Development of Si-Based Anodes for All-Solid-State Li-Ion Batteries

Abstract

All-solid-state Li-ion batteries (ASSBs) promise higher safety and energy density than conventional liquid electrolyte-based Li-ion batteries (LIBs). Silicon (Si) is considered one of the most promising anode materials due to its high specific capacity (3590 mAh g⁻¹) but suffers from poor cycling performance because of large volumetric effects leading to particle pulverization, unstable solid electrolyte interphase (SEI), and electric disconnection. In ASSBs, additional issues such as poor solid–solid contacts and interfacial side reactions between Si and solid-state electrolytes (SSEs) are also hindering their practical application. This review first outlines the prospects and recent research achievements of Si-based anodes with special focuses on various Si structures and composite materials, then analyzes the issues of electrochemical–mechanical effects, and finally summarizes key factors and promising strategies for further improving Si-based anodes for high-performance ASSBs.

1. Introduction

Li-ion batteries (LIBs) are widely used in portable electronic products and electric vehicles (EVs) because of their merits of high energy density (HED), a long cycle life, and a low memory effect [1,2,3]. Traditional LIBs based on organic liquid electrolytes have safety hazards and are approaching their energy density limits. In comparison, all-solid-state batteries (ASSBs) using less flammable solid-state electrolytes (SSEs) and enabling aggressive high-capacity active materials can meet the persistent pursuit of high safety performance, HED from battery manufacturers, and end consumers [4,5].

Apart from cathode materials, anode materials play a decisive role in the key performance indicators, especially energy density, of LIBs [6,7]. Candidates of anode materials for ASSBs include graphite, Li metal, and alloy anodes such as Si, etc. [8,9,10]. Graphite is stable, low-cost, and is the dominating anode material in commercial LIBs. However, with a theoretical specific capacity as low as 372 mAh g⁻¹, graphite presents a limit on the energy density of LIBs [11]. Li metal poses an ultra-high specific capacity (3860 mAh g⁻¹) and a low electrochemical potential (−3.04 V vs. SHE) and can greatly increase the energy density of the battery when matched with high-voltage cathode materials [12]. However, it compromises the inherent safety promise of ASSBs because of their high flammability and high-probability dendrite formation which can lead to internal short circuits (ISCs) in the cell. In contrast, Si also has a high specific capacity (3590 mAh g⁻¹) but is chemically much more stable and environmentally friendly and thus is also regarded as a promising and practical anode candidate for HED ASSBs, especially in the near-to-medium term. As shown in Figure 1a, compared to Li metal, Si is advantageous with high annual production and low cost [13]. On the other hand, while Li metal can undergo severe chemical reactions with some SSEs (e.g., sulfide SSEs) [14], forming by-products with low ionic conductivities, Si is chemically much more stable and thus compatible with most SSEs [15]. Moreover, Li metal and Si behave differently under high stacking pressures that are generally required in the ASSB manufacturing process to realize close solid–solid contacts [16]. At pressures above 25 MPa [17], Li metal may propagate through SSEs and lead to short circuits with the cathodes, whereas Si can bear much higher pressures because of its high Young’s modulus of 130 GPa [18]. In addition, Li metal is very sensitive to ambient environment and must be processed in gloveboxes or dry rooms [19]. Conversely, Si has good air stability and is suitable for large-scale manufacturing.

Nonetheless, Si has its problems, which mostly originate from the alloying lithiation mechanism and resulting huge volumetric change (up to ≥300%). These problems are illustrated in Figure 1b and include (1) instability at the material level: Si is prone to particle pulverization during the charging and discharging process because of high internal stress; (2) instability at the electrode level: the repeated expansion and contraction of the electrode as a whole may lead to electric disconnection and isolation; and (3) instability at the interface: the volume effect may cause the repeated rupture and excessive growth of the surface solid electrolyte interphase (SEI) film, which consumes active Li and increases interfacial impedance. The above three unstable processes collectively account for the generally observed poor cycling performance of Si anodes. In addition, (4) the rate performance of Si is generally not good because of its low electric conductivity and low Li-ion diffusion coefficient, compared to conventional anode materials such as graphite.

These issues could be intensified or relieved when Si anodes are used in ASSBs. In 2009, Lee et al. first investigated the application of Si nanoparticles in ASSBs with a sulfide SSE [22]. In this work, Si nanoparticles demonstrated higher capacity retention in ASSBs than in liquid batteries after the same number of cycles. The volume expansion of Si material during cycling is inevitable. However, sulfide-based SSEs with good mechanical ductility may be beneficial in buffering the volume and stress effects of Si anodes. Moreover, the cracks formed in the electrodes were reported to be vertical to the current collectors and may not necessarily cause electric connection failure [19,23]. The problem of the surface instability of the Si anode related to excessive SEI growth could also be alleviated in ASSBs. In 2021, Meng’s group first reported an ASSB with a carbon-free Si anode that can achieve a capacity of 2800 mAh g⁻¹, a capacity retention of 80% over 500 cycles, and an average columbic efficiency (CE) of >99.9% at room temperature [21]. As shown in Figure 1c, unlike liquid electrolytes, SSEs do not completely wet the electrode active material, and therefore the excessive growth of the SEI and Li-ion consumption during cycling could be largely suppressed by the reduced Si/SSE interface area.

Stimulated by the urgent demand for HED and safe ASSBs, there are some review studies on Si-based anodes for ASSBs from different perspectives. Wang et al. [24] and Pilgun Oh et al. [14] reviewed the developments of HED anodes for ASSBs based on sulfide SSEs. Lewis et al. summarized the progress and promise of alloy anode materials in ASSBs [25]. Hanyu Huo et al. examined Si-based anodes for ASSBs with different cell and electrode types [26]. Liu et al. discussed the challenges and solutions for Si-based anodes in various LIBs including liquid, quasi-solid, and all-solid batteries [27]. There are also several review reports on Si-based anodes for solid-state batteries with a focus on different SSEs [28,29,30]. However, only a few reviews mainly focus on the engineering and modification of Si, as well as fundamental issues in Si-based ASSBs.

In this review, we summarize the current development and application of Si anodes in ASSBs. The development of Si structures is first introduced, and this includes μm-, nm-, porous Si, thin-film Si, and columnar Si. In addition, we discuss the composition and alloying of Si such as Si/C and Li-Si alloys, etc. Later, we highlight the main factors that could impact the performance of Si in ASSBs, including interface issues and the electrochemical–mechanical coupling effect between Si and SSEs. Finally, we propose some prospects and strategies for the current hindrance that may put forward the research of Si in ASSBs.

2. Si Anodes

2.1. Si Anodes with Different Particle Sizes

The size of Si particle anodes is a significant factor influencing their accessible capacities and cycling performances. It has been found that by reducing the size of Si particles to nanometers, the stress concentration caused by volume expansion can be effectively relieved, which can prevent Si from pulverization and powdering. However, the high specific surface area of nm-Si could also bring more side reactions consuming more electrolytes and active Li ions. In addition, the large-scale production of nm-Si generally involves complicated processes and demands equipment and thus leads to high costs.

In 2018, Dunlap et al. [31] investigated the effect of Si particle sizes on the electrochemical performance of Si/C anodes. They found that the Si/C anode prepared with 50 nm Si has a higher first-cycle capacity and cycling stability than the Si/C anodes with μm-Si (1–3 μm and 325 mesh Si) because the amorphous carbon matrix was not robust enough for μm-Si (Figure 2a). Pores and cracks were observed in electrodes with Si/C anodes with μm-Si, which could increase the cell impedance. Trevey et al. [32] compared the cycling performances of nm-Si (50–100 nm) and μm-Si (1–5 μm) in ASSBs with sulfide SSEs (Figure 2b). At the same cutoff voltages, the nm-Si anode exhibits a higher discharge capacity and improved cycling performance over the μm-Si anode. Rana et al. [33] investigated the influence of the Si particle size on the rate performance and found that the rate performance as well as the first-cycle capacity of the Si anode improves as the particle size decreases. The surface-to-volume ratio of active materials plays a decisive role in the rate performance. While the larger sizes of μm-Si may be advantageous for achieving a higher effective ionic conductivity of the electrodes [34,35,36], nm-Si with a high surface area-to-volume ratio can provide more contact sites, more effective reaction interfaces, and thus can achieve higher specific capacity (Figure 2c,d).

Takahashi et al. [23] presented a slurry-mixing method for fabricating Si-75Li₂S-25P₂S₅(LPS) electrodes applicable to binder-free sheet-type ASSBs. Unlike powder-pressed Si composite anodes, this method is suitable for the current manufacturing process for traditional LIBs and is promising for large-scale production [26,37,38]. The μm-Si-LPS sheet can deliver a discharge capacity of 3058 mAh g⁻¹ and an initial coulombic efficiency (ICE) of 90%, while the nm-Si-LPS anode using the same preparation method delivers a higher capacity of 3168 mAh g⁻¹ but a lower ICE of 86% (Figure 3a,b). The lower ICE of nm-Si can be ascribed to its higher specific surface area and irreversible side reaction from the surface oxide layer [39]. Jung et al. [38] came to the same conclusion by fabricating a sheet-type Si composite anode with a different method of infiltrating electrodes with SSEs in a Li₆PS₅Cl/ethanol solution (Figure 3c). Both μm-Si and nm-Si anodes showed high first-cycle discharge capacities over 3000 mAh g⁻¹; however, the nm-Si anode delivered a lower ICE than the μm-Si anode. nm-Si demonstrates good electrochemical performances in sulfide-based full-cell ASSBs. In 2022, Zhu et al. [19] designed a high-energy ASSB based on a nm-Si composite anode, a Li₂SiO_x@S-NMC composite cathode, and a thin sulfide SSE membrane, which shows great potential for industrial applications. The full-cell exhibited remarkable cell-level energy densities of 285 and 177 Wh kg⁻¹ at current densities of 0.158 and 3.16 mA cm⁻² (Figure 3d). When cycled at C/3, the cell showed good cycling stability for 1000 cycles (Figure 3e). There are also some reports on the performance of full-cell ASSBs using μm-Si. Meng et al. [21] developed an ASSB that utilized a μm-Si anode, a sulfide Li₆PS₅Cl SSE, and a LiNi_0.8Co_0.1Mn_0.1O₂(NMC811) cathode. The full-cell achieved a capacity retention of 80% after 500 cycles and an average CE of >99.9% (Figure 3f). A distinctive feature of this battery is that the anode does not contain carbon. Conductive carbon and electrolytes were generally added to ensure good electron and ion transport paths, but they reduced the weight proportion of the active material and the effective electrode loading [31,40,41]. Eliminating carbon additives is also favorable in maintaining the stability of the SSE after forming a relatively stable two-dimensional interfacial SEI layer with a limited area. This is in contrast to the uncontrolled growth of the SEI of the Si anode during prolonged cycling [42].

