant. The largest deviations of the interfacial stability from the intrinsic stability of the LSPS occur in the 1-3 V region. In this "medium" voltage range, those compounds with the lowest chemical decomposition energies (compounds containing S, Se, I, P) deviate the least from the LSPS, while those compounds with large decomposition energies (compounds containing N, F, O, O + This trend suggests that the low and high voltage ranges are dominated by electrochemical reduction and oxidation at the material level, respectively, while the mid-range is dominated by chemical reactions at the interface level. For example, at 0 V, the interface between Al2O3 and LSPS is expected to decay into {Li9Al4, Li2O, Li3P, Li2S, Li 21 Si5}, which is the same set of decay products that would result from each material decomposing independently at 0 V. Therefore, the presence of the interface has no energetic effect.

[0259] The average interfacial contribution of electrochemical decomposition is shown in Fig.28C All anion species tend to G′ at 0V. hull (x m )=0, which means that the material tends to become completely reduced at 0V, in which case the interface effects are negligible compared to the material-level instabilities. Obvious interface instabilities appear in the medium voltage range, while the interface instabilities decrease again in the high voltage range. Again, this means that interface-level chemical effects dominate in the medium voltage range, while material-level reduction [oxidation] dominates at low [high] voltages. At high voltages, the interface contribution to the instability approaches the reaction energy between the most oxidized material and the LSPS. As a result, for any voltage above 4V, the interface will increase the energy instability to equal that of this chemical reaction. This explains the high voltage asymptotic behavior, while the low voltage behavior always tends to 0eV atomic -1 For example, for any voltage above 4 V, LFPO will decompose into {Li, FePO4}}, while LSPS will decompose into {Li, P2S5, SiS2, S}. The introduction of the interface allows these oxidation products to chemically react and form FeS2 and SiO2.

[0260] Effect of anionic species on functional stability

[0261] Fig.29A (anode range) and Fig.29BThe total number of each anion class determined to be functionally stable or potentially functionally stable is given in (cathode range), where they are inherently stable at the material level and form a stable interface with the LSPS within the specified voltage range. For the anode range, nitrogen, phosphorus, and iodine containing compounds have the highest percentage of stable compounds (2-4%), while all other classes are below 1%. The cathode range shows much higher percentages, with sulfur containing compounds reaching 35%. Both iodine and selenium are above 10%.

[0262] Experimental comparison

[0263] The chemical compatibility between various coating materials and LSPS was experimentally tested by hand-grinding the mixture powders of LSPS and coating materials with and without high temperature annealing, followed by X-ray diffraction (XRD) measurements at room temperature. Any chemical reaction between the powders will cause compositional and structural changes in the original phase, which can be detected by changes in peak position and intensity in the XRD spectrum. It is worth noting that even if interfacial reactions are predicted to occur based on thermodynamic calculations, a certain amount of energy may be required to overcome the kinetic energy barrier for these reactions to occur. 4 Therefore, the mixed powders were annealed at high temperatures (300°C, 400°C, 500°C) to determine the onset temperature of the interfacial reaction as well as the reaction products, and the role of kinetics was further evaluated by comparing these results with the DFT-calculated thermodynamic reaction products.

[0264] Figures 30A-30D compare the XRD patterns of this powder mixture annealed at room temperature and at 500°C. 12 , SiO2) mixed with LSPS (FIGS. 30C-30D), while the mixed powder of LCO+LSPS was used for comparison (FIG. 30A). Each single phase (i.e., SnO2, Li4Ti5O 12 , LiCoO2, SiO2 and LSPS) were used as reference ( Figures 31A-31E). By comparing these XRD patterns, it is clear that at room temperature, none of the coated materials reacted with LSPS, as the XRD patterns showed only peaks of the original phase. However, after annealing at 500°C for 6 hours, the different materials showed completely different reactivity with LSPS. LCO was observed to react severely with LSPS, as the peak intensity and position of the XRD pattern of the mixed powder changed completely over the entire 2-θ range of 10-80 degrees (Figure 309A). The original LCO and LSPS peaks disappeared or decreased, while additional peaks belonging to new reaction products (such as SiO2, Li3PO4, cubic Co4S3, and monoclinic Co4S3) appeared, indicating that LCO is incompatible with LSPS. In sharp contrast, the peak intensity and position of the XRD pattern of the SiO2+LSPS mixture never changed, showing only the original peaks both before and after annealing at 500°C. This is direct evidence showing that no interfacial reaction occurs when SiO2 is in contact with LSPS, despite the provision of a large amount of external energy. SnO2 and LTO also show incompatibility with LSPS as new peaks belonging to the reaction products appear in the XRD patterns of their samples annealed at 500°C, however, the peaks of the reaction products are much weaker than in the case of LCO+LSPS. The 2-θ range of the peak positions and intensity variations of the four materials are highlighted by color regions in Figures 30A-30D as an indication of the incompatibility of the different materials with LSPS. It can be observed from Figures 30A-30D that the order of incompatibility is LCO>SnO2>LTO>SiO2, which is completely consistent with our theoretical prediction based on thermodynamic calculations. The onset temperatures of the interfacial reactions of various materials with LSPS are shown in Figures 32A-32D middle.

[0265] The electrochemical stability of typical coating materials is characterized by cyclic voltammetry (CV) techniques, where the decomposition of the tested coating material can be visualized by current peaks at certain voltages relative to lithium. Two typical coating materials are used as demonstrations to show the good correspondence between our theoretical predictions and experimental observations. The CV test of Li2S (Figure 30E) shows a relatively flat region between 0-1.5V, while a large oxidation peak dominates in the region of 2-4V. In contrast, the CV test of SiO2 (Figure 30E) shows a relatively flat region between 0-1.5V, while a large oxidation peak dominates in the region of 2-4V. Fig.30F ) shows a net decrease in the region of 0-1.5 V, and there is a neutral region with little decomposition between 2 V and 4 V. These results are also direct evidence confirming our theoretical predictions based on thermodynamic calculations.

[0266] method

[0267] Data collection

[0268] The data used in this work are the results of previous density functional theory calculations performed as part of the Materials Project (MP) and interfaced with the Materials Application Programming Interface (API). The convex hull was calculated using the Python Materials Genomics (pymatgen) library. Of the initial 69,640 structures evaluated, 2,578 structures were not considered due to the requirement that the hull dimension be equal to or greater than 9.

[0269] Element set generation

[0270] In order to minimize the computational cost of analyzing all 67,062 structures, the minimum number of element sets across all materials is determined. For this reason, the element sets in each structure are combined with the elements of LSPS to produce a list of element sets, wherein the length of each set is equal to the dimension of the required hull of the material. This list is sorted based on the decreasing length of the set (e.g., sorted with the decreasing dimension of the required hull). Then, iterate this set, and remove any set that is equal to the previous set or is a subset of the previous set. The result is the minimum number of element sets, wherein each material can be described.

[0271] The chemical decomposition hull is calculated using the energy and composition from MP. The changes in volume and entropy (ΔG≈ΔE) are ignored. Similarly, by using the lithium grand canonical free energy and subtracting the term μ from the energy Li N Li (ΔΦ≈ΔE-μ Li ΔN Li ) to obtain the electrochemical decomposition hull, where μ Li is the target chemical potential and N Li is the number of lithium ions in the structure. After the hull is calculated, it is used to evaluate each material that exists within the span of its element set.

[0272] Pseudo-binary

[0273] As described in Section 2, pseudo-binary seeks to find the ratio of LSPS to coating material such that the decomposition energy is most severe and, therefore, is the strongest kinetically driven. This problem is simplified by using vector notation to represent a given composition by mapping atomic occupancy to vector elements. For example, based on (Li Co O), LiCoO2→(112) means that there is 1 lithium, 1 cobalt, and 2 oxygen in the unit formula. Using this notation, the decomposition in Equation 1 can be written in vector form.

[0274]

[0275] use To represent the vector and Representing the matrix, Equation 5 becomes:

[0276]

[0277] The relative composition derivative of each decay product can be obtained by adding Seek the inverse to get it.

[0278]

[0279] Equation 7 allows the calculation of the derivative of the hull energy with respect to the fractional parameter x.

[0280]

[0281] By using Equation 7, and the fact that the hull is a convex function of its x, a binary search can be performed to find G hull The maximum value and its occurrence x m The process consists of first defining a two-element vector that defines where it is known that x m The range of x 范围 =(0,1) and the initial guess x0 = 0.5. Evaluating the convex hull under the initial guess yields the decomposition product {D i} and the corresponding energy Then, we can use Equation 7 and Equation 8 to find the slope of the hull energy. If the hull energy is positive, then x 范围 →(x0,1), and if it is negative, then x 范围 →(0, x0). This process is repeated until the upper and lower limits differ by a factor less than a specified threshold of 0.01%, which will always be achieved in 14 steps (2 -14 ≈0.006%).

[0282] Equations 5-8 are defined for chemical stability. In the case of electrochemical (lithium open circuit) stability, the free energy is given by Φ i =G i -μN i Instead, where μ is the chemical potential and N i is the number of lithium in structure i. In addition, the lithium composition is not included in the composition vector of Equation 6 to allow the number of lithium atoms to vary.

[0283] X-ray diffraction

[0284] The compatibility of candidate materials and solid electrolytes was studied by XRD at room temperature (RT). XRD samples were prepared by manually grinding candidate materials (LCO, SnO2, SiO2, LTO) with LSPS powder (weight ratio = 55:30) in an Ar-filled glove box. In order to test the starting temperature of the reaction of candidate materials and LSPS solid electrolyte, the powder mixture was fully spread on a hot plate to heat to different nominal temperatures (300°C, 400°C and 500°C) and then characterized by XRD.

[0285] XRD tests were performed on a Rigaku Miniflex 600 diffractometer equipped with Cu Kα radiation in the 2-θ range of 10–80°. All XRD sample holders were sealed with Kapton film in an Ar-filled glove box to avoid exposure to air during the tests.

[0286] Cyclic voltammetry

[0287] The candidate coating materials (Li2S and SiO2), carbon black and poly(tetrafluoroethylene) (PTFE) were mixed together in a weight ratio of 90:5:5 and hand-ground in an Ar-filled glove box. The powder mixture was sequentially hand-rolled into a film from which discs (5 / 16 inch diameter, ~1-2 mg loading) were punched out to form working electrodes for cyclic voltammetry (CV) tests. These electrodes were assembled into a Swagelok cell with Li metal as the counter electrode, two glass fiber separators and a commercial electrolyte (1 M LiPF6 in 1:1 (volume ratio) ethylene carbonate / dimethyl carbonate (EC / DMC) solvent).

[0288] CV tests were performed by Solartron 1455A at a voltage sweep rate of 0.1 mV / s in the range of 0–5 V at room temperature to investigate the electrochemical stability window of the candidate coating materials (Li2S and SiO2).

[0289] in conclusion

[0290] Our high-throughput pseudo-binary analysis of the Materials Project DFT data has revealed that interfaces with LSPS decay primarily chemically in the range of 1.5 V to 3.5 V and by electrochemical reduction [oxidation] at lower [higher] voltages. The fraction of the decomposition energy attributable to interface effects vanishes as the voltage approaches 0 V. This result suggests that all material classes tend to decay to the most lithiated Li binary and elemental compounds at low voltages, where the presence of an interface has no effect.

[0291] Regarding anion content, we see that it is most important to properly match the operating conditions to the coating material. For example, compounds containing sulfur and selenium show a very high chance of functional stability in the 2-4V cathode range (>25% in all sulfides and selenides). However, less than 1% of these same materials form functionally stable coating materials in the 0-1.5V anode range, with iodine, phosphorus, and nitrogen having the highest performance. Oxygen-containing compounds have a large number of phases that are functionally stable in both voltage regions, but the percentages are lower because the number of oxygen-containing data points is even higher.

[0292] Example 4

[0293] We show that advanced mechanical compression methods can improve the stability of lithium metal anodes in solid-state batteries with LGPS as electrolyte. More importantly, we demonstrate that in symmetric cells, even at 10 mA cm -2 There is no Li dendrite formation and penetration after high-rate testing under 400 ℃ and 800 ℃ conditions. The mechanical compression method is technically achieved by applying an external pressure of 100 MPa to 250 MPa on the battery cell, where the Li metal anode is covered by a graphite film (G) that separates the LGPS electrolyte layer in the battery assembly. At the optimal Li / G capacity ratio, it exhibits excellent cycling performance in both the Li / G-LGPS-G / Li symmetric cell and the Li / G-LGPS-LiCoO2 (LiNbO3 coated) cell. During cycling, the Li / G anode transforms from two layers to an integrated composite layer. Comparison between density functional theory (DFT) data and X-ray photoelectron spectroscopy (XPS) analysis produces the first direct observation of mechanical compression controlling the LGPS decomposition reaction. In addition, under the optimal compression conditions, it can be seen that the degree of decomposition is significantly suppressed.

[0294] Design of Li / graphite anode

[0295] We first investigated the chemical stability between LGPS and (lithiated) graphite by subjecting the mixture of LGPS and (lithiated) graphite to a high temperature treatment at 500 °C for 36 hours in an argon-filled glove box to accelerate the reaction. XRD measurements were performed on the different mixtures before and after the heat treatment, as shown in Figure 33 (A, B, C). Severe decomposition of LGPS in contact with lithium was observed, accompanied by the formation of Li2S, GeS2, and Li5GeP3 ( Fig.33A ). On the contrary, Fig.33B As shown in , after heating, the mixture of LGPS and graphite did not undergo peak changes, proving that graphite and LGPS are chemically stable. After heating the mixture of Li and graphite powder, the synthesized lithiated graphite ( Fig.38 ). When lithiated graphite is further mixed with LGPS, it is chemically stable, such as Fig.33C As shown in , the 26° peak has only a slight change in intensity.

[0296] A Li / graphite anode is designed as shown in Figure 33(D). A protective graphite film is made by mixing graphite powder with poly-PTFE and then covering it on lithium metal. The three layers of Li / graphite, electrolyte and cathode membranes are stacked together in sequence and then mechanically pressed. During the battery test, the pressure is maintained at 100-250MPa. Based on common sense in the art, this pressure helps to obtain good contact between the anode and the electrolyte, but more importantly, it acts as a mechanical compression to improve the electrochemical stability of the solid electrolyte. Scanning electron microscopy (SEM) shows that the graphite particles are transformed into a dense layer under such a high pressure (Figure 39). Fig.33E , Fig.33F In the figure, the anode thus prepared before battery testing can be directly observed by SEM and focused ion beam (FIB)-SEM. The intimate interface contact of the three layers of Li, graphite and LGPS is clearly seen.

[0297] Cycling and rate performance of Li / graphite anode

[0298] The electrochemical stability and rate performance of Li / graphite (Li / G) anode were tested under an external pressure of 100 MPa using an anode-LGPS-anode symmetric cell design. The cycling performance comparison between Li / G-LGPS-G / Li cell and Li-LGPS-Li cell is shown in Fig.34A The Li symmetric battery failed at 0.25 mA cm -2 The Li / G symmetric cell operated for only 10 h at a current density of 0.1 V, while the Li / G symmetric cell still operated after 500 h of cycling, with the overvoltage slowly increasing to 0.28 V. Fig.40 As shown in , according to another battery in which the overvoltage slowly increases from 0.13V to 0.19V after 300 hours of cycling, the stable cycling performance is repeatable, indicating that this slight overvoltage change is different from the battery assembly. SEM shows that the Li / graphite anode is transformed from two layers to an integrated composite layer after long-term cycling, and the total thickness does not change significantly (Figure 41). In Figure 34B, the SEM image of the Li / G anode after 300 hours of cycling in a symmetrical cell is compared with the SEM image of the Li anode after 10 hours of cycling. After long-term cycling, the Li / G anode maintains a dense layer of lithium / graphite composite material ( Fig.34B1 , Figure 34B2). In contrast, after 10 hours of testing, numerous pores appeared in the Li anode, which were most likely induced by the severe decomposition reaction of LGPS with Li metal. The pores are detrimental to both ionic and electronic conductivity, which may be the reason for the sharp voltage increase at 10 hours, when the Li symmetric cell failed.

[0299] We also compared the rate performance of the Li / G symmetric cell under different external pressures of 100 MPa or 3 MPa, as shown in Figure 34C. By changing the working time per cycle, the same charge and discharge capacities were set for different current densities. The Li / G symmetric cell can stably charge from 0.25 mA cm -2 Cycling to 3 mA cm -2 , where the overvoltage increases from 0.1 V to 0.4 V. It can then cycle back to 0.25 mA cm normally. -2 ( Fig.34C1 ). At 3MPa, the battery was tested at 2mAcm -2 Next failure ( Fig.34C2 ). Note that at the same current density, the overvoltage at 100 MPa is only about 63% of that at 3 MPa. -2 After the rate test, the SEM image of the Li / G-LGPS interface shows a close interface contact at 100MPa ( Fig.34D1 ), while cracks and voids were observed after testing at 3 MPa ( Fig.34D2 ). Therefore, the external pressure plays a role in maintaining intimate interfacial contact during battery testing, thereby contributing to better rate performance.

[0300] In order to further understand the influence of Li / G composite formed by battery cycling on its high rate performance, a battery test similar to Figure 34 (E1) was designed. Here, a higher external pressure of 250 MPa was maintained during the test. It was at 0.25 mA cm -2 Start 1 cycle at 5 mA cm -2 We then immediately restarted the battery, again at 0.25 mA cm -2 Ten cycles were run, followed by 5 mA cm -2 The next ten cycles were run. This time, the cell performed normally at 5 mA cm -2 The average overvoltage is 0.6V, and it can still return to 0.25mA cm -2 At a fixed current of 5 mA cm -2The initial voltage surge at 0.25 mA cm indicates a resistance jump, which is most likely related to the fact that Li and graphite are two layers when assembled, and therefore there is not enough Li in the graphite to support such a high current density. -2 After 20 hours of cycling, Li / G begins to transform into a composite material, as shown in FIG34B and FIG41 , in which more Li is stored to support high-rate cycling tests.

[0301] Based on the above understanding, we further reduced the current density of the initial cycle to 0.125 mA cm -2 , and at 0.25 mAh cm -2 The same capacity cycle can be achieved to obtain a more uniform Li distribution and storage in the Li / G composite, thereby improving the lithium transfer kinetics. As shown in Figure 34 (E2), the battery can be used at 10 mA cm -2 When the current density was set back to 0.25 mA cm -2 Note that there is no obvious overvoltage increase at the same low current rate before and after the high rate test, as shown in FIG34E and the inset of FIG42 , where the SEM of the Li / G anode of this cell also shows the clear formation of the Li / G composite material, and no obvious Li dendrites are observed at the interface.

[0302] Li / graphite anode in all-solid-state batteries

[0303] We first performed DFT simulations of the decomposition pathway of LGPS in the low voltage range of 0.0–2.2 V relative to lithium metal. The mechanical compression at the material level is determined by the effective bulk modulus (K eff ). Based on the value of this modulus, the system can be transformed from an isobaric (K eff =0) to isochoric (K eff =∞). The expected value of Kefr in an actual battery system is on the order of 15 GPa. Below, these simulation results are used to explain the XPS results of the valence changes of Ge and P from LGPS in solid-state batteries after CV, rate and cycle tests.

[0304] like Fig.36A As shown in , the decomposition ability of LGPS is low at the high efficiency modulus, indicating that the decomposition of LGPS at low voltage is largely suppressed by mechanical compression. The predicted decomposition products and fraction values ​​are listed in Fig.36B and Table 4. In K eff = 0 GPa (i.e., no mechanical constraint / isobaric pressure), when the voltage approaches zero, the reduction products are close to the lithium binary compounds Li2S, Li3P and Li 15However, after mechanical compression is applied and the effective modulus is set at 15 GPa, Ge, Li x P y and Li x Ge y The formation of x Ge y , GeS and P2S appear. This is also consistent with the known P x Ge y The fact that it is a high pressure phase is consistent with that shown in Figure 36. eff The voltage distributions and reduction products under φt indicate that the decomposition of LGPS follows different reduction pathways at low voltages after applying mechanical compression.

[0305] Table 4. (A)-(D) at different K eff LGPS decomposition products and fractional values ​​down to low voltage

[0306] (A)K eff =0GPa

[0307]

[0308]

[0309] (B)K eff =5GPa

[0310]

[0311] (C)K eff =10GPa

[0312]

[0313]

[0314] (D)K eff =15GPa

[0315]

[0316] It is worth noting that although the applied pressure and effective modulus (K eff ) are both measured in units of pressure and are independent. The effective modulus represents the intrinsic bulk modulus of the electrolyte that increases in parallel with the finite rigidity of the battery system. Therefore, K effThe mechanical compression that can be achieved at the material level in any single particle is measured, and the external pressure applied for operation of solid-state batteries emphasizes the effectiveness of this compression at the interfaces between particles or between electrode and electrolyte layers. This is because the exposed surfaces are most susceptible to chemical and electrochemical decomposition, and the intimate interfacial contact imposed by the external pressure will minimize this. Therefore, even if the applied pressure is only in the order of 100 MPa, the effective bulk modulus is expected to be much larger. In fact, a close-packed LGPS particle will experience a K of approximately 15 GPa. eff Applied pressures of 100-250 MPa are an effective tool for achieving this close-packed structure. In short, the applied pressure minimizes the gaps in the bulk electrolyte, allowing the effective modulus, representing mechanical compression on the material level, to approach its ideal value of about 15 GPa.

[0317] XPS results for LGPS in direct contact with lithium or lithium-graphite anodes and bulk LGPS during battery cycling are provided in Figure 37. Fig.36B The measurements of these valence changes can be well understood based on the phase predictions of the ions. The LGPS in the membrane region away from the anode interface shows the same Ge and P peaks as the pristine LGPS (Figure 37A).

[0318] We first studied the Li / G composites compared with pure lithium metal at 0.25 mA / cm 2 The function of slow rate under 100 MPa external pressure (Figure 37B, Figure 37C). Using pure lithium metal (Figure 37C), the reduction of both Ge and P at the Li-LGPS interface is significant, indicating the formation of Li x Ge y Alloy, element Ge and Li3P. Note that Li x Ge y The valence of Ge in and the valence of P in Li3P are negative or below zero, consistent with the Bader charge analysis from DFT simulations (Figure 44). In contrast, using Li / G anodes, reduction is suppressed at the Li / G-LGPS interface, and the valences of both Ge and P remain above zero in the decomposed compounds (Figure 37B). The Li and LGPS interface is chemically unstable, leading to decompositions including those observed in Figure 37C. These decompositions are also consistent with the Fig.36B K at 0 GPa eff Further electrochemical cycling of this chemically decomposed interface will lead to an increase in the decomposed volume fraction, eventually consuming all of the LGPS. In contrast, the graphite layer in the Li / G anode prevents chemical interfacial reactions between LGPS and Li, and under moderate mechanical compression, the electrochemical decomposition appears to be driven by Fig.36B Medium-high K eff≥10GPa path, where GeS, P x The valences of Ge and P2S match those observed from XPS in Figure 37B.

[0319] When the cycle rate increases to 2 mA / cm 2 and 10mA / cm 2 37D, 37E, the decomposition observed at the L / G-LGPS interface under external pressure changes to a metastable pathway, which is different from that in FIG. 37B at 0.25 mA / cm 2 This means that while FIG. 37B is consistent with the thermodynamics predicted in FIG. 36 , at high current densities the decomposition becomes kinetically dominated. Furthermore, it can be concluded that the Li / Ge alloy formation seen in FIG. 37D and FIG. 37E is the kinetically preferred phase, replacing the reduced P. Specifically, based on the valence from XPS, Ge 0 and Li x Ge y and Li3PS4 and Li7PS6 are most likely to decompose. Note that at an external pressure of 3 MPa and thus at the interface the K eff Even at 2mA / cm 2 At high magnification, the reduction of both Ge and P was also observed (FIG. 37F), which is consistent with Fig.36B Medium to Low K eff However, P reduction may still be kinetically rate-limited, as the most reduced state of Li3P was not observed, as in Fig.36B In K eff = 0 GPa as predicted and as observed from the interfacial chemical reactions in Figure 37C.

[0320] These two competing reactions, thermodynamically and kinetically preferred, can be understood by considering the current-dependent overvoltage (η′(i)) for each of the two competing reactions (η→η+η′(i)). The η′ term will be caused by kinetic effects such as ohmic losses, etc. When the current is small (i≈0), η′ disappears, so the thermodynamic overvoltage (η) dominates and favors the ground state decomposition products of Figure 36. However, at high currents, η′ begins to dominate and favor those metastable phases, such as at high K eff Li x Ge y , which are not shown in FIG. 36 in our calculations, since they are the ground state phase in each voltage range.

[0321] Impedance distribution before and after CV test at 100 MPa or 3 MPa ( Fig.45A ) in use Fig.45D After fitting the model shown in Fig.45Band Fig.45C The calculated R 本体 (body resistance) and R ct (charge transfer resistance, here mainly interface resistance) are listed in Table 5. Due to better contact under high pressure, R at 100 MPa ct (38.8Ω) compared to R at 3MPa ct (395.4Ω) is much smaller. After the CV test, the R 本体 There is almost no change, while the R 本体 The significantly increased resistance is attributed to the more severe decomposition of LGPS under ineffective mechanical compression. Again, according to electrochemical tests, it is proved that the decomposition degree is significantly suppressed under the optimal compression conditions.

[0322] Table 5. Calculated R 本体 and R ct

[0323]

[0324]

[0325] in conclusion

[0326] The lithium-graphite composite allows the application of high external pressure during the testing of solid-state batteries with LGPS as electrolyte. This generates high mechanical compression at the material level, which contributes to the excellent rate performance of the Li / G-LGPS-G / Li symmetric cell. For this solid-state battery, the rate performance of the Li / G-LGPS-G / Li symmetric cell was significantly improved at up to 10 mA cm -2 After cycling at a high current density, the cycle can still proceed normally at a low rate, suggesting that there is no lithium dendrite penetration or short circuit. The reduction pathway of LGPS decomposition under different mechanical compressions was analyzed by using both experimental XPS measurements and DFT calculation simulations. For the first time, it was shown that under appropriate mechanical constraints, LGPS reduction follows a different pathway. However, this pathway may be kinetically affected by the overvoltage induced by high current density. Therefore, the decomposition of LGPS is a function of both mechanical compression and current density. According to the battery cycling performance and impedance tests, it is shown that high mechanical compression and kinetically restricted decomposition pathways reduce the overall impedance and achieve LGPS-lithium metal batteries with excellent rate performance.

[0327] method

[0328] Electrochemistry

[0329] The graphite film was made by mixing the active material with PTFE. The weight ratio of the graphite film was graphite:PTFE = 95:5. All cells were assembled using a homemade pressurized cell in an argon-filled glove box with oxygen and water < 0.1 ppm. Symmetric cells (Li / G-LGPS-G / Li or Li-LGPS-Li) were made by cold pressing three layers of Li( / graphite)-LGPS powder-(graphite / )Li together and kept at different pressures during the battery testing. The cells were charged and discharged at different current densities with a total capacity of 0.25 mAh cm per cycle. -2 . The LiCoO2 half-cell was made by cold pressing the Li / graphite composite-LGPS powder-cathode film with a hydraulic press, and the pressure was maintained at 100-250 MPa. LiCoO2 was coated with LiNbO3 using a sol-gel method. The weight ratio of all cathode films was active material: LGPS: PTFE = 68:29:3. The battery cycling data was obtained on a LAND battery testing system. The cycling performance was tested at 0.1C at 25°C. CV tests (Li / G-LGPS-LGPS / C) were performed on a Solartron 1400 unit testing system between OCV and 0.1V at a scan rate of 0.1mV / s. The LGPS cathode film used for the CV test was made of LGPS:super P:PTFE=87:10:3.

[0330] Material characterization

[0331] XRD: XRD samples were prepared by hand grinding LGPS powder with lithium metal and / or graphite in a weight ratio of 1:1 in a glove box. The powder mixture was placed on a hot plate and heated to a nominal temperature (500°C) for 36 hours before characterization by XRD. XRD data were obtained using a Rigaku Miniflex 6G. The mixture of LGPS and graphite before and after high temperature treatment was sealed with a Kapton film in an argon-filled glove box to prevent air contamination.

[0332] SEM and XPS: Cross-sectional imaging of the spherulites of Li / graphite-LGPS-graphite-Li was obtained by Supra 55SEM. The spherulites were broken into small pieces and attached to the sides of the nut with carbon tape so that it was perpendicular to the beam. The nut with the sample was mounted on a standard SEM short column and sealed into two plastic bags in an argon-filled glove box. FIB-SEM imaging was performed on a FEI Helios 660 dual beam system. XPS was obtained from Thermo Scientific K-α+. The sample was mounted on a standard XPS sample holder and also sealed with a plastic bag. All samples were transferred to a vacuum environment within about 10 seconds. All XPS results were fitted by peak differentiation and simulation via Avantage.

[0333] Computational methods All DFT calculations were performed using the Vienna Ab Initio Simulation Package (VASP) following the material project calculation parameters. 32 A K-point density of 1000 kppa, a cutoff of 520 eV, and pseudopotentials recommended by VASP were used. Mechanically constrained phase diagrams were calculated for effective moduli at 0 GPa, 5 GPa, 10 GPa, and 15 GPa using the Lagrange minimization scheme as outlined in Ref. 13. All Li-Ge-PS phases in the Materials Project database were considered. Bader charge analysis and spin polarization calculations were used to determine the charge valence.

[0334] Example 5-

[0335] In this work, we focus on how to use externally applied high voltage or isochoric conditions to stabilize LGPS at the material level by controlling it at the cell level. This advance exceeds the mechanical constraints at the microstructural level present in previous work, where particle coatings were used to induce metastability. Under appropriate mechanical conditions, we show that the stability window of LGPS can be broadened to the upper limit of the tool test of 9.8V. Synchrotron radiation X-ray diffraction (XRD) and X-ray absorption spectroscopy (XAS) that measure the structural changes of LGPS before and after high voltage maintenance show direct evidence of strain of LGPS during these electrochemical processes for the first time. Both thermodynamic and kinetic factors are further considered by comparing density functional theory (DFT) simulations and X-ray photoelectron spectroscopy (XPS) measurements that perform decomposition analysis outside the voltage stability window. These results suggest that mechanically induced metastability stabilizes LGPS until about 4V. In addition, from 4-10V, local stresses experienced by decomposition in rigid mechanical constraints lead to kinetic stability. The combined mechanically induced metastability and kinetic stability allow the voltage window to be extended from 2.1V to nearly 10V. To demonstrate the utility of this approach for practical battery systems, we used this approach to construct all-solid-state batteries with various cathode materials. 12 (LTO) anode and LiCo 0.5 Mn 1.5 O4(LCMO), LiNi 0.5 Mn 1.5 O4(LNMO) and LiCoO2(LCO) cathodes are paired to demonstrate the high voltage stability of confined LGPS. To further probe the electrochemical window of LGPS, we report on Li metal and LiCo 0.5 Mn 1.5 O4's first all-solid-state battery, which can be charged to 6-9V and cycled to 5.5V.

[0336] result

[0337] To illustrate how mechanical confinement affects the electrochemical stability of LGPS, cyclic voltammetry (CV) tests of the LGPS+C / LGPS / Li cell were performed ( Fig.46A ). Three cells were pre-pressed in an assembly with a force of 1, 3 or 6 tons (T) (78 MPa, 233 MPa and 467 MPa, respectively) and then tested in a normal Swagelok cell. The external pressure of the tightened Swagelok cell was calibrated to a few MPa, giving a quasi-isobaric cell test condition. In addition, one cell was initially pressed at 6T and then tightened in a homemade pressurizing unit, and the external pressure constantly applied during the cell test was calibrated to about 200 MPa, applying a quasi-isobaric cell test environment. The density of the LGPS pellets after pre-pressing at 1T, 3T and 6T was 62%, 69% and 81% of the theoretical density of single crystal LGPS, respectively. The morphology of the LGPS pellets after pressing is shown in Fig.51 A. However, the density of the pellet in the pressurized cell calculated from in situ force-displacement measurements ( Fig.51 B) is close to 100% at an external pressure exceeding 30 MPa.

[0338] like Fig.46A As shown in , in the cyclic voltammetry (CV) test, there is a threshold voltage above which each cell begins to decompose severely. For those isobaric cells pre-pressed at 1T, 3T, and 6T, these thresholds are 4.5V, 5V, and 5.8V, respectively. However, the isobaric cell was charged to 9.8V and showed no obvious decomposition. In the low voltage region ( Fig.46B ), for the isobaric cell, two smaller decomposition peaks are visible at ~3 V and ~3.6 V, where reduced peak intensity is observed at increased pressure during the pre-pressing step. In contrast, the isochoric cell avoids these peaks completely. During the CV test, the in situ resistance of the cells in these four cells was measured by impedance spectroscopy at different voltages ( Fig.46C ). It was found here that higher pressures in the pre-pressing improve the contact between the particles and thus reduce the initial resistance in the solid-state battery system (at 3 V, Fig.46C However, when the CV test was performed toward higher voltages, the resistance in the isobaric cell increased faster, indicating that the LGPS in the cathode underwent some decomposition under weak mechanical compression. In contrast, the resistance of the battery tested with the isocathic cell showed almost no change. It is noteworthy that the voltage stability window of crystalline LGPS toward higher voltages was extended from 2.1 V to about 4.0 V by mechanical compression-induced metastability. Fig.46A The observed stability of 5 V to 10 V in the cell, far exceeding 4 V, suggests a different phenomenon.

[0339] like Fig.46DAs shown, synchrotron radiation XRD of LGPS from an isochoric cell indicated that the general crystal structure of LGPS remained unchanged after CV testing up to 9.8 V. However, XRD peak broadening was observed after high voltage CV scans at 7.5 V and 10 V ( Fig.46E and Figure 52). It is found that the peak broadens with increasing 2θ angle ( Fig.46F ) follows a strain widening mechanism rather than a size widening. Note that no obvious strain widening is observed at 3.2 V.

[0340] This strain effect is further illustrated from XAS measurement and analysis. Figure 46G shows the comparison of the P and S XAS peaks of the original LGPS with those scanned to 3.2V and 9.8V by CV in liquid or solid-state batteries. Under the condition of no mechanical constraint (expressed as 3.2VL), where LGPS and carbon are mixed with a binder and tested in a liquid battery, both P and S show obvious peak shifts and shape changes towards high energy, indicating a significant overall oxidation reaction in LGPS in the liquid cell and the rearrangement of the local atomic environment. While in the solid-state battery, the P peak and the S peak do not show any signs of overall oxidation because no peak shift is observed. However, it is worth noting that the shoulder intensity increases at 2470eV and 2149eV in the P and S spectra, respectively. The ab initio multi-scattering simulation of P XAS in LGPS is shown in 46H, where various strains are applied to the unit cell. The comparison between the experiment and the simulation suggests that the increase in shoulder intensity in this XAS may be caused by negative strain (i.e., the compression experienced by the crystallized LGPS after CV scanning and holding at high voltage). If we relate the strain broadening in XRD to the increase in shoulder intensity in XAS and also consider that no significant decomposition current is observed in CV tests up to 10 V, a physical picture emerges related to small local decomposition under moderate mechanical compression. Under constant external pressure around 150 MPa and almost zero porosity in the LGPS pellets, the macroscopic voltage decomposition of LGPS is kinetically greatly suppressed beyond the voltage stability window (i.e., 4.0 V), giving no evidence of Li +The overall transfer of ions and electrons, and therefore no decomposition current in the CV test. However, small local decompositions can still be formed inside and between LGPS particles. Since decomposition in LGPS has a positive reaction strain, such small local decompositions will impose compression on the adjacent crystalline LGPS under a mechanical compressive environment, inducing the strain widening observed in XRD and the increase in shoulder intensity observed in XAS. The fact that both XRD and XAS are ex situ measurements supports our picture at the material level that once the local strain induced by such local decomposition is formed, it will not be easily released due to kinetic barriers even after the external pressure at the battery cell level has been removed. That is, appropriate mechanical conditions can lead to mechanically induced metastability from 4.0 V to 10 V in LGPS without obvious decomposition current in CV tests. Our results here provide direct evidence that the electrochemical window of ceramic sulfides can be significantly widened by appropriately applying mechanical constraints.

[0341] Theoretically, considering the LGPS decomposition and Gibbs energy variation AG chem <0 unconstrained response if:

[0342] ΔG chem +K eff ∈ RXN V>0 (1)

[0343] The reaction can be achieved by applying a bulk modulus K eff The mechanical constraints are used to suppress, where V is the reference state volume, and ∈ RXN is a stress-free reaction expansion, in other words, ∈ RXN is the volume fraction change of LGPS after decomposition in the absence of any applied stress. The effective bulk modulus of Equation 1 is the same as that of Equation 2. 8 The bulk modulus (K 材料 ):

[0344]

[0345] Minimization of free energy in a mechanically constrained ensemble allows calculation of an extended voltage window and ground state decomposition products. Using ab initio data, FIG47A shows that the voltage range of 0-10 V is within four levels of mechanical restraint (K eff =0 GPa, 5 GPa, 10 GPa, 15 GPa) for LGPS. Figure 47A1 shows the energy above the hull, or the magnitude of the decomposition energy. -1The energy above the hull indicates that LGPS is a ground state product thermodynamically, while the elevated values ​​indicate that LGPS will decay. It can be seen that the upper voltage limit of the region where the energy above the hull is almost zero (<50 meV for thermoresistance) increases from about 2.1 V to nearly 4 V. FIG47A2 shows the ground state pressure corresponding to the minimization of the free energy. The pressure is given by K eff ∈ RXN Given, where ∈ RXN corresponds to the volume fraction of LGPS converted to products that minimize the free energy. eff = 15 GPa, the ground state pressure reaches 4 GPa, which corresponds well to the level of local strain used in the XAS simulation of the strained LGPS in Figure 46H. Figure 47A3 The total lithium specific capacity of the ground-state products is shown, which is predicted to be eff Under conditions exceeding 5V, the LGPS electrolyte will not provide greater lithium capacity or cannot decompose further.

[0346] Without considering heat resistance, eff The exact decomposition products predicted by DFT over the entire voltage range are shown in Fig.47B The exact reaction equations are listed in Table 7. This simulation actually thermodynamically predicts how small local decomposition reactions induced by electrochemical driving forces, as discussed in Figure 46, change quantitatively under mechanical compression. The elemental valence states in the decomposition can therefore be directly compared with XPS measurements, which are sensitive to chemical valence information on the particle surface (Figure 47C, Figure 47D), providing complementary information to bulk-sensitive XAS. The stoichiometric LGPS is determined by the valence states Li 1+ ,Ge 4+ , P 5+ , S 2- When LGPS forms lithium metal (Li 1+ →Li 0 ), the remaining elements must be oxidized. eff =0GPa, we are Fig.47B The simulations in Figure 2 suggest that sulfur is most likely oxidized to form S4 above 2.3 V. 1- (LiS4), and S is formed above 3.76V 0 (elemental sulfur). According to DFT simulation of Bader charge, S4 1- or S show very similar charge states and are significantly higher than S in LGPS 2-This is consistent with the large amount of oxidized S observed for LGPS in the liquid cell in XPS after CV scanning to 3.2 V and holding for 10 hours (Figure 47C2). Similarly, oxidation of P to form PS4 was observed in the same 3.2 V liquid cell. 3- P 5+ (Figure 47D2). This suggests that the thermodynamically favorable decomposition actually represents eff = 0 (as opposed to the alternative kinetically favorable decomposition under mechanical compression).

[0347] In contrast, the calculated thermodynamic stability limit of LGPS is in K eff = 15GPa reaches nearly 4V. Therefore, Figure 47C3 and In Figure 47D3, no oxidation of S is observed under the condition of strongly constrained LGPS at 3.2 V and very small amounts of oxidized P are observed. This small amount of oxidized P can be attributed to ineffective confinement from the device or the voltage being close to the thermodynamic voltage. In addition, at 9.8 V, beyond the voltage stability limit, the solid-state battery shows less oxidized S or P than expected. Note that according to Fig.47B , assuming that LGPS decomposes into elemental S and oxidized P in Li7PS6 or Li2PS3. However, this thermodynamic pathway is bypassed. Beyond this thermodynamic stability, there are kinetic factors that stabilize sulfide electrolytes under high mechanical constraints.

[0348] The application of mechanical constraints can greatly reduce the decay rate of ceramic sulfides, such as Fig.53 When the decay rate is slowed sufficiently, the effective stability ("mechanically induced kinetic stability") is high enough to allow the battery to operate. For example, if the electrolyte decays by only one part per million per charge cycle, it is sufficiently stable for a practical battery design that only needs to last for thousands of cycles.

[0349] The proposed mechanism of mechanically induced kinetic stability is depicted in Fig.53 In a given LGPS particle that has undergone decomposition, the particle can be divided into three regions. The first two regions are the decomposed region and the original region, which are Fig.53 (Top) The mole fraction of LGPS decomposed by decomposition (for pure decomposed LGPS, x D =1; for the original LGPS, x D =0). The third region is the interface where the mole fraction transitions from 0 to 1. The propagation direction of the decomposition front is controlled by the thermodynamic relationship of equation 1. If equation 1 is satisfied, the front will propagate inward, with the original LGPS taking precedence. Therefore, the LGPS will not decompose. When equation 1 is violated, the front will propagate into the LGPS and eventually consume the particle.

[0350] However, even when Eq. 1 is violated, the speed at which the front propagates into the original LGPS will still be affected by the imposition of mechanical constraints. Fig.53 (bottom). As the decomposition front propagates, there must be an ionic current tangent to the curvature of the front. This requires the presence of an overvoltage to accommodate the finite conductivity of the front for each elemental species. The ohmic part of the overvoltage is given by the sum of Equation 3, where p i (p) is the resistivity of each species i at the leading edge at the pressure (p) present at the leading edge, l t is the characteristic length scale of the decomposition morphology, and j i is the ion current density.

[0351]

[0352] Considering ρ i (p) can grow rapidly with compression, and this overvoltage would be expected to become significant at high pressures. This effect can be seen by comparing the expected compression with previous molecular dynamics results of compression cells. The pressure on the decomposition front is given by p = K eff ∈ RXN Given, and the elastic volume strain of the material at this pressure is p = K 材料 ε V Since the strain of a single lattice vector is approximately So the strain of the ab-plane of LGPS near the front is expected to be For K eff ≈K 材料 For a fully constrained system, this strain can easily reach 4% because at high voltage ∈ RXN Considering that the activation energy for Li migration in LGPS is predicted to increase from 230 meV to 590 meV at 4% compression, the rate at which lithium reordering can occur is reduced by the following factor:

[0353]

[0354] This many orders of magnitude reduction in the possible reordering rate may explain why isocapacitive cells show virtually no decomposition current for any voltage below 10V.

[0355] Figure 48 shows the galvanostatic cycling of all-solid-state cells using LCO, LNMO and LCMO as cathodes, LGPS as separators and LTO as anodes and their cycling performance. The cell tests were conducted in a pressurized cell where the cells were initially pressed with 6T and then fastened in a bolted [quasi]isotropic cell. It should be noted that LCO is the most common and widely used cathode material included in commercial lithium-ion batteries, and has a high cycling performance relative to Li+ / Li has a platform at about 4V, and LNMO is considered to be one of the most promising high-voltage cathode materials. + / Li has a flat operating voltage at 4.7V. The high rate test of LCO full battery is shown in Fig.55 The charge and discharge curves of LCO and LNMO are depicted in Fig.48A1 and Figure 48B1 In the first discharge cycle, both cells show a flat operating platform, centered at 2 V (vs. Li) for LCO. + / Li is 3.5V), while for LNMO, the center is 2.9V (relative to Li + / Li is 4.4V). In addition, both of them show excellent cycle performance, such as Figure 48A2 and Figure 48B2 As can be observed in Figure 46, for LCO, the capacity fade is only 9% in the first 360 cycles, and for LNMO, the capacity fade is 18% in the first 100 cycles. This indicates that the decomposition or interfacial reaction of the cathode material with LGPS is not very serious. These results are in good agreement with the CV tests reported in Figure 46, which show that mechanical constraints can suppress the decomposition of LGPS and widen its operating voltage range to values ​​much higher than those previously reported. In addition, in order to further probe the stability of LGPS, the previously synthesized LCMO was selected as the cathode because it presents an even higher working platform than LNMO. Figure 48A3 Plotting the battery test curves of LCMO vs. LTO. In both the charge and discharge profiles, two plateaus can be observed, centered at approximately 2.2 V and 3.2 V (vs. Li + / Li is 3.7V and 4.7V), which are respectively 3+ / Mn 4+ and Co 3+ / Co 4+ The oxidation reaction of Figure 48B3 As shown in Figure 2, some capacity decay was observed upon cycling, which may be attributed to the side reactions between LCMO and LGPS at high voltages and corresponds to a 33% decay in the 50th cycle. Thus, in contrast to previously reported results that claimed that the stability window of LGPS is limited to a low voltage range, here we show that LGPS can be used as an electrolyte material in high voltage cathode all-solid-state batteries even when the charging plateau is as high as 3.8 V (vs. Li + / Li at 5.3 V) also showed relatively good cycle performance. Figures 48C1-48D3The XPS measurements of the electron binding energies in LGPS are shown before and after cycling of cells using LCO, LNMO, and LCMO as cathodes. Each element can be oxidized by chemical reaction with the cathode material (chemical oxidation) or by delithiation of LGPS by application of voltage (electrochemical oxidation). Figures 48C1-48D3 As depicted in , after cycling, those electrons in the characteristic region of sulfur-bound electrons show a peak shift to a higher energy state, indicating that the sulfur has been electrochemically oxidized. The presence of oxidized sulfur in the original sample indicates the extent of the chemical reaction with the cathode material.

[0356] Considering that S, rather than P, is combined with the transition metal (either from the coating material or the cathode material), the XAS measurement shows a pre-edge in the intensity of the S element, while no pre-edge from P is found (Figure 48E and Fig.56 Although the interfacial reactions are circumvented by mechanical constraints, there is still a certain amount of side reactions that occur due to the direct contact between the cathode material and LGPS. More interfacial reactions occur after battery cycling.

[0357] The interfacial reaction between two materials (i.e., LGPS and cathode material) presents computational challenges because ab initio simulation of the interface presents a unique burden. Instead, the preferred method to simulate both the chemical stability and electrochemical stability of the interface is the so-called pseudophase (also called pseudo-binary) method. In these methods, a linear combination of the target materials is adopted and represented as a single phase, where both the composition and energy are given by the linear combination. This phase is a pseudophase. Conventional stability calculations can then be applied to the pseudophase to estimate the reaction energy of the interface. Fig.49A -D and Table 6 give the results of chemical reaction pseudophase calculations of LGPS+LNO, LCO, LNMO and LCMO. Fig.49A -D, the atomic fraction of the cathode material (or LNO) is swept from 0 to 1 (representing pure LGPS to pure cathode or LNO). Whichever value of the atomic fraction makes the reaction energy most negative represents the worst-case reaction and is referred to as x m Table 6 gives these x values ​​for each interface. m values, as well as the worst-case reaction energies, decomposition products, and an additional pseudophase representing the decomposition interface. This pseudophase representing the decomposition interface (also called an intermediate phase) can be used to calculate how the decomposition interface will further decay when the battery is cycled. Fig.49E-G shows the electrochemical stability of the LGPS+LNO interphase. Note that the chemical reaction between LGPS and the cathode material occurs as soon as the materials come into contact during cathode membrane assembly. This is in contrast to the electrochemical reaction, which does not occur until the external circuit assembly is attached. Therefore, the main difference between the two is that the chemical reaction occurs before pressurization / cell assembly, while the electrochemical reaction occurs after. Since the chemical reaction occurs in the absence of a fully assembled cell, the initial reaction is always at K eff = 0 (the electrochemical reaction occurs at K eff occurs below).

[0358] Table 6. Chemical reaction data at the interface between LGPS and LNO, LCO, LCMO or LNMO. RXN is the worst-case reaction energy between the two phases, and x m is the atomic fraction of the non-LGPS phase consumed in this worst-case scenario. 'Products' lists the phases produced by this worst-case reaction. 'Chemical Decomposition Pseudophases' is an application of pseudophase theory to the product set in 'Products'. It represents an artificial phase with a linear combination of the composition, energy, and volume of its constituent phases.

[0359]

[0360] Fig.49B -D shows that the chemical reaction energies of LCO, LNMO and LCMO are 345 meV atomic -1 , 322meV atoms -1 and 335meV atoms -1 Although the 124 meV atom -1 The much lower reaction energy of LNO ( Fig.49A ) coating, but the coating was not perfect, allowing some contact with the LGPS which resulted in the chemical oxidation of sulfur seen in the original samples of Figures 48C-48E. Fig.49E-G shows that the products produced by the chemical reaction of LGPS and LNO (constituting the LGPS-LNO intermediate phase) also experience mechanically induced metastability. Therefore, in a full cell in which the cathode particles are coated with LNO, appropriate compression (such as those depicted in Figure 48) will lead to mechanically induced metastability in the bulk of the solid electrolyte and at the interface with the cathode material. As a general rule, the interface of LGPS with an insulator (such as LNO) is more likely to experience mechanically induced metastability than the interface of LGPS with a conductor (such as LCO, LNMO and LCMO). The reason is that when the intermediate phase is oxidized to form lithium metal, if the interface is between two electronic insulating materials, the lithium metal will be formed locally. However, if one of the two phases is conductive, the lithium ions can migrate to the anode and thus form a non-local phase. In the latter case, since the volume of the lithium phase formed will not be included in the local volume change, the local reaction expansion will be greatly reduced. On the contrary, if the lithium metal phase is formed locally, it brings a larger local volume change and therefore brings a larger reaction expansion. To this end, it is necessary to coat the cathode material with an insulator (such as LNO) for confinement to induce mechanically induced metastability at the interface of LGPS.

[0361] Typically, lithium metal is soft and it is difficult to apply pressure due to the instantaneous passage of lithium through the bulk solid electrolyte. In order to explore the high voltage performance of pressurized LGPS in a lithium metal solid-state battery system, lithium metal was used as the anode and a graphite layer was used as a protective layer, which allowed high voltage to be applied during battery testing. First, as shown in Figure 57, lithium metal-LCO batteries were fabricated under different mechanical conditions using Swagelok, aluminum pressurized cells, and stainless steel pressurized cells. Similarly, the interfacial reactions and decomposition reactions under the strongest confinement conditions were the lowest. A similar structure was applied to fabricate higher voltage lithium metal batteries using LCMO as the cathode, where the cell was initially pressed with 6T. Figure 58 shows that the graphite protective layer mitigates the interfacial reactions between lithium metal and LGPS. As shown in Figure 59, the decomposition of LGPS itself is very small under strong mechanical confinement conditions, and it contributes very little to the decomposition current, as shown in Figure 59. As Fig.50A As depicted in Figure 2, the LCMO cathode can then be charged to 9 V, which simulates the high voltage charging state of the yet-to-be-discovered high voltage redox chemistry. By charging the LCMO at 6 V, 7 V, 8 V, and 9 V, discharge capacities of 99 mAh / g, 120 mAh / g, 146 mAh / g, and 111 mAh / g were obtained, respectively. Fig.50A). This indicates that the additional lithium capacity comes from the higher voltage state of the LCMO. Although there are more side reactions after the battery is charged to a voltage above 8V, it can be seen that the battery maintains cycling performance even up to 9V. This high voltage cycling demonstrates a high electrochemical window for constrained LGPS exceeding 9V. In a highly delithiated state, cathode materials generally show poor electrochemical stability, and the reaction between the cathode material and the electrolyte is also more severe.

[0362] To compare this performance with conventional electrolytes, Fig.50B Depicts an organic liquid electrolyte that fails at nearly 5 V. However, solid-state batteries tested under isochoric conditions can be charged to 9 V ( Fig.50A ) without evidence of a decomposition platform. In addition, compared with liquid batteries ( Fig.50B ) in contrast, the battery ( Fig.50C ) showed stable cycling performance and high Coulombic efficiency even at a high cut-off voltage of 5.5 V. Although the performance of the Li-metal-LCMO battery was not as good as that of the full battery due to the mechanical softness of Li metal, this result still suggests that solid electrolytes are a better platform for running high-voltage cathode materials, unlike liquid electrolytes.

[0363] In summary, we demonstrate how mechanical constraints broaden the stability of ceramic solid electrolytes, pushing their electrochemical window to levels exceeding those of organic liquid electrolytes. CV tests show that properly designed solid electrolytes operating under isochoric conditions can operate at voltages up to nearly 10V without obvious signs of decomposition. A mechanism for this mechanically induced kinetic stability of sulfide solid electrolytes is proposed. Furthermore, based on this understanding, it has been shown how several high-voltage solid-state battery cells using some of the most commonly used and promising cathode materials can operate up to 9V under isochoric conditions. Therefore, the development of high-voltage solid-state cells is no longer compromised by the stability of the electrolyte. We expect this work to be an important breakthrough in the development of new energy storage systems and cathode materials focusing on very high voltage (>6V) electrochemistry.

[0364] method

[0365] Sample characterization

[0366] Structural analysis

[0367] CuKα radiation (wavelength 2.0 Å) was used in a Rigaku Miniflex 6G diffractometer operating at 45 kV and 40 mA. ) to collect conventional XRD data. The operating conditions were a 2θ scan between 10° and 80° with a step size of 0.02° and a scan rate of 0.24 seconds per step.

[0368] Electrochemical characterization

[0369] The LGPS+C / LGPS part of the cell was made into pellets by pressing powders at 1T, 3T, 6T respectively and put into a Swagelok or homemade pressurized cell. In the CV test, the voltage was ramped up from the open circuit voltage to 10V, during which the decomposition current was measured at each voltage. The CV tests were performed on a Solartron 1400 electrochemical testing system between OCV to 3.2V, 7.5V and 9.8V respectively with a scan rate of 0.1mV / s. The voltage was maintained for 10 hours after the CV scan to ensure complete development of the decomposition and it was scanned back to 2.5V before any other characterization. Electrochemical impedance spectroscopy (EIS) was performed on the same machine in the range of 3MHz to 0.1Hz.

[0370] For all-solid-state batteries, the electrode and electrolyte layers are manufactured by a dry process that uses polytetrafluoroethylene (PTFE) as a binder and allows obtaining films with a typical thickness of 100-200 μm. 12 Two different types of all-solid-state batteries were assembled using either lithium ion (LTO) or lithium (Li) metal as the anode. In either case, the active materials (LiCo 0.5 Mn 1.5 O4、LiNi 0.5 Mn 1.5 O4 or LiCoO2) and Li 10 GeP2S 12 The composite cathode was prepared by mixing LGPS (LGPS) powder and 3% additional PTFE. This mixture was then rolled into a film. On the one hand, for those all-solid-state batteries using LTO as the anode, a separator with a weight ratio of LGPS and PTFE membrane of 95:5 was used. The anode composition consisted of a mixture of LGPS, LTO and carbon black in a weight ratio of 60:30:10 and 3% additional PTFE. Finally, the cathode membrane (using LiCo 0.5 Mn 1.5 O4、LiNi 0.5 Mn 1.5O4 or LiCoO2 as active material) / LGPS membrane / LTO membrane. The specific capacity was calculated based on the amount of LTO in the anode film (30 wt%). Constant current battery cycling tests were carried out on an Arbin BT2000 workstation at room temperature. On the other hand, when lithium metal was used as the anode, a Li metal foil with a diameter and thickness of 1 / 2" and 40μm, respectively, was connected to the current collector. In order to prevent interfacial side reactions, the Li foil was covered with a carbon black film with a diameter of 5 / 32" with a weight ratio of carbon black to PTFE of 96:4. After the negative electrode was loaded into the Swagelok battery cell, 70mg of pure LGPS powder acting as a separator was added and slightly pressed. Finally, ~1mg of the cathode composite material LCMO membrane was inserted and pressed to 6Tn (0.46GPa) to form a battery, whose final configuration was LCMO / LGPS pellets / graphite film + Li metal. For Fig.50A For the high voltage test in Figure 50, the battery was charged to 0.3 C, then rested for 30 minutes and discharged at 0.1 C. All batteries in Figure 50 were tested at an elevated temperature of 55°C.

[0371] Computational simulation

[0372] All ab initio calculations and phase data were obtained with the Vienna Ab-initio Software Package (VASP) following the Materials Project Calculation Guide. Mechanically induced metastability calculations were performed following the LaGrangian optimization method outlined in Small 1901470, 1-14 (2019) and J. Mater. Chem. A (2019). doi: 10.1039 / C9TA05248H). Pseudophase calculations were performed following the methods of J. Mater. Chem. A4, 3253-3266 (2016), Chem. Mater. 28, 266-273 (2016) and Chem. Mater. 29, 7475-7482 (2017).

[0373] Further embodiments are described in the claims.

Claims

1. A rechargeable battery comprising a first electrode, a second electrode and a solid electrolyte disposed therebetween, wherein: The solid electrolyte comprises a sulfide containing an alkali metal, The rechargeable battery includes a coating disposed on at least one of the first electrode or the second electrode, the coating including at least one of LiNbO3, AlF3, MgF2, Al2O3, SiO2, Si, Ag, C, or graphite, the coating forming an interface between at least one of the first electrode or the second electrode and the solid electrolyte; and an external stress on the battery in the range of 1 MPa to 50 MPa, causing the solid electrolyte to be subjected to mechanical contraction, wherein the solid electrolyte is pre-compressed at a pressure of about 70 MPa to about 1,000 MPa prior to inclusion in the battery, wherein "about" means ±10%; and The combination of coating, pre-compression and external stress is configured to increase the stability of the cell to suppress dendrite growth.

2. The rechargeable battery of claim 1, wherein the solid electrolyte has a core-shell morphology.

3. The rechargeable battery of claim 1, wherein the alkali metal is Li, Na, K, Rb or Cs.

4. The rechargeable battery of claim 1, wherein the solid-state electrolyte comprises SiPS, GePS, SnPS, PSI or PS.

5. The rechargeable battery of claim 1, wherein the solid electrolyte is Li 10 SiP2S 12 , Li 10 GeP2S 12 or Li 9.54 Si 1.74 P 1.44 S 11.7 Cl 0.3 .

6. The rechargeable battery of claim 1, wherein the first electrode is a cathode and comprises LiCoO2, LiNi 0.5 Mn 1.5 O4, VLi2CoPO4F, LiNiPO4, Li2Ni(PO4)F, LiMnF4, LiFeF4 or LiCo 0.5 Mn 1.5 O4.

7. The rechargeable battery of claim 1, wherein the second electrode is an anode and comprises lithium metal, lithiated graphite, or Li4Ti5O 12 .

8. The rechargeable battery of claim 1, wherein the solid electrolyte has an effective modulus of about 1 GPa to about 100 GPa, wherein "about" means ±10%.

9. The rechargeable battery of claim 1, wherein the coating comprises graphite, and the alkali metal and the graphite form a composite material.

10. A method of storing energy comprising applying a voltage across a first electrode and a second electrode and charging the rechargeable battery of any one of claims 1 to 9.

11. A method of supplying energy, comprising connecting a load to a first electrode and a second electrode and allowing the rechargeable battery of any one of claims 1 to 9 to discharge.

12. The rechargeable battery of claim 1, wherein the dendrites grow at a rate of up to 10 mA cm -2 is suppressed at a cycling current density of .

13. The rechargeable battery of claim 1, wherein the coating is configured to form a composite material with a corresponding electrode of at least one of the first electrode or the second electrode during cycling.

Images