Toward Scalable Liquid-Phase Synthesis of Sulfide Solid Electrolytes for All-Solid-State Batteries

Abstract

All-solid-state batteries (ASSBs) are promising to be next-generation battery that provides high energy density and intrinsic safety. Research in the field of ASSBs has so far focused on the development of highly conductive solid electrolytes (SEs). The commercialization of ASSBs requires well-established large-scale manufacturing for sulfide SEs with high ionic conductivity. However, the synthesis for sulfide SEs remains at the laboratory scale with limited scalability owing to their air sensitivity. The liquid-phase synthesis would be an economically viable manufacturing technology for sulfide SEs. Herein, we review a chemical perspective in liquid-phase synthesis that offers high scalability, low cost, and high reaction kinetics. This review provides a guideline for desirable solvent selection based on the solubility and polarity characterized by the donor number and dielectric permittivity of solvents. Additionally, we offer a deeper understanding of the recent works on scalable liquid-phase synthesis using solubilizers and reactant agents. We present an outlook on a universal liquid-phase synthesis of sulfide SEs toward the commercialization of sulfide-based ASSBs.

1. Introduction

Lithium-ion batteries (LIBs) are attracting attention as a power source for electric vehicles (EVs) [1]. EV sales are expected to grow tenfold by the end of this decade, even though they were just 2–3% of total passenger vehicle sales in 2020 [2]. However, traditional LIBs with flammable organic liquid electrolytes have safety issues, such as leakage and ignition. These issues remain critical obstacles that hinder the market adoption of LIBs in the EV field [3]. All-solid-state lithium-ion batteries (ASLBs) with non-flammable inorganic solid electrolytes (SEs) offer inherently higher safety [3,4]. Inorganic SEs are crucial for the development of ASLBs that meet the demanding requirements of the EV sector for safer and greater batteries. Extensive material exploration to date has led to the discovery of highly conducting crystalline Li₇P₃S₁₁ and Li₁₀GeP₂S₁₂, which exhibit an ionic conductivity in the order of 10⁻² S cm⁻¹ at room temperature, comparable with those of organic liquid electrolytes [5,6]. This discovery has largely advanced the application of sulfide-based ASLBs [7,8]. Recent research has focused on the material design of a family of argyrodite-type SEs with high ionic conductivity and moderate electrochemical stability. Sulfide SEs are generally synthesized by the solid-phase method using quartz ampoules or the mechanochemical method using ball-milling. These synthetic techniques are used on a laboratory scale for the discovery and design of novel materials. The synthesis is commonly performed in a glove box under an inert atmosphere because sulfide SEs and their raw materials react with moisture in the air to release toxic H₂S gas. Given the low scalability of both synthetic methods, the scientific community developing sulfide-based ASLBs is currently interested in manufacturing technologies involving a level of practical application [9].

The liquid-phase method would be an economically viable technology for ASLB manufacturing [10,11]. However, the liquid-phase synthesis of sulfide SEs is not described by only the simple chemical reactions of raw materials [12]. The liquid-phase synthesis involves complex interactions influenced by a solvent coordination bond, competition between the solvent and chemical species, and other factors [13]. Such interactions are significantly influenced by raw materials and their compositions, which determine the solubility and reactivity in the liquid-phase synthesis of sulfide SEs. The reaction mechanism, solubility, and reactivity of Li₂S-P₂S₅ compounds (common sulfide SEs) in solvents are becoming evident, but there is sparse understanding of these chemical perspectives in the other family of sulfide SEs, which consists of SiS₂ and GeS₂ in solvents. In addition, the reaction mechanism of sulfide SEs doped with lithium halides, which are often introduced to improve ionic conductivity and electrochemical stability, in solvents is unclear because of the coordination structure altered by the presence of the halide ions. The liquid-phase synthesis involves several difficulties: estimation of increasing increments for the nucleophilic and electrophilic contributions, respectively; the choice of optimal empirical parameters for the nucleophilic and the electrophilic functions; and the understanding of the reaction mechanism [13,14,15]. Therefore, the liquid-phase synthesis of sulfide SEs has been advanced by an empirical approach so far [15,16,17,18]. This review aims to discuss the chemical perspective in the liquid-phase synthesis for sulfide SEs and provide a design guideline for the liquid-phase synthesis that realizes the high scalability, low cost, and short reaction processing time.

2. The Liquid-Phase Synthesis Method

In the liquid-phase synthesis, sulfide SEs are generally synthesized through a chemical reaction in solvents, solvent removal, and the crystallization process. This synthesis method is classified into suspension and solution synthesis, depending on the state of the solution in the reaction process, as shown in Figure 1. The state during the reaction process is determined by the properties of the solvent.

2.1. Suspension Synthesis

In suspension synthesis, precursor suspensions are obtained through the chemical reactions of raw materials in a solvent with moderate solubility [10]. The synthesis of sulfide SEs via suspension is summarized in

Table 1. Synthesis of sulfide SEs via suspension.

Product	Treatment	Solvent	Reaction Time (h)	Ionic Conductivity (mS cm−1)	Reference
Li3PS4	Stirring	THF	12	0.074	[19]
Li3PS4	Stirring	ACN	48	0.12	[14]
Li3PS4	Shaking	EP	6	0.20	[33]
Li7P3S11	Stirring	DME	72	0.27	[20]
Li7P3S11	Stirring	ACN	24	1.5	[35]
Li7P3S11	Sonication	ACN	0.5	1.0	[32]
Li7P3S11	Microwave	anisole	0.5	0.13	[27]
Li7P2S8I	Stirring	ACN	36	0.63	[31]
Li7P2S8I	Shaking	EP	6	0.34	[23]
Li7P2S8I	Shaking	EP	1	1.0	[36]
Li6PS5Cl	Stirring	THF+EtOH	36	2.4	[37]
Li6PS5Cl	Stirring	ACN+PTH	24	2.8	[38]

. Li₂S-P₂S₅ systems are generally synthesized via the suspension because of their incomplete dissolution in most aprotic polar solvents. Liang et al. first synthesized nanoporous β-Li₃PS₄ via the suspension process using tetrahydrofuran (THF) [19]. A mixture of Li₂S and P₄S₁₀ (commonly referred to as P₂S₅ in the literature) is reacted overnight in THF to generate an Li₃PS₄·3THF complex. This compound releases THF molecules under a vacuum at 80 °C, followed by heating at 140 °C under a vacuum to produce the β-Li₃PS₄ with an ionic conductivity of 7.4 × 10⁻⁵ S cm⁻¹ at room temperature. Subsequent research found that 1,2-dimethoxyethane (DME) [20], acetonitrile (ACN) [15,16], ethyl acetate (EA) [21,22], ethyl propionate (EP) [23,24], pyridine [25,26], and anisole [27] are also used as an effective reaction solvent for the suspension synthesis owing to their moderate solubility and reactivity. Lithium sulfide and P₂S₅ reactants are individually insoluble in most aprotic polar solvents, but form highly soluble solvate lithium thiophosphates at a 1:1 molar ratio of Li₂S and P₂S₅ [28]. The intermediate reacts with Li₂S to generate Li₃PS₄ precursors. The reaction from insoluble Li₂S to insoluble Li₃PS₄ involves a long processing time (1–3 days) because of the sluggish reaction kinetics. The suspension synthesis of Li₇P₃S₁₁ and Li₇P₂S₈I involves the rate-limiting process of the Li₃PS₄ precursor formation [29,30,31]. This process involves the formation of heterogeneous precursors, which may increase the engineering cost of ASLB manufacturing, particularly for large-scale manufacturing. On the other hand, giving additional energy by mechanical [23], super sonication [32], and microwave [27] treatments allows us to boost the reaction kinetics and reduce the processing time. The liquid-phase synthesis assisted by the mechanical treatment provided the rapid synthesis of the Li₃PS₄ SEs with an ionic conductivity over 10⁻⁴ S cm⁻¹ at room temperature [33]. Amorphous Li₃PS₄ prepared by the mechanochemical reaction shows higher ionic conductivity than the crystalline β-phase [34]. Thus, in the liquid-phase synthesis, the solvent removal process should be conducted at a lower temperature than the crystallization temperature of β-Li₃PS₄ to achieve high ionic conductivity. The evaporation temperature of a coordinated solvent depends on the strength of the coordination bond between the solvent and intermediate molecular complex, which is determined by the polarity of the solvent. In the liquid-phase synthesis of Li₃PS₄ via acetate solvent systems, solvents with lower polarity resulted in a lower temperature of the solvent removal [22]. Li₃PS₄ synthesized using acetate solvents with lower polarity showed higher ionic conductivity because of the formation of low crystallinity Li₃PS₄, even after the solvent removal.

In the suspension synthesis of Li₆PS₅Cl, alcohol solvents are used as effective reaction mediators. For instance, ethanol (EtOH) solvent shows an extremely high solubility against Li₂S and P₂S₅, but causes a substitution reaction between the O atom in the solvent and the S atom in the P₂S₅ molecule. Despite this limitation, the Li₆PS₅Cl precursor was synthesized by the reaction of the Li₃PS₄ THF complex precursor, Li₂S, and LiCl in an alcohol solvent [37,39]. We believe that this is because the PS₄^3– anion unit is protected from EtOH by steric hindrance from THF coordinated to 3Li⁺-PS₄³⁻. The Li₆PS₅Cl SEs did not return to their original argyrodite structure after dissolving in EtOH and heating again [40]. This should be derived from the severe decomposition of PS₄³⁻ by EtOH. The decomposition of the Li₆PS₅Cl by EtOH can be minimized by decreasing the exposure time to EtOH [41,42]. Considering the side reactions caused by EtOH, a synthesis method using an ACN/1-propanethiol (PTH) solvent was developed [38]. The Li₆PS₅Cl SEs synthesized using the THF/EtOH solvent contained an Li₃PO₄ crystal phase as a secondary phase (see Figure 2a). Even Li₆PS₅Cl synthesized using THF/PTH formed oxides in the final product, which originated from the ring-opening reaction of THF solvent caused by the nucleophilic attack of the dissociated Cl ions and the –S⁻ anion of PTH. On the other hand, highly pure Li₆PS₅Cl SEs without Li₃PO₄ were successfully synthesized in an oxygen-free system using ACN/PTH solvent (Figure 2b). Additionally, Li₆PS₅Cl SEs with relatively high purity were synthesized using only the THF single solvent by employing a heat treatment to fully react the raw materials [43].

2.2. Solution Synthesis

The solution synthesis involves the formation of a homogeneous precursor solution in an organic solvent with high solubility. Hence, the solution synthesis method provides faster reaction kinetics in comparison with suspension synthesis.

Table 2. Synthesis of sulfide SEs via solution.

Product	Treatment	Solvent	Reaction Time (h)	Ionic Conductivity (mS cm−1)	Reference
Li3PS4	Stirring	EDA	3	2.3 × 10−3	[44]
Li7P3S11	Stirring	EA	2	0.12	[21]
Li6−xPS5−xCl1+x (0 ≤ x ≤ 0.5)	Stirring	EDA	―	2.9	[25]
Li10Ge2PS12	Stirring	EDA+ET	3	1.2	[49]

summarizes the synthesis of sulfide SEs via the solution. Recent works reported that ethylene diamine (EDA) and EA dissolve 75Li₂S·25P₂S₅ [44,45] and 70Li₂S·30P₂S₅ (mol%) [21], respectively. It should be noted that the solubility of the solvent may be influenced by the presence of slight moisture in the solvent. In addition, polar hydrophilic protonic solvents, such as dimethylformamide (DMF) and 1-methyl-2-pyrrolidone (NMP), exhibit high solubility against Li₂S and P₂S₅ [46]. However, their use is limited because of their high basicity, which leads to the decomposition of anion units in intermediate molecular complexes [17,46,47]. Thiol solvents were found to be effective solvents for the solution synthesis of sulfide SEs [48,49]. Thiol solvents exhibit desirable properties, such as high polarity, high solubility, moderately low boiling point, and no side reactions with the raw materials and intermediates. A mixed solvent containing thiol solvent and EDA is a useful reaction solvent for a broad range of sulfide SEs, including Li₆PS₅Cl and Li₁₀Ge₂PS₁₂ [49]. With the use of ethanedithiol (EDT) solvent, carbonized solvent molecules remained in the sulfide SEs after heat treatment. This should be ascribed to the strong coordination bonds between the solvent molecules and solutes. The presence of residual solvent molecules led to high electronic conductivity of the SEs, which negatively affected their performance as an electrolyte for ASLBs. Generally, there is a relationship between the strength of the coordination bonds with solvent molecules and the evaporation temperature of the coordinated solvent. The use of the highly polar solvent may increase the industrial cost due to the large energy consumption in the solvent removal process. Taking this into consideration, an ideal manufacturing process would involve the complete dissolution of the precursor and precipitation of the final product, thereby improving yield, reducing production costs, and minimizing waste and solvent removal.

3. Solvent Selection

The solvent selection in the liquid-phase synthesis is a critical consideration, taking into account factors such as solvent polarity, solubility of reactants, products, and intermediates, solvent removal processes, and industrial concerns [12]. This review focuses on the topics of polarity and solubility. The polarity of organic solvents varies the bond properties in the intermediate molecular complexes. The extent of polarization induced by a given bond relies on the strength of the coordination bonds with the nucleophile and/or electrophile. The stronger the coordination interaction, the greater the induced polarization. The reaction between Li₂S and P₂S₅ is initiated by the dissociation reaction of the P₂S₅ through nucleophilic attack by the electron donor [50]. Therefore, the introduction of empirical functional parameters related to nucleophilic property is important for the effective synthesis of sulfide SEs via the liquid phase. Donor number (DN) and dielectric permittivity are commonly used as empirical parameters to quantify polarity.

Table 3. Functional parameter and physical properties of various solvents [13,51].

Solvent	Donor Number [kcal·mol−1]	Dielectric Permittivity	Boiling Point [°C]
1,4-dioxan	15	2.2	101
Carbon sulfide	2	2.63	46
Anisole	9	4	154
Tetrahydropyran	22	5.5	88
Ethyl acetate	17	6	77
1,2-Dimethoxyethane	24	7.2	83
Tetrahydrofuran	20	7.6	66
Pyridine	33	12	115
Ethylene diamine	55	14	116
Ethanol	32	19	78
Dimethylformamide	27	36	153
Acetonitrile	14	38	82

summarizes the DN, dielectric permittivity, and boiling point of various solvents.

3.1. Donor Number

The DN is defined as the negative enthalpy for the reaction of complexation with SbCl₅ in a highly diluted solution of dichloroethane. The concept of DN was proposed to express quantitatively the Lewis basicity of solvents. Aprotic solvents, in general, exhibit poor accepter properties, and their DN values are related to the extent of charge transfer associated with ion solvation [13]. Hence, the DN of a solvent serves as an indicator of its reactivity in liquid-phase synthesis. Previous studies reported that organic solvents with DN higher than 14 kcal mol⁻¹ were unsuitable for solution processing of Li₃PS₄ SEs because of their decomposition caused by nucleophilic attacks [18]. While THF has 20 kcal mol⁻¹ of DN, it is often used for the solution processing of sulfide SEs [40,43]. However, the guideline for solvent selection is not contingent on DN indicators only. EDA solvent shows the highest DN among the solvents listed in Table 3. The EDA with high DN acts as a strong nucleophilic agent, breaking the P-S bond in P₂S₅ and facilitating the complete dissolution of P₂S₅ [44]. The 75Li₂S 25P₂S₅ solution in EDA contains various units such as PS₄³⁻, P₂S₆⁴⁻, P₂S₆²⁻, and PS₃ ⁻, which convert into β-Li₃PS₄ consisting exclusively of PS₄³⁻ units after heat-treatment at 200 °C under vacuum.

3.2. Dielectric Permittivity

Dielectric permittivity is an important indicator for evaluating the polarity of solvents and determining their reactivity in wet-chemical synthesis. It was found that the dielectronic permittivity of solvents used in the synthesis relates to the ionic conductivity of Li₇P₃S₁₁ prepared by the suspension synthesis [15]. Figure 3a illustrates the correlation between the ionic conductivity of Li₇P₃S₁₁ at room temperature and the dielectronic permittivity of the used organic solvents. Higher dielectronic permittivity of the solvents leads to increased ionic conductivity in the Li₇P₃S₁₁. There is no relationship between the DN of the solvent and the ionic conductivity of Li₇P₃S₁₁ SEs. Figure 3b displays differential thermal curves of Li₇P₃S₁₁ precursors prepared using ACN, THP, and Dox solvents. An endothermic peak was observed between 100 and 200 °C, which is attributed to the evaporation of the solvent coordinated with lithium thiophosphates. Differential thermal analysis showed that the strength of the chemical interaction between the Li₂S-P₂S₅ system and solvent molecules increased with an increasing dielectric permittivity of the coordinated solvent. Therefore, the correlation between the conductivity and dielectric permittivity can be described using the argument that the use of an organic solvent with high dielectronic permittivity increases the reactivity of the precursor, resulting in the formation of the highly pure Li₇P₃S₁₁. Besides dielectronic permittivity, other selection criteria for the suspension synthesis of Li₂S-P₂S₅ systems include the absence of side reactions with raw materials, intermediates, and products, as well as a reasonably low boiling point. Given these requirements, ACN solvent is the optimal choice for the suspension synthesis of Li₇P₃S₁₁ at present.

3.3. Solubility

The formation of soluble precursors is preferred for the reaction to proceed quickly. EA solvent completely dissolves the intermediate resulting from 70Li₂S 30P₂S₅, enabling the rapid synthesis of Li₇P₃S₁₁ SEs [21]. Individual Li₂S and P₂S₅ are insoluble in EA solvent during solution synthesis. This indicates that a rapid synthesis for sulfide SEs does not necessarily require the solubilization of Li₂S and P₂S₅ raw material. In other words, the overall reaction rate in the synthesis process depends on the solubility of solvents for intermediate complexes, as follows: ACN and EP dissolve 1Li₂S 1P₂S₅ (molar ratio) [24,28]; THF dissolves 1.5Li₂S 1P₂S₅ [52]; and a mixed solvent of ACN and THF at a volume ratio of 1:1 dissolves 2Li₂S 1P₂S₅ [53]. Of course, the solubility of the solvent depends on a given compound. The solubility of Li₂S-GeS₂ and Li₂S-SnS₂ in organic solvents remains poorly understood. The strong covalent bonding between the metal and chalcogen atoms typically brings a challenge for the direct dissolution of these compounds. In 2009, it was reported that a bulk chalcogenide semiconductor and chalcogen in hydrazine solvent form a highly soluble hydrazinium-based salt of the chalcogenide semiconductor through a redox reaction [54]. Hydrazine is an effective solvent for the solution processing of metal chalcogenide, but its highly toxic, explosive, and carcinogenic properties make it less attractive for scale-up. A binary thiol/amine mixture was developed as an ideal alkahest solvent system for the solution processing of chalcogenide semiconductors [55]. A mixed solvent of EDA and EDT fully dissolves SnS₂ and GeS₂ [49,56]. EDA solvents do not cause a nucleophilic attack to the germanium moiety on GeS₂, even though EDA solvent with high DN shows a strong nucleophilic attacker capable of breaking P-S bonds. In contrast, the thiolate anion in thiol solvents dissociates Ge-S bonds in the GeS₂ through the nucleophilic attack. The high solubility of the mixed solvent of EDA and EDT for sulfides was first demonstrated in studies on the dissolution of bulk semiconductors, such as As₂Ch₃, Sb₂Ch₃, and Bi₂Ch₃ (Ch = S, Se, Te) which are insoluble in common organic solvents [57]. As mentioned above, depending on the solvent combination, mixed solvents may show higher solubility than a single solvent. Recent work reported that the mixtures of Li₂S and SiS₂ with a molar ratio of 2:1 are soluble in ACN solvent [58]. However, no comprehensive guidelines for gaining high solubility in starting materials for sulfide SEs have been proposed.

4. Strategies for Rapid Synthesis

4.1. Solubilizer

Solution synthesis is a desirable manufacturing technology for sulfide SEs. However, there are restrictions on the selection of solvents that dissolve raw materials and intermediates, e.g., Li₂S and Li₃PS₄, and no leading guidelines for solvent selection have been established. The introduction of a solubilizer has been examined as one of the strategies to dissolve Li₂S and Li₃PS₄. Excess elemental sulfur reacts with these insoluble materials to form intermediate molecular complexes that are soluble in specific solvents, such as THF and diethylene glycol diethyl ether (DEGDME) [59,60]. Lim et al. first proposed solution processing using excess elemental sulfur as the manufacturing technology for sulfide- based ASLBs [59]. In the sulfide SE solution processing, Li₂S and S reacted in DEGDME solvent to form fully soluble lithium polysulfides. Subsequently, P₂S₅ was added to the obtained solution to generate the β-Li₃PS₄ precursor. The β-Li₃PS₄ after drying treatment exhibited an ionic conductivity of 2.8 × 10⁻⁵ S cm⁻¹ at room temperature. The low conductivity might be attributed to residual sulfur due to an insufficient drying process at 140 °C. The synthesis method using excess sulfur requires a heat-treatment process over 250 °C for the removal of excess sulfur, and is therefore unsuitable for synthesizing SEs that exhibit high ionic conductivity in the low-temperature phase, for example, amorphous Li₃PS₄ and Li₇P₂S₈I. Subsequent work successfully synthesized Li₆PS₅Cl SEs via soluble lithium thiophosphates with excess sulfur, which showed a high ionic conductivity of 1.8 mS cm⁻¹ at room temperature [61]. However, the reported solution processing via polysulfide involved a long reaction time of 24 h or more. To achieve rapid solution synthesis of Li₇P₃S₁₁, a synthesis method based on polysulfide chemistry was introduced (Figure 4a) [53]. Conventional suspension synthesis of highly conductive Li₇P₃S₁₁ requires a long reaction process to ensure sufficient reaction completion (Figure 4b). As shown in Figure 4c, the 7Li₂S 3P₂S₅ system in ACN solvent formed a white suspension and a yellowish solution after stirring for 3 days; they correspond to the Li₃PS₄ complex and soluble 1Li₂S 1P₂S₅, respectively. According to the reaction mechanism proposed in the previous reports, Li₇P₃S₁₁ SEs are formed through a solid-state reaction of the Li₃PS₄ complex and Li₂S P₂S₅ (a molar ratio of 1:1) intermediate during heat treatment above 200 °C [29]. On the other hand, 7Li₂S 3P₂S₅ 5S in a mixed solvent of ACN, THF, and a trace amount of EtOH formed a black colored solution without additional energy in an extremely short time of 2 min (Figure 4d). The formation of the black solution suggests the presence of polysulfide ions. In the UV-Vis absorption spectra of 7Li₂S 3P₂S₅ 5S in the mixed solvent, S₈, S₆²⁻, S₄²⁻, and S₃^·− were detected at 265, 350, 410, and 610 nm, respectively (see Figure 4e). Research in the field of lithium sulfur batteries reported that lithium polysulfides with different chain lengths are stabilized in specific organic solvents through disproportionate and dissociation reactions [62,63]. Tri-sulfur radical anions with high reactivity are the dominant chemical species in the complex system of equilibria. The reaction kinetics in the synthesis method with excess sulfur is significantly influenced by the chemical stability of the S₃^·− radical anions. Intrinsically unstable S₃^·− radical anions are stabilized in highly polar organic solvents because of the competition for Li ions between strong coordinated bonds of highly polar solvents and electrostatic interactions of polysulfide anions. The rapid solution synthesis shown in Figure 4a can be achieved by the existence of highly polar EtOH, which enhances the chemical stability of the S₃^·− radical anion in the precursor solution.

Figure 5 outlines the proposed reaction of lithium phosphates with excess elemental sulfur in the precursor solution [53]. First, lithium ions form strong coordination bonds with highly polar EtOH and ACN solvents (step 1). Next, the polysulfides generated by the reaction of Li₂S and excess sulfur are shielded from Li ions (Step 2). Tri-sulfur radical anions are stabilized owing to the competition between the polysulfide dianions and salt anions for the Li cation (Step 3). Then, the S₃^·− radical anions with unpaired electrons attack the phosphorus moiety on P₄S₁₀ (step 4). The P₄S₁₀ with an adamantane-like cage structure dissociates to reactive P₂S₅ (step 5). The reactive P₂S₅ reacts with lithium polysulfides to form a soluble lithium thiophosphate. The added excess elemental sulfur is removed without deteriorating influence on the formation of the final products during the heat treatment for crystallization. Recent works demonstrated that Li₆PS₅Cl and Li₁₀GeP₂S₁₂ are rapidly synthesized via the precursor solution with excess elemental sulfur in the ACN/THF/EtOH solvent [64,65]. A mixture of Li₂S, P₂S₅, LiX (X = Cl, Br, and I), and S at a molar ratio of 5:1:2:15, which represents the raw materials of Li₆PS₅X SEs, formed a black solution in the ACN/THF/EtOH solvent, as shown in Figure 6a. This black coloration arises from the formation of polysulfide ions (see Figure 6b–d). The existence of a trace EtOH enhanced the solubility of the precursor solution and the chemical stability of polysulfides, which is explained by the coordination chemistry shown in Figure 5.

Interestingly, 5Li₂S P₂S₅ 2LiCl 15S precursor in ACN/THF solvent formed a reddish-brown solution, indicating lower chemical stability of the polysulfides than other precursor solutions. A Raman study showed that this observation may be caused by the strong competitive interaction between Cl ions and ACN molecules in the precursor solution [64]. In this solution synthesis method using excess elemental sulfur as a solubilizer, the coordination environment of solvent molecules, Li-ions, and polysulfides determines the chemical stability and reactivity of each chemical species. The coordination chemistry is also of interest in the research field of lithium-sulfur batteries [63,66,67]. Thus, this method can promise further advances by integrating the insight gained from coordination chemistry studies in the field of lithium-sulfur batteries.

4.2. Reactant Agent

As mentioned above, the wet-chemical synthesis of lithium thiophosphates initiates the dissociation reaction of P₄S₁₀. The dissociation reaction is driven by nucleophilic attack from lone-pair electrons of Li₂S and/or donor solvents [26,29]. The introduction of reactant agents for activating P₄S₁₀ is a useful strategy for the rapid synthesis of sulfide SEs. The addition of lithium thioethoxide and lithium halide facilitates the formation of soluble chemical species [64,68,69]. Halide ions are nucleophilic, and quickly dissociate P₂S₅ in organic solvents, as shown in Figure 6e [64]. Lithium ions are strongly solvated by donor solvents owing to their strong Lewis acid. Therefore, the bond of Li ions with counter ions is strongly polarized and more easily dissociated. In the coordination environment, where the cation is highly stabilized, the anion species would be available as reactive chemical species in the system. Taking this into consideration, the reaction mechanism in the wet-chemical synthesis of Li₇P₂S₈I SEs may intrinsically differ from that of Li₃PS₄ [23,24].

5. Low-Cost Synthesis

The price of ASLBs per energy density is one order of magnitude higher than that of conventional LIBs, and is dominated by the cost of raw materials for SEs [61]. Among these raw materials, Li₂S, an essential raw material, is one to two orders of magnitude more cost-expensive per gram than P₂S₅ [12]. Hence, it is necessary to develop a synthetic method using inexpensive starting materials instead of Li₂S. A liquid-phase synthesis of Li-ion SEs using inexpensive Na₂S raw material was recently developed: the synthesis of Li₄SnS₄ from Na₄SnS₄ using the ion-exchange method [70]; and the synthesis of Li₆PS₅Cl using Na₂S as a sulfurizing agent [71,72,73].

5.1. Ion-Exchange Method

Li₄SnS₄ was synthesized from Na₄SnS₄ using the ion-exchange method via water [70]. Firstly, an Na₄SnS₄ solution was prepared through the reaction between Na₂S and SnS₂ in water. The obtained Na₄SnS₄ solution flowed through a column packed with Li-type ion exchange resin to exchange from Na⁺ to Li⁺ ions in Na₄SnS₄. The resultant Li₄SnS₄ aqueous solution was dried and heated to obtain Li₄SnS₄ SEs (see Figure 7). The Li₄SnS₄ obtained from this method showed an ionic conductivity of approximately 1 × 10⁻⁴ S cm⁻¹ at room temperature, which is the same value as Li₄SnS₄ prepared by the solid-phase method and mechanochemical reaction. Unlike Li₂S, Na₂S can be handled in air, and it is easy to process the reaction in an aqueous solution. This method using inexpensive Na₂S and H₂O as reaction solvents resolves environmental loads owing to the use of organic solvents and economic concerns.

5.2. Sulfurizing Agent

Lithium chloride reacts with Na₂S in EtOH solvent to generate precipitated Na₂S and Li₂S/LiCl EtOH solution [71,72,73,74]. The Li₂S/LiCl mixture powder, separated by the distillation of the product, and P₂S₅ are added into THF solvent to form Li₃PS₄ intermediate molecular complexes. The Li₃PS₄, Li₂S, and LiCl mixture powder, prepared by the drying treatment of the precursor suspension, dissolves in EtOH solvent, followed by drying and heat treatment to obtain Li₆PS₅Cl SEs.

$${5\mathrm{N}\mathrm{a}}_2\mathrm{S}\left(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n}\right)+12\mathrm{L}\mathrm{i}\mathrm{C}\mathrm{l}\left(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n}\right)\underset{\mathrm{E}\mathrm{t}\mathrm{O}\mathrm{H}}{\to }{5\mathrm{L}\mathrm{i}}_2\mathrm{S}\left(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n}\right)+2\mathrm{L}\mathrm{i}\mathrm{C}\mathrm{l}\left(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n}\right)+10\mathrm{N}\mathrm{a}\mathrm{C}\mathrm{l}\left(\mathrm{s}\right)$$

(1)

$$5{\mathrm{L}\mathrm{i}}_2\mathrm{S}\left(\mathrm{s}\right)+2\mathrm{L}\mathrm{i}\mathrm{C}\mathrm{l}\left(\mathrm{s}\right)+{\mathrm{P}}_2{\mathrm{S}}_5\left(\mathrm{s}\right)\underset{\mathrm{T}\mathrm{H}\mathrm{F}}{\to }{2\mathrm{L}\mathrm{i}}_3\mathrm{P}{\mathrm{S}}_4\left(\mathrm{s}\right)+2{\mathrm{L}\mathrm{i}}_2\mathrm{S}\left(\mathrm{s}\right)+2\mathrm{L}\mathrm{i}\mathrm{C}\mathrm{l}\left(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n}\right)$$

(2)

$${\mathrm{L}\mathrm{i}}_3\mathrm{P}{\mathrm{S}}_4\left(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n}\right)+{\mathrm{L}\mathrm{i}}_2\mathrm{S}\left(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n}\right)+\mathrm{L}\mathrm{i}\mathrm{C}\mathrm{l}\left(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n}\right)\underset{\mathrm{E}\mathrm{t}\mathrm{O}\mathrm{H}}{\to }{\mathrm{L}\mathrm{i}}_6\mathrm{P}{\mathrm{S}}_5\mathrm{C}\mathrm{l}\left(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n}\right)$$

(3)

$${\mathrm{L}\mathrm{i}}_6\mathrm{P}{\mathrm{S}}_5\mathrm{C}\mathrm{l}(\mathrm{s}\mathrm{o}\mathrm{l}\mathrm{n})\stackrel{\mathsf{\Delta }}{\to }{\mathrm{L}\mathrm{i}}_6\mathrm{P}{\mathrm{S}}_5\mathrm{C}\mathrm{l}(\mathrm{s})$$

(4)

This method allows us to synthesize Li₂S from Na₂S and LiCl on the basis of the difference in the solubility of EtOH for Li₂S and NaCl. A further examination of the sulfurizing agent such as Na₂S can contribute to the development of the liquid-phase synthesis for sulfide SEs from inexpensive Li salts other than Li₂S.

6. Environmental Aspects

The industrial adaption of the liquid-phase synthesis for sulfide SEs to the electric vehicle fields requires consideration of environmental aspects, such as the toxicity of sulfide SEs and organic solvents, yield rate, energy-saving route, recycling of sulfide SEs. Sulfide SEs consisting of P or Si element release toxic gas, that is, H₂S gas, in contact with atmospheric moisture [75,76,77]. The air-stability of these materials is based on the classical hard-soft acid-base theory (HSAB) [76,77,78,79,80,81]. According to HSAB theory, the sulfur anion is a softer base than the oxygen anion and thus metal cations classified as softer acids form stronger bonds with sulfur than oxygen. Using softer acids, such as Sn⁴⁺ and Sb⁵⁺, as central cations has been an effective strategy to improve the air stability of sulfide SEs. Importantly, Li₄SnS₄ can be processed in a solution using water, which is the most environmentally friendly solvent [78]. This stability is desirable for recycling technology of sulfide SEs through energy-saving routes. Indeed, the Li₄SnS₄ subjected to exposure to water, followed by vacuum drying and heat treatment, showed almost the same ionic conductivity as the as-synthesized Li₄SnS₄ [78]. However, a family of Li₄SnS₄ SEs showed an ionic conductivity in the order of 10⁻⁴ S cm⁻¹ at room temperature, which is insufficient for practical application in ASLSBs [70,78,79,80]. On the other hand, the solution processing of a family of Li₂S-P₂S₅ with high ionic conductivity in water is unable to avoid detrimental effects on the electrochemical and structural properties of the Li₂S-P₂S₅ SEs [75]. The use of green solvent EtOH in solution processing may potentially contribute to the environmentally friendly synthesis of sulfide SEs [39,42,71,72]. In some cases, the introduction of protective agents may be necessary. The influence of green solvents on the nature of sulfide SEs still requires further exploration [82,83]. Such a chemical perspective would offer significant insights into SE recycling.

7. Summary

In this review, we have highlighted a chemical perspective on the liquid-phase synthesis of sulfide SEs for ASLBs. This article has discussed the role of an additive (e.g., reactant, solubilizