Recent progress on anode for sulfide-based all-solid-state lithium batteries

—— Part 1 Lithium metal anode

Author:

JIA Linan, DU Yibo, GUO Bangjun, ZHANG Xi

1. School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200241, China

2. Shanghai Yili New Energy Technology Co. , LTD. , Shanghai 201306, China

Abstract

All-solid-state lithium batteries (ASSLBs) exhibit higher energy density and more safety than current liquid lithium batteries, which are the main research direction for next-generation energy storage devices. Compared with other solid-state electrolytes, sulfide solid-state electrolytes (SSEs) have the characteristics of ultra-high ionic conductivity, low hardness, easy processing, and good interfacial contact, which are one of the most promising routes to realize all-solid-state batteries. However, there are some interfacial issues between anodes and SSEs that limit their applications such as interfacial side reactions, poor rigid contact, and lithium dendrite. This study outlines the current progress in anode materials used for sulfide-based ASSLBs, summarizes the development status, application advantages, interface problems and mainstream solution strategies of the main anode materials including lithium metal, lithium alloys, silicon anode for sulfide-based ASSLBs, and provides guiding suggestions for the next development of anode materials and the solution of interfacial issues.

Keywords： all-solid-state lithium batteries ; sulfide electrolyte ; lithium anode ; alloy anode ; anode/electrolyte interfaces

Introduction

Lithium-ion batteries are widely used in various portable devices due to their high voltage and high energy density. They are a key industrial product for vehicle electrification and the deployment of energy storage systems in a low-carbon society. However, liquid lithium-ion batteries use graphite negative electrodes, organic liquid electrolytes, and metallic lithium oxide positive electrodes (such as LiCoO2). On the one hand, the specific energy of the assembled batteries is limited to the range of 200~250 W·h·kg-1, making it difficult to achieve further breakthroughs in specific energy. On the other hand, organic electrolytes have disadvantages such as poor thermal stability and flammability. Moreover, the lithium dendrites generated during the battery cycle will also bring huge risks of battery short circuit or even explosion. This series of problems has caused many researchers to pay attention to and think about the safety of lithium-ion batteries. Replacing flammable organic liquid electrolytes with solid electrolytes can fundamentally prevent thermal runaway and solve the safety hazards caused by flammable liquid electrolytes used in liquid lithium-ion batteries. At the same time, the high mechanical properties of solid electrolytes are also considered to be one of the breakthroughs in inhibiting the growth of lithium dendrites.

Currently, the mainstream solid-state electrolytes include four types: sulfide solid-state electrolyte, oxide solid-state electrolyte, polymer solid-state electrolyte and halide solid-state electrolyte. Among them, oxide electrolytes have the advantages of good stability and moderate ionic conductivity, but have poor interface contact. Polymer electrolytes have good stability to lithium metal and have relatively mature processing technology, but poor thermal stability, narrow electrochemical windows, and low ionic conductivity limit the scope of application. As a new type of electrolyte, halide electrolytes have received widespread attention due to their high ionic conductivity. However, the high valence metal elements in halide electrolytes determine that they cannot directly contact lithium metal to form a stable anode interface. Research on halide electrolytes requires further exploration. Sulfide electrolytes are considered to be one of the most promising routes to realize all-solid-state lithium batteries (ASSLBs) electrolytes due to their high ionic conductivity, low hardness, easy processing, good formability, and good interface contact.

In recent years, related research on sulfide electrolytes has been further developed, and its ionic conductivity has reached a level comparable to that of liquid organic electrolytes. Typical sulfide electrolytes include glassy Li-P-S sulfide (LPS) and derivative glass ceramics, silver sulfide germanium ore (Li6PS5X, X=Cl, Br, I), and lithium sulfide ion superconductors (thio-lithium superionic conductor, thio-LISICONs), Li10GeP2S12 (LGPS) and similar compounds.

Among these different sulfide materials, LGPS-type electrolytes show by far the best ionic conductivity. In 2016, Kato et al. reported the superlithium ion conductor Li9.54Si1.74P1.44S11.7Cl0.3 (LSPSCl), whose ionic conductivity is as high as 25×10-2 S·cm-1 at room temperature. LGPS also has an ultra-high ion conductivity of 1.2×10-2 S·cm-1 at room temperature. The weak anisotropic ion conductivity of single crystal LGPS in the (001) direction even reaches 27×10-2 S·cm-1. Glass ceramics (Li7P3S11) and sulfide-germanite (Li6PS5Cl) can achieve high ionic conductivities of 10-3 S·cm-1. All-solid-state batteries combining sulfide electrolytes with high-nickel layered cathodes and high-energy anodes (such as Si or metallic lithium) can even exhibit ultra-high specific energy of 500 kW·h·kg-1. However, the application of sulfide electrolytes in all-solid-state lithium batteries still has problems such as narrow electrochemical window, poor electrode-electrolyte interface stability, poor air stability, lack of large-scale manufacturing methods, and high cost. The narrow electrochemical window determines that the reduction reaction of the electrolyte will occur when the active sulfide electrolyte comes into contact with most negative electrodes, resulting in interface instability, which is an important bottleneck restricting the development of all-solid-state lithium batteries. This article mainly summarizes the development status of mainstream anode materials for all-solid-state lithium batteries based on sulfide electrolytes, and further summarizes the interface problems and solution strategies between sulfide solid electrolytes and anode materials. Provide guiding suggestions for the development and commercial application of all-solid-state lithium batteries based on sulfide electrolytes.

1 Lithium metal anode

1.1 Lithium/sulfide electrolyte interface chemical stability

Fig.1   Types of interfaces between lithium metal and solid-state electrolyte

(1) Thermodynamically stable interface: As shown in Figure 1(a), the two phases in contact are in a state of thermodynamic equilibrium. Metal lithium does not react at all with the electrolyte, forming a sharp two-dimensional plane, such as LiF, Li3N and other lithium binary compounds.

(2) Thermodynamically unstable interface: Due to the thermodynamically driven chemical reaction between the contacting electrolyte and electrode, a three-dimensional interface layer can be formed. Depending on whether the interface layer formed by the reaction product has sufficient electronic and ionic conductivity, it can be further distinguished into the following two interfaces.

①Mixed conductive interface layer: When the product has sufficient electronic and ionic conductivity, the interface layer can stably grow into the solid electrolyte. The formation of this hybrid conductive interlayer will ultimately allow electron transport through the electrolyte, leading to self-discharge of the battery [Figure 1(b)]. The interfacial instability of sulfide solid electrolytes leads to the generation of interfacial side reactions, which can cause rapid attenuation of battery capacity or even failure. Wenzel et al. used in situ X-ray photoelectron spectroscopy (XPS) combined with time-resolved electrochemical measurements. Detailed information on the chemical reaction at the interface between LGPS and metallic lithium is provided, and it is verified that the decomposition of LGPS leads to the formation of a solid electrolyte interface phase composed of Li3P, Li2S, and Li-Ge alloys. Among them, Li3P and Li2S are ionic conductors, and Li-Ge alloy is an electronic conductor. The mixed conductive interface layer formed will cause LGPS to continue to decompose, and the negative electrode interface impedance will continue to increase, eventually leading to battery failure.

②Metastable solid electrolyte interface layer: If the reaction product is non-conductive or has only low electronic conductivity, the interface layer can be limited to grow into a very thin film, and a stable solid-state electrolyte interphase, SEI, may be formed. . As shown in Figure 1(c), the performance of this battery will depend on the ion conduction properties of the SEI. The sulfide-germanite type electrolyte is relatively stable, and its decomposition products Li2S, Li3P and LiX (X=Cl, Br and I) have low enough electronic conductivity to avoid continued decomposition of the electrolyte and easily form a stable SEI. At the same time, Li3P has high ionic conductivity, ensuring efficient transmission of lithium ions in solid-state batteries.

1.2 Research on the mechanical properties of lithium metal

The current solid-solid interface contact between the negative electrode and the solid electrolyte is a limited point contact, which easily leads to an increase in interface resistance. However, the mechanical properties of metallic lithium, especially the creep of metallic lithium, will further affect the interface contact effect, leading to the formation of interface voids and even negative electrode delamination at high current densities. Therefore, studying the mechanical properties of metallic lithium, especially the creep behavior of metallic lithium, is crucial to the cycle stability of all-solid-state batteries.

Tian et al. conducted contact mechanics research and established relevant theoretical models to obtain the boundary conditions that affect the stress distribution function of elastic, plastic and viscous contacts on the lithium metal anode. Predict the contact area of the metallic lithium-sulfide solid electrolyte interface and calculate the capacity loss caused by ion diffusion at the interface and loss of contact area. Experiments show that at a lower cut-off voltage (3.8 V), the relationship between the decrease in battery capacity and the loss of contact area is almost linear, with a slope of 1. While at a higher cutoff voltage (4.0 V), the slope is less than 1, and the capacity drop rate decreases with increasing discharge rate. Fincher et al. used tensile experiments to test the mechanical effects of commercial lithium foil and found that the yield strength of metallic lithium ranged from 0.57 to 1.26 MPa at a strain rate of 5×10-4~5×10-1 s-1. For the indentation test with a target of 0.05 s–1, the hardness dropped sharply from nearly 43.0 MPa to 7.5 MPa as the indentation depth increased from 250 nm to 10 µm. The plastic properties measured from nanoindentation tests showed strong strain rate dependence with stress exponents of 6.55 and 6.90 respectively. Finite element analysis is used to relate indentation depth to relevant length scales in battery applications. It can provide important guidance for optimizing the structure of lithium anodes and ensuring charge and discharge stability, so as to reduce the uneven deposition of lithium during electrochemical cycles. Masias et al. systematically measured the elastic, plastic and time-dependent mechanical properties of polycrystalline lithium at room temperature. Its Young's modulus, shear modulus and Poisson's ratio were determined to be 7.82 GPa, 2.83 GPa and 0.38 respectively, and the yield strength was between 0.73 and 0.81 GPa. Power-law creep dominates under tension, with a stress index of 6.56. Compression testing was performed within the battery-relevant stress range (0.8~2.4 MPa), and significant banding and a decrease in strain rate with time were observed. Narayan et al. established a response model for an all-solid-state battery lithium anode based on large deformation theory, simulating the interaction between the lithium anode and the sulfide solid electrolyte in the elastic-viscoplastic reaction of lithium. It shows that the reaction of strain is related to the volume deformation of lithium anode, which is the main reason for the failure of solid-state batteries. Through batch tensile and nanoindentation tests, lithium metal shows obvious strain rate dependence and size decay during creep. showed that fine-tuning of deformation mechanics can be achieved by adjusting lithium deposits to improve the robustness of the lithium anode and mitigate unstable lithium growth during electrochemical cycling.

In addition to the overall mechanical study of metallic lithium, the study of nanomechanics provides quite important and extremely detailed surface and local information at small scales. Nanoindentation experiments are one of the most commonly used analysis tools for surface and local characteristics. Nanoindentation experiments performed in inert gas can more comprehensively analyze the mechanical, electrochemical, and morphological coupling behaviors of metallic lithium. Herbert et al. conducted a series of nanoindentation experiments on high-purity evaporated lithium films and collected data on plastic flow characteristics, including elastic modulus, hardness and yield strength. The evolution of the above data with key variables such as length scale, strain rate, temperature, crystallographic orientation and electrochemical cycling was studied, indicating that the plastic flow of lithium is mainly related to steady-state creep under constant load or pressure. The creep of lithium during electrochemical charging and discharging can induce buckling at the interface and generate additional stress. At the same time, the viscoplastic behavior of lithium will further affect the interface contact area, leading to the deterioration of ion diffusion channels and interface instability. However, the current nanomechanical research on metallic lithium is still in its preliminary stage, and further research is very important. Some new technologies such as nanocolumn compression and in-situ real-time observation of metallic lithium nanomechanics have also been proposed to analyze the coupling of the metallic lithium anode interface and provide high-fidelity information about the interface to further understand the mechanical coupling effect of metallic lithium, thus providing the possibility for the design of nanoscale metallic lithium anodes.

1.3 Nucleation and growth of lithium dendrites

Lithium dendrites are one of the fundamental issues affecting the stability and safety of lithium-ion batteries. Solid electrolytes have long been considered as a potential solution to lithium dendrite growth due to their high mechanical strength. However, numerous research results show that the problem of lithium dendrites in solid electrolytes still exists, and is even more serious than in liquid lithium batteries. In solid-state batteries, there are many reasons for the growth of lithium dendrites, including uneven contact at the interface between the electrolyte and metallic lithium, defects, grain boundaries, voids within the electrolyte, space charges, etc. Monroe et al. reported a lithium dendrite growth model based on metallic lithium anode and solid electrolyte. Factors such as electrolyte elasticity, compression force, surface tension and deformation force were considered in the model. Simulation results show that when the shear modulus of the electrolyte is equivalent to that of lithium, a stable interface will be formed. When the shear modulus of the electrolyte is approximately twice that of lithium (4.8 GPa), the generation of lithium dendrites can be suppressed. However, in actual all-solid-state lithium battery research, it was found that lithium dendrites are still produced in solid electrolytes with high shear modulus [such as Li7La3Zr2012 (LLZO), elastic modulus ≈ 100 GPa]. Therefore, this model is only applicable to ideal interfaces without any microscopic defects and uneven distribution. Porz et al. found that the high shear modulus of the electrolyte will lead to high ultimate current density, inducing the nucleation and growth of metallic lithium in the grain boundaries and voids of the solid electrolyte. Nagao et al. used in-situ scanning electron microscopy to observe the lithium deposition and dissolution process at the negative electrode interface in all-solid-state lithium batteries, revealing the changes in the lithium deposition morphology with different applied current densities. When the current density exceeds 1 mA·cm-2, local lithium deposition will cause larger cracks, resulting in a reduction in the reversibility of lithium deposition and dissolution, and the cracks will further expand until lithium dendrites are formed. On the other hand, uniform and reversible lithium deposition and dissolution can be achieved at a low current density of 0.01 mA·cm-2, with almost no cracks. Therefore, focusing only on the high shear modulus of the electrolyte cannot solve the problem of lithium dendrite growth, and may reduce the ionic conductivity of the electrolyte and affect the energy density of solid-state batteries.

Porz et al. studied the nucleation and growth mechanism of lithium dendrites in various electrolytes and showed that the onset of lithium penetration depends on the surface morphology of the solid electrolyte. In particular, the size and density of defects, and the deposition of lithium in defects can create tip stresses that drive crack propagation. In addition, differences in conductivity between grains, grain boundaries or interfaces can also lead to the generation of lithium dendrites. Yu et al. theoretically studied the energetics, composition and transport properties of three low-energy symmetrically tilted grain boundaries in solid electrolytes. It shows that the transport of lithium ions at grain boundaries is more difficult than in grains and is sensitive to temperature and grain boundary structure. Raj et al. theoretically studied the effect of grain boundary resistance on the nucleation of lithium dendrites at the solid electrolyte/lithium interface. They proposed that the high ionic resistivity of the grain boundaries and the physical irregularities of the anode interface would lead to an increase in the local electrochemical mechanical potential of lithium, thereby promoting the formation of lithium dendrites. Therefore, compared with crystal grains, grain boundaries with high ion resistivity are more likely to induce the nucleation and growth of lithium dendrites. The growth mechanism of lithium dendrites in all-solid-state batteries has gradually become clearer with further research. However, there is still a lack of effective ways to completely suppress lithium dendrites, and related research needs to continue to be in-depth to realize the application of metallic lithium anodes in all-solid-state batteries as soon as possible.

1.4 Interface problem solving strategies

Many methods have been proposed to solve the challenges in the application of lithium anodes, including applying external pressure, using SEI layers, optimization of electrolytes, and modification of metallic lithium. This reduces the impact of lithium creep on the battery, increases the contact area of the solid-solid interface, inhibits side reactions at the interface between the sulfide solid electrolyte and the metallic lithium anode, improves the lithophilicity of the anode interface, and avoids the formation and growth of lithium dendrites.

1.4.1 Apply external pressure

Applying external pressure can increase the contact area of the solid-solid interface, reduce the damage caused by creep to the negative electrode interface, and improve the cycle stability of the battery. Zhang et al. reported a multi-scale three-dimensional time-dependent contact model to describe the evolution of the solid electrolyte/lithium anode interface under stack pressure. Theoretical calculations show that high stack pressures of about 20 GPa tend to inhibit the formation of voids, a promising method to ensure consistent interface contact, potentially achieving stable battery performance. Higher stack pressure is not more beneficial to battery performance. Lower stack pressure cannot fundamentally solve the contact problem at the solid-solid interface. Excessive stack pressure can easily form lithium dendrites and cause short circuits in the battery. Wang et al. studied the effect of stack pressure on the performance of lithium/sulfide electrolyte batteries and found that during the lithium stripping process, the maximum allowable stripping current density is proportional to the applied external pressure. During the deposition process, higher applied pressure will reduce the maximum allowable deposition current, that is, high stacking pressure will easily lead to the generation of lithium dendrites (Figure 2).

Fig.2   Relationship between maximum allowed current density (MACD) and external pressure for stripping and deposition in ASSLBs

1.4.2 Artificial solid electrolyte interface layer

Placing a stable SEI at the sulfide solid electrolyte/lithium interface can avoid direct contact between metallic lithium and the sulfide solid electrolyte, effectively inhibiting the occurrence of interface side reactions and the formation and growth of lithium dendrites. Generally, there are two methods of forming SEI: in-situ SEI and ex-situ SEI. Wang et al. established an in-situ ion conductive protective layer on the surface of polished lithium metal through spin coating technology. A mixture of polyacrylonitrile (PAN) and fluoroethylene carbonate (FEC) is used to embed an artificial protective layer (LiPFG) composed of an organic matrix of inorganic Li3N and LiF on the lithium surface. Effectively promotes uniform deposition of lithium and improves interface stability and compatibility. Li et al. designed an in-situ polymerized interlayer of 1,3-dioxolane in lithium difluoro(oxalate)phosphate. The SEI formed at the Li/LGPS interface has a double-layer structure. The upper layer is rich in polymers and is elastic, and the lower layer is full of inorganic substances to inhibit the nucleation and growth of lithium dendrites. At the same time, the seamless contact of the Li/LGPS interface is achieved, which promotes the uniform transmission of lithium ions and inhibits the continuous decomposition of LGPS. Lithium symmetric batteries with this gel polymer coating exhibit stable cycling over 500 h under the conditions of 0.5 mA·cm-2/0.5 mA·h·cm-2. Gao et al. reported a nanocomposite based on organic elastic salts [LiO-(CH2O) n -Li] and inorganic nanoparticle salts (LiF, -NSO2-Li, Li2O), which can be used as an intermediate phase to protect LGPS. The nanocomposite material is formed in situ on Li through the electrochemical decomposition of liquid electrolyte, which reduces the interface resistance, has good chemical and electrochemical stability and interface compatibility, and effectively inhibits the occurrence of LGPS reduction reaction. Stable lithium deposition of more than 3000 h and a cycle life of 200 times were achieved. The mechanical strength of SEI is extremely important for the cycle stability of all-solid-state batteries. If the mechanical strength of SEI is too low, dendrite penetration will occur. If the SEI is not tough enough, bending cracking will occur [Fig. 3(a)]. Duan et al. prepared a structured LiI layer through chemical iodine vapor deposition as an artificial SEI between metallic lithium and LGPS [Figure 3(b)]. The LiI layer generated in situ has a unique, slender rice-shaped LiI crystal intertwined structure, which provides high mechanical strength and excellent toughness, and can effectively inhibit the growth of lithium dendrites. and adapts well to changes in lithium volume, thereby maintaining a strong Li/LGPS interface [Figure 3(c)]. At the same time, this LiI layer has high ionic conductivity and certain chemical inertness, and shows high stability to both lithium and LGPS. The prepared Li/LiI/LGPS/S battery showed a high capacity of 1400 mA·h·g-1 at 0.1 C, and showed a high capacity retention rate of 80.6% after 150 cycles at room temperature. Even under harsh conditions of 1.35 mA·h·cm-1 and 90°C, it still exhibits a high capacity of 1500 mA·h·g-1 and excellent stability for 100 cycles. Showing its great potential in various application scenarios. Based on the solution method, Liang et al. synthesized a Li x SiS y layer in situ on the surface of metallic lithium as SEI to stabilize the Li/Li3PS4 interface. This Li x SiS y layer is air-stable and can effectively prevent side reactions between lithium and the surrounding environment. It can be stably cycled for more than 2000 hours in a symmetrical battery. The team also reported a solution strategy using polyacrylonitrile-sulfur composites (PCE) as an ex-situ artificial SEI. Using PCE as an intermediate layer at the interface between lithium metal and LGPS significantly suppresses the interface reaction between LGPS and Li metal. The assembled all-solid-state battery exhibits high initial capacity. 148 mA·h·g-1 at 0.1 C rate. It is 131 mA·h·g-1 at 0.5 C rate. The capacity remains 122 mA·h·g-1 after 120 cycles at 0.5 C rate. Demonstrate excellent performance.

Fig.3   Schematic diagram of interface between LGPS and Li anode

1.4.3 Electrolyte optimization

Electrolyte optimization can not only improve the ionic conductivity of the sulfide electrolyte, but also avoid or reduce the reduction of the electrolyte by the lithium anode to a certain extent. Among them, using appropriate element substitution is an effective strategy to improve ionic conductivity and stabilize the anode interface. Experiments by Sun et al. show that oxygen doping can increase ion conductivity (Li10GeP2S11.7O0.3: 8.43×10-2 S·cm-1; LGPS: 1.12×10-2 S·cm-1). At the same time, interfacial reactions are prevented, thereby improving the stability of the lithium/sulfide electrolyte interface. In addition to oxygen, metal sulfide doping can also reduce the impedance of the lithium/sulfide electrolyte interface. For example, Li7P2.9S10.85Mo0.01 (improved Li2S-P2S5 glass ceramics using MoS2 doping) exhibits lower interface impedance than L7P3S11. Li3.06P0.98Zn0.02S3.98O0.02 (ZnO doped in Li3PS4) also shows good cycle stability (100 cycles capacity retention rate of 81%, bare Li3PS4 is only 35%). Although appropriate element substitution has shown good results for the lithium/sulfide electrolyte interface. However, these modification methods still have problems such as the occurrence of side reactions and the formation of lithium dendrites during long cycles. The upper limit of the role of kinetics on interface issues should be further confirmed, and other strategies should be combined to improve the chemical stability of the lithium/sulfide electrolyte interface. The electrolyte structure design can also inhibit the occurrence of side reactions and prevent the nucleation and growth of lithium dendrites. Ye et al. proposed an ingenious design of a sandwich-structured electrolyte [Figure 4(a)]. Sandwiching the unstable electrolyte between more stable electrolytes avoids direct contact through good local decomposition in the layer of the less stable electrolyte. It can both prevent the growth of lithium dendrites and fill the generated cracks. This expansion screw-like design concept achieves a stable cycle of metallic lithium anode paired with LiNi0.8Mn0.1Co0.1O2 cathode [as shown in Figure 4(b), the capacity retention rate is 82% after 10,000 cycles at 20 C ]. More importantly, this work is not limited to specific materials. Stable cycles can be observed using LGPS, LSPSCl, Li9.54 Si1.74P0.94S11.7Cl0.3 (LSPS), Li3YCl6, etc. as central layer materials. It provides a highly applicable design method to improve the stability of the lithium anode/sulfide electrolyte interface.

Fig.4   Schematic diagram of sandwich structure electrolyte design and long cycle electrochemical performance curve

1.4.4 Lithium anode modification

Modification of the lithium anode can reduce or avoid the occurrence of electrolyte cracks caused by the creep behavior of metallic lithium during cycling, thereby inhibiting the formation of lithium dendrites. As shown in Figure 5, Su et al. used a graphite film to protect the lithium negative electrode, separate the LGPS electrolyte layer from the lithium metal, and inhibit the decomposition of LGPS. Based on the mechanical shrinkage mechanism, an external pressure of 100~250 GPa is applied to the battery system. This external force constraint optimizes the interface contact between electrolyte particles and between the electrolyte layer and the Li/G anode. The all-solid-state battery achieves excellent cycle performance. In addition, alloying metallic lithium is also an important way to solve the lithium anode interface problem of all-solid-state lithium batteries. In current reports, lithium alloys have shown certain advantages in solving problems such as serious interface side reactions and the generation of lithium dendrites in lithium anodes, which will be introduced in detail below.

Fig.5   Protection design of graphite film for Li/LGPS interface