Although both nm- and μm-Si have shown promise for applications in ASSBs, some problems remain to be addressed. nm-Si may exacerbate the interfacial side reactions with SSEs due to its higher specific surface area and could also have size-matching issues with micro-sized SSEs, which may result in a lower ionic conductivity of the electrode [33]. The electrodes consisting of μm-Si are subject to void formation and porosity change, which affect the cycling performance of ASSBs [42].

2.2. Si Anodes with Different Structures

2.2.1. Porous Si Anode

Porous materials have many advantages, such as high specific surface areas, plenty of active sites, and short ion diffusion distances, which favor good rate performances in LIBs [43]. The porous structure also improves the cycling performance of ASSBs because the internal connections of porous Si improve the structure integrity, and the pores can act as a buffer region accommodating the volume changes during cycling. Okuno et al. [44,45] investigated the effects of the structure and dispersion method of porous Si on the performances of ASSBs with a sulfide SSE. The nano-porous Si half-cell exhibits an excellent capacity retention of 89% after 50 cycles, with a remaining capacity approximately 11 times higher than that of the non-porous Si half-cell (Figure 4a) [46]. The improved cycling stability could be because the volume expansion of Si during charging can be accommodated by the nano-sized pores and the elastic deformation of the surrounding SSE. Furthermore, Okuno et al. [47] studied the effect of the dispersion methods of mechanical milling and hand milling. The highly dispersed nano-porous Si electrodes prepared by mechanical milling enabled a capacity retention of 80% up to 150 cycles (Figure 4b). It is believed that the stress generated by nano-porous Si could be relieved by the elastic deformation of the surrounding electrolyte so that the electrolyte can maintain good contact with nano-porous Si during delithiation (Figure 4c). In the case of a low degree of dispersion (prepared by hand milling), the stress generated by aggregated nano-porous Si may cause the electrolyte to deform plastically, and thus voids can be formed between the nano-porous Si and the electrolyte during delithiation, leading to a reduction in capacity. It should also be noted that the highly dispersed nano-porous Si electrodes (prepared by mechanical milling) exhibited a lower first-cycle charge–discharge capacity as well as a lower ICE than that prepared by hand milling. This could be related to the larger contact area between the nano-porous Si and the SSE and the more surface oxide layers of the nano-porous Si prepared by mechanical milling.

2.2.2. Columnar Si Anode

Columnar Si is also a promising architecture [50,51]. The low surface area-to-volume ratio of columnar Si favors a high ICE [52]. Cangaz et al. [48] prepared a columnar Si anode (col-Si) by the physical vapor deposition process (PVD) and coupled it with a Ni-rich layered oxide cathode to form ASSBs with a Li₆PS₅Cl SSE. Figure 4d shows the macroscopic structure of col-Si, which is compactly aligned with each other and vertically connected to the foil. The full-cell delivered a good capacity retention of 82% after 100 cycles. No short circuits were observed even at a relatively high charging rate (1 mA cm⁻²). The SSE contacts closely with columnar Si but does not penetrate the gaps. This condition facilitates a 2D lateral SEI that can be further stabilized by external pressure (Figure 4e). As shown in Figure 4f, due to the close contacts between Si pillars prohibiting SSE penetration, during lithiation/delithiation, the volume expansion of col-Si occurs mainly in the longitudinal direction and not inside the anode, which ensures the stability of the SEI during the cycling process. The volumetric effects can also be accommodated by the external pressure.

2.2.3. Film Si Anode

Thin-film Si anodes generally exhibited lower volume expansion than Si particles and also showed promise in ASSBs [53]. Thin-film Si anodes have been successively deposited on the surface of SSEs and show good compatibility [54,55,56]. Miyazaki et al. [57] reported an amorphous Si(a-Si) film anode for ASSBs with a specific capacity of 2400 mAh g⁻¹ after 100 cycles at 0.1 mA cm⁻². To alleviate accumulated stress and enhance cycling performance, Sakabe et al. [49] designed porous amorphous Si films (Figure 4g), which delivered a high capacity of 3128 mAh g⁻¹ and a very low capacity fading rate of 0.06% per cycle even at a high practical areal capacity of 2.3 mAh cm⁻² (Figure 4h). Despite its promise, film Si anodes also face limitations, and some important factors should be considered. For example, film thickness affects cycling performance [58], and the crystallinity and microstructure affect ion/electron transport properties.

3. Si-Based Composite Anodes

Compositing Si with other conductive, active, buffering, or coating materials to form Si-based composite anodes is a prevailing strategy to improve the cycling and rate performances of Si towards HED LIBs [59,60]. Herein, the electrochemical performances of Si-based anodes in ASSBs that are mentioned in this section are partly summarized in

Table 1. Electrochemical performances of Si-based anodes.

Si-Based Anode	Solid-State Electrolytes	Ionic Conductivity	Counter Electrode and Voltage Range	ICE	Current	Cycling Performance	Refs
Si-C	Li6PS5Cl	/	Li 0.01–2.00 V	73.2%	0.2 mA cm−2	53rd: >1000 mAh g−1Si-C	[61]
Si@SiO2@LPO@C	PEO@LATP	2.86 × 10−5 S cm−1 (50 °C)	Li 0.005–1.5 V	88.7%	0.5 A g−1 (50 °C)	200th: 1001.9 mAh g−1	[62]
Si@MgO@C	PEO@LATP@ NCF	1.68 × 10−4 S cm−1 (50 °C)	Li 0.005–1.5 V	81.4%	0.3 A g−1	100th: 1658.9 mAh g−1	[63]
Si/CNF@LPSCl	Li6PS5Cl	/	Li-In 0.05–1.5 V	/	0.5 C (25 °C)	Capacity retention: 83.4% (50 cycles)	[64]
Si-VACNF	Gel Polymer	~2.2 × 10−3 S cm−1	Li 0.05–1.5 V	76%	2.6 A g−1	100th: ~1050 mAh g−1	[65]
Si/CNTs/C	Li6PS5Cl	2.13 × 10−3 S cm−1 (30 °C)	Li	59.7%,	50 mA g−1	50th: 1226 mAh g−1	[66]
Si/Graphite	Li6PS5Cl	/	Li	/	1.77 mA cm−2 (50 °C)	Capacity retention: 88.8% (100 cycles)	[67]
PL-Si/Graphite	Li6PS5Cl	/	Li 0.01–2.0 V	~108.9%	0.5 C (60 °C)	Capacity retention: 79.5% (150 cycles)	[68]
LiSH46	Li6PS5Cl	/	NCM811 2.5–4.2 V	/	1 C (55 °C)	Capacity retention: 61.5% (5000 cycles)	[69]
Si/Li21Si5	Li6PS5Cl	/	LCO/Li3InCl6 2–4.2 V	97.8%	0.5 C (25 °C)	Capacity retention: 80% (175 cycles)	[70]
Si-PAN	77.5Li2S-22.5P2S5	/	Li-Ln 100 mV–1 V	~84%	0.1 C (60 °C)	200th: 1606 mAh g−1	[41]
Si-N-MXene	PEO@LATP	3.4 × 10−4 S cm−1 (50 °C)	Li 0.005–1.5 V	82.02%	0.4 A g−1 (50 °C)	90th: 881 mAh g−1	[71]
Si@MOF	PVDF/PEO/garnet	81 × 10−4 S cm−1 (25 °C)	Li 0.01–1.5 V	72.0	0.2 A g−1 (60 °C)	50th: 1442 mAh g−1	[72]

.

3.1. Si/C Composite Anodes

Carbon materials in Si/C composite anodes can serve several roles: (1) as coating materials restricting the volume change in Si and preventing surface side reactions; (2) as conductive substrates/networks improving the conductivity of the electrode; or (3) as active materials contributing capacity. Following these concepts, many Si/C composite anodes, such as simple Si/C composites [31], core–shell-structured Si/C [73], Si/graphene [74], Si/carbon nanofibers (CNFs) [64], and Si/carbon nanotubes (CNTs) [66], have been designed and studied in Si-based anodes in ASSBs.

Poetke et al. [61] prepared Si-C void composites using polyvinyl butyral (PVB) as a void template (Figure 5a,b). The void inside the composite material was designed to provide free space for volume expansion. And the carbon shell can prevent Si from having direct contact with the electrolyte. As an anode in ASSBs, the Si-C void composite exhibited a higher capacity retention and CE compared to that delivered in LIBs with a liquid electrolyte because of the inhibited side reactions and suppressed excessive growth of the SEI. Han et al. [62] designed a novel μm-Si/SiO₂/Li₃PO₄ (LPO)/C multilayer-coated Si composite anode (Figure 5c,d). The SiO₂/Li₃PO₄@C coating layer can provide fast ionic and electronic pathways and improve the integrity of Si particles. In addition, theoretical calculations suggested that the SiO₂ layer could reduce the energy barrier of Li⁺ transport. The μm-Si/SiO₂/LPO/C electrode showed a reversible specific capacity of 2482.1 mAh g⁻¹ with an ICE of 88.7%, and a capacity of 1001.9 mAh g⁻¹ was maintained after 200 cycles at 0.5 A g⁻¹. In sharp contrast, the Si@C anode without the SiO₂/Li₃PO₄ interlayer only delivered a capacity of 1130.8 mAh g⁻¹ at 0.2 A g⁻¹ and maintained a capacity of 202 mAh g⁻¹ after 100 cycles (Figure 5e). Recently, a Si@MgO@C monothetic anode electrode was developed by the same group [63]; see Figure 5f. The conductive carbon network prepared with a polyacrylonitrile (PAN) backbone facilitated the electron transport and the charge transfer kinetics at the Si/SSE interfaces. Matched with a poly(ethylene oxide(PEO)/Li_1.3Al_0.3Ti_1.7(PO₄)₃(LATP)/nano cellulose (NCF) SSE, the Si@MgO@C electrode exhibited a high capacity of 2576.1 mAh g⁻¹ with a CE of 81.4% during the first cycle and a remaining capacity of 792.1 mAh g⁻¹ after 200 cycles at a current density of 0.5 A g⁻¹ (Figure 5f,g).

Paik et al. [64] reported Si nanoparticles embedded in carbon nanofiber (CNF) sheathed with Li₆PS₅Cl (LPSCl) (Si/CNF@LPSCl, Figure 6a) as anode materials to achieve high energy density and stable cyclability for sulfide-based ASSBs. CNFs with good mechanical properties and high electronic conductivity [75,76] can build good electronic transport pathways and buffer the volumetric effect of Si particles. Furthermore, the conformal coating of LPSCl on the fiber enhances the ionic transport and stability at the active material/SSE interface. The Si/CNF@LPSCl electrode exhibits a capacity retention of 84.3% after 50 cycles at 0.5 C in a half-cell with a Li-In counter electrode (Figure 6b). Li et al. [65] reported a Si/CNF composite anode electrode with unique vertical 3D architecture, a nanostructured coaxial Si shell coated on vertically aligned carbon nanofibers (VACNFs) (Figure 6c), which act as a good electron intercalation medium. Matched with a good thermal and stable mechanical gel polymer electrolyte, the Si-VACNFs showed a high capacity of 3450 mAh g⁻¹ at 0.36 A g⁻¹ and 1732 mAh g⁻¹ at 3.8 A g⁻¹. Huang et al. [66] designed a Si/CNTs/C anode with a “reinforced concrete” structure (Figure 6d), in which CNT frameworks can prevent the disruption of electron/ion transport pathways and can mitigate the volume expansion of Si. The Si/CNTs/C@Li₆PS₅Cl electrode showed a stable capacity of 1226 mA h g⁻¹ after 50 cycles at 50 mA g⁻¹ in a half-cell. Kim’s group studied diffusion-dependent [77] electrodes (DDEs) consisting mainly of graphite and Si without SSEs to meet the high power and high energy density requirements of ASSBs [67,68]. Adding high-capacity nm-Si in graphite enabled much thinner coating layers and therefore can shorten the diffusion distance of Li ions in graphite DDEs and increase the energy density (Figure 6e). Graphite can accommodate the volume change in Si and provide a good electron transport path. The electrode with optimized composition demonstrated high areal and volumetric capacities of 2.76 mAh cm⁻² and 935 mAh cm⁻³ at a high current density of 1.77 mA cm⁻². Furthermore, they developed dry pre-lithiated DDEs (PL-DDEs) by introducing Li metal powder [68], which can supply Li ions to the active materials even in a dry state to enhance the CEs and compensate for the Li loss prolonging the cycling (Figure 6f). In contrast to other methods, this pre-lithiation method is simple and can avoid additional side reactions [78,79]. Compared to the bare DDE, the capacity retention of the ASSBs with the PL-DDE was significantly improved to 85.2% after 200 cycles.

3.2. Effects of Carbon Additives for Si Anode

Due to the low electronic and ionic conductivity of Si, carbon additives are usually added to the Si anode to ensure electron transport [80,81]. The type and content of carbon additives have a large impact on ASSBs. Okuno et al. [82] systematically investigated the effects of the content of acetylene black (AB) in the composite anode on battery performance. They found that a higher content of AB can increase the electronic conductivity and first charging capacity of the anode (Figure 7a) but does not necessarily lead to an increased discharge capacity and capacity retention (Figure 7b). Compared to granular AB, which is disadvantageous to form a continuous channel, one-dimensional materials like carbon nanotubes (CNTs) and two-dimensional graphene are superior choices for constructing a stable conducting network and are also beneficial for alleviating the volume effects of Si anodes. James et al. [32] compared the effects of multi-walled carbon nanotubes (MWCNTs) and AB on nm-Si anodes (Figure 7c). After 100 cycles, the capacity of the electrode using MWCNTs was ~100% higher than that using AB (Figure 7d). The reason was ascribed to that MWCNTs can provide a higher specific surface area for contact with Si and pose a higher flexibility to adapt to the volume change in Si during the cycling process.

3.3. Li-Si Alloy Anode

Compared to pure Si, Li-Si alloys, generally prepared by the controlled pre-lithiation of pure Si, pose higher ICE [83,84,85], improved electronic and ionic conductivity, decreased hardness, and a smaller volume expansion and therefore can give rise to improved cyclability and rate performances [86,87,88,89].

Ji et al. [90] reported a facile lubricant-assisted ball milling technique [91] using μm-Si powder and Li metal pieces as precursors, and hexane as a solvent, to synthesize all-electrochem-active pre-lithiated Si (Figure 8a). The anode was relatively soft, highly conductive, and had a high Li diffusivity and therefore cannot only enhance the energy density but also prevent electrolyte-related interfacial degradation [21]. As a result, the Li_xSi–Li_xSi symmetric cell displayed stable cycling over 320 h at 4 mAh cm⁻² (Figure 8b). Zhang et al. [92] tried in situ pre-lithiation by placing Si in direct contact with the ultra-thin Li foil based on a short circuiting mechanism under stack pressure. The reversible delithiation capacity and kinetics of the Si anode were improved, which enabled a full-cell to achieve a mass energy density of 402 Wh kg⁻¹ (at a rate of 0.1 C) and stable cycling over a wide temperature range (−30–50 °C). Lim’s group [93] reported a Li₃PS₄+Li–Si composite anode for ASSBs to mitigate the short circuit issues (Figure 8c) caused by Li dendrites and chemo-mechanical failure because of uneven lithiation/delithiation reactions [94]. Non-uniform alloying distribution and Li plating close to the SSE were effectively mitigated by the more pathways for Li ions in the anode (Figure 8d,e). The Li₃PS₄+Li–Si composite anode exhibited stable cycling performance for 355 cycles without any short circuits.

Recently, Wu et al. [69] developed a hard carbon-stabilized Li-Si alloy anode. Si can form a continuous amorphous film with minimized vacancy and close contact, due to the boundary fusion mechanism during the initial lithiation [95], but is prone to longitudinal cracks after many cycles [96] (Figure 8f). The Li-Si alloy was proven to be effective in suppressing the cracks (Figure 8g) but still subjected to short circuit issues (Figure 8h). By introducing hard carbon (HC) materials with high disorder, low potential plateaus, and sufficient defects [97], they further restricted the volume change and cracks in the electrode, as well as the Li dendrite growth. The Li-rich phases Li₁₅Si₄ and LiC₆ after lithiation formed a three-dimensional conducting network, which can enlarge the active area and improve the dynamics and structural stability of the electrode (Figure 8i). The hard carbon-stabilized Li-Si alloy anode enabled high-load and long-cycled sulfide-based ASSBs. Matching with the NMC811 cathode, it delivered a capacity retention of 80% after 1033 cycles and 61.5% after 5000 cycles (Figure 8j). It can also be cycled at high rates, delivering a capacity retention of 80% after 15,000 cycles at 30 C (Figure 8j). In another study, to avoid Li dendrite growth and short circuits, Huang et al. [70] prepared an all-electrochemically active Si/Li₂₁Si₅ composite anode, in which Li₂₁Si₅ was uniformly distributed and acted as a continuous soft buffering medium against the volume expansion of Si particles. At the same time, it could build excellent ionic/electronic conductive channels in the electrode and therefore can direct Li ions to Si particles, avoiding the formation of Li dendrites. The LCO/Li₃InCl₆|SSE|Si/Li₂₁Si₅ ASSBs using this composite anode achieved an ICE of 97.8% at 25 °C, a first area discharge capacity of 17.9 mAh cm⁻² at 55 °C, and an anode expansion rate of only 18.1%.

3.4. Other Si-Based Composite Anodes

Apart from carbon and alloys, many other materials with advantageous properties were also exploited to combine with Si. Among them, conductive polymers [98] and 2D materials with superior mechanical properties to inorganic materials were widely used to stabilize the structure and improve the performances of Si anodes.

Polyacrylonitrile (PAN) can be made electrically conductive without carbonization by controlling the heat treatment temperature in the appropriate range [99]. Lee et al. [41] utilized PAN as both a binder and conductive additive for Si-rich anodes (70 wt%) and investigated the electrochemical properties of the Si-PAN anode at different cutoff voltages. The Si-PAN half-cell delivered a high ICE of around 84% and an initial discharge capacity close to the theoretical capacity of Si (3579 mAh g⁻¹) in the cutoff voltage range of 5–200 mV, indicating that PAN can effectively mitigate the volume expansion of Si to maintain a stabilized electrode structure. Improved stable cycling can be achieved by increasing the cutoff voltage, which prevents Si from undergoing large structure changes, although at the cost of capacity loss [32,100]. After 200 cycles at 60 °C, a cell with a cutoff voltage of 100 mV delivered a capacity retention rate of 77.4%. By adjusting the heat treatment process, PAN can be transformed into amorphous carbon. Using PAN as a precursor, Han et al. [71] prepared a self-integrated Si-N-MXene anode with a Si-N chemical bond between Si and amorphous carbon and a N-MXene chemical bond between MXene and amorphous carbon which can improve the structural stability of the whole electrode. The Si-N-MXene/PEO@LATP/Li half-cell delivered an initial capacity of 2305 mAh g⁻¹ with an ICE of 82.02% at a current of 0.2 A g⁻¹. The capacity remained at 881 mAh g⁻¹ after 90 cycles at a current of 0.4 A g⁻¹. For the LiFePO₄ (LFP)/PEO@LATP/Si-N-MXene full-cell, the first-cycle capacity was 1659.2 mAh g⁻¹ at 0.32 A g⁻¹, and no capacity fading was observed after 60 cycles. Zhao et al. [72] developed high-capacity ASSBs using a metal–organic framework (MOF)-derived carbon-hosted Si (Si@MOF) anode and a fiber-supported PEO/garnet composite electrolyte. The Si nanoparticles were uniformly encapsulated in a micron-sized MOF carbon skeleton, which could buffer the volume change in Si while providing a conductive channel. Benefiting from this composite strategy, the Si@MOF anode exhibited excellent interfacial stability toward the composite polymer electrolyte and delivered an initial capacity of 1967 mAh g⁻¹ with an ICE of 72.0%. As a comparison, bare Si electrodes showed a low ICE of 50.8%, probably because of a severe interfacial reaction between the unprotected Si and the electrolyte. After 50 cycles, the Si@MOF maintained a reversible capacity of 1442 mAh g⁻¹, which was much higher than that of the bare Si electrode (636 mAh g⁻¹ after 50 cycles).

4. Key Factors Influencing Performance of Si in ASSBs

4.1. Interface Issues between Si and SSEs

The interface issues between the Si-based anode and SSEs are a focus of current research. Interfacial issues typically involve the chemical/electrochemical stabilities of SSEs, the surface substances of Si, and side reactions between them. In liquid batteries, the SEI forms on the anode surface due to solvent reduction and may break and re-generate as Si expands/shrinks in volume, which continuously depletes the Li⁺ and electrolytes and increases the cell impedance [101]. However, the growth mechanism of the SEI in ASSBs also deviates from that in liquid batteries. Unlike the liquid electrolyte, the SSEs will not completely wet the electrode material. By reducing the contact interface between Si and SSEs, the SEI can be limited to a 2D plane, which can be kept relatively stable and inhibit excessive side reactions and Li ion consumption. Meng et al. [21] investigated the interfacial issues of an anode consisting of 99.9 wt% μm-Si (Figure 9a) with Li₆PS₅Cl as the electrolyte. It was found that the columbic efficiency was low at the first cycle but stabilized afterward, indicating that the interface and the side reactions between Si and the electrolyte remained stable during cycling. This implied that the volumetric change in Si does not lead to the continuous formation and growth of the SEI film that is frequently observed in the liquid electrolyte (Figure 9b).

Although the solid–solid interfacial contact helps to mitigate the overgrowth of the SEI, the side reactions and product composition at the Si/SSE interfaces also play important roles in the performance of ASSBs. Among sulfide, oxide, and polymer SSEs [103], sulfide SSEs have attracted intensive attention due to their high ionic conductivity, good mechanical properties, and simple synthesis processes [104]. However, sulfide SSEs also suffer from low stability at low potentials when matched with Si anodes. The electrochemical stability of SSEs in the Si-based anode and the effect of carbon additives were investigated through in situ XANES by Zhu et al. (Figure 9c) [102]. It was found that Li₆PS₅Cl decomposition mainly occurs during the lithiation of Si in the first cycle and can be exaggerated by conductive carbon, forming products of Li₂S, Li₃P, and LiCl [105,106]. The peaks remained unchanged in the delithiation process, indicating that the decomposition is irreversible, and the products are relatively stable. Han et al. [15] performed first-principles calculations to understand the thermodynamic stability between various SSEs and Si. The calculations revealed that a variety of electrochemical reactions could occur between Si and sulfide electrolytes (75Li₂S-25P₂S₅(LPS) and 70(0.75Li₂S-0.25P₂S₅)-60LiI(LPSI)) with an onset potential of 1.5 V vs. Li⁺/Li (Figure 9d) and that they are mainly due to the interactions between Si and P, with the formation of SiP₂, SiP, and Li₅SiP₃ as possible products. In contrast, the stability between Si and 3LiBH₄-LiI(LBHI) is much better, and reactions occur only in a very narrow range of potentials (0.38–0.36 V vs. Li⁺/Li). The reaction mechanism was further analyzed by XPS tests (Figure 9e), which showed that sulfide SSEs can decompose into products such as Li₃P and can also react with Si to form products such as SiP, whereas no side reaction occurs between Si and LiBHI. As a result, a high-performance solid-state Si anode with 96.2% ICE and excellent electrochemical stability can be realized by using LBHI SSE.

4.2. Electrochemical–Mechanical Coupling Effects

A full understanding of electrochemical–mechanical coupling effects in Si-based ASSBs is key to elucidate the failure mechanism of Si anodes. The volume expansion of Si can lead to stress and strain evolution and variation in the electrode structure, which may decrease effective contacts, increase the impedance, and therefore cause the cycling decay of ASSBs.

Appropriate external pressure can suppress the volume expansion of Si, which is effective in maintaining the interfacial contact of Si/SSEs and facilitating the cycling of the battery. In 2012, Lee et al. [107] reported the effect of applied external stress on the electrochemical performance of nm-Si. The capacity of the Si half-cell nearly reaches the theoretical capacity when the applied pressure reaches 3 MPa. With an increased external pressure of 230 MPa (Figure 10a), the capacity of the battery is only half of that of the external pressure of 3 MPa when the applied pressure is 230 MPa. However, the increase in external pressure is also beneficial to the cycling stability performance of the battery. The capacity of the battery with an external pressure of 3 MPa decreases rapidly to 76.1% of its initial value after eight cycles. In contrast, the capacity of half-cells with an external pressure of 150 MPa and 230 MPa still retains 87.3% and 99% after 20 cycles, respectively (Figure 10b). Higher external pressure decreases the capacity because constraining the volume expansion of the Si material may result in the incomplete lithiation of Si. And lower external pressure affects the cycling performance because the electrode integrity degrades faster because of the unrestrained volume change in electrodes. Therefore, a trade-off between capacity and cycling performance exists when choosing the optimized magnitude of external pressure [24]. Yamamoto et al. [40] found that as the pressure increased from 50 MPa to 75 MPa, the initial discharge capacity of the Si|Li-In half-cell increased from 2849 mAh/g to 3412 mAh/g also with a higher capacity retention of 64% after 140 cycles. By comparing the structure change of the Si anode under different pressures, it was suggested that a higher pressure of 75 MPa can form large perpendicular cracks which relieve the stress and result in stable cycling instead of fine cracks.

McDowell et al. [108] investigated the internal stress evolution of the Si anode during cycling in an ASSB with an argyrodite-type SSE and NMC cathode (Figure 10c). The stress of the Si anode increases gradually during charging, reaches a maximum, and then decreases during discharging (Figure 10d). During the following cycles, the stress immediately after discharging does not return to the original value but increases somewhat, indicating that the electrode structure is irreversibly damaged. They also investigated the effect of Si particle size on stress. It was found that nanometer particles induced a smaller stress change compared to microparticles, as the volumetric effect of smaller particles could be better accommodated by the internal voids between them (Figure 10e). Based on the dP/dt curves (stress change/time), it can be found that the rate of stress change is the maximum at the end of charging and close to zero at the end of discharging. At the beginning of charging, the rate of stress change is small because the pores inside the electrodes buffer the stress change due to volume expansion. However, the void diminishes with the charging process, which leads to an increased rate of stress change (Figure 10f). Zhu et al. [102] investigated strain and stress change in three types of anodes, Si, Si+SSE, and Si+SSE+C (Figure 10j). During the cycling, the plastic deformation and stress of the Si anode are the largest, while those of the Si+SSE electrode are smaller, and those for Si+SSE+C are the smallest, suggesting that both the SSE and C can alleviate the stress generated in Si. The deformation of the Si anode does not return to the original volume ratio (zero) at the end of lithiation, indicating that the electrode structure is irreversibly damaged (Figure 10i). Recently, Tao et al. [109] investigated the performance decay of the μm-Si anode in sulfide-based ASSBs with different Si loading and different external pressures. As the Si loading increases, the Li-ion concentration gradient and the hydrostatic stress gradient inside the Si electrode also increase, causing uneven stress distribution and making the electrode susceptible to cracking, which is harmful to the cycling performance of the battery. Through the simulation, it was found that the applied pressure helps to reduce the unevenness of the stress inside the electrode, which reduces the possibility of electrode structure damage (Figure 10g). They found that the applied pressure has little effect on Li⁺ diffusion but is advantageous to reduce internal stress, the degree of electrode cracking, and the Li⁺ concentration gradient (Figure 10h).

5. Summary and Perspective

With the merits of a high specific capacity, high chemical stability, and low cost, Si anodes hold great promise to fulfill the promise of the intrinsic safety and high energy density of ASSBs. This review summarizes the advantages and drawbacks, as well as the current research progress of Si and Si-based composite anodes with various structures and compositions, discusses the key issues of interface instability between Si and SSEs, and probes the electrochemical–mechanical coupling effects of Si-based anodes. The main perspectives and suggestions are summarized as follows:

(1) Since the designs, formulas, manufacturing processes, and working conditions of ASSBs deviate, the basic physiochemical requirements and application standards of the Si anode remain not converged. Various Si anode structures with different electrode formulas and testing conditions have been investigated. Key influencing factors include sizes, dimensions, and architectures. In general, μm-Si is prone to pulverization, void formation, and porosity change in the electrodes, which affect its cycling performance. nm-Si is beneficial because of the better structure integrity; however, when applied in composite anodes with SSEs, it may exacerbate the interfacial side reactions with SSEs due to its higher specific surface area and could also have size-matching issues with micro-sized SSEs. Porous-structured Si is an attractive choice because the connections between domains can improve the whole integrity and the pores can act as a buffer region accommodating the volume changes in Si during cycling. Columnar Si arrays are also a promising architecture with volumetric changing in the lateral direction and a stable 2D lateral SEI that can be further stabilized by external pressure.

(2) Compositing Si with other materials proves an effective strategy to improve the cyclability and rate performances of Si. Carbon is a promising consistency in Si-based composite materials since it can serve as a coating material to restrict the volume change in Si, as a conductive network to improve the conductivity of the electrode, or as an active material contributing capacity. Li-Si alloys pose higher ICE, improved electronic and ionic conductivity, decreased hardness, and a smaller volume expansion and therefore represent an attractive solution. However, simple, low-cost, and safe pre-lithiation or alloying methods and processes are required for wide application. Apart from carbon and alloys, conductive polymers and 2D materials with superior mechanical properties to inorganic materials are widely used to stabilize the structure and improve the performances of the Si anodes.

(3) The interface issues between the Si-based anode and SSEs are a focus of current research. The growth mechanism and composition of the SEI in ASSBs deviate from that in liquid batteries. By reducing the contact interface between Si and SSEs, the SEI can be limited to a 2D plane, which can be kept relatively stable and inhibit excessive side reactions and successive Li ion consumption. The volume expansion of Si can lead to stress and strain evolution, which can cause cracks, voids, and the degradation of the electrode structure that is harmful to the cycling performance of Si anodes. Appropriate external pressure can suppress the volume expansion of Si, and maintain the stabilized interfacial contact of Si/SSEs, and thus is an important factor in assessing the performances of the Si anodes.

To conclude, motivated by the urgent pursuit of safe and HED ASSBs, Si-based anodes have attracted intensive attention and witnessed significant advances in their performances and understanding of failure mechanisms. Nonetheless, there remain many issues hindering their practical application. Some of these issues could be naturally solved by the progress on highly electrochemically stable and mechanically favorable SSEs. But others are inherent to Si itself and need to be conquered by collaboration between academia and the industry. We envision that the perspectives and suggestions in this review could shed some light on the design and application of high-performance Si-based anodes accelerating the commercialization of ASSBs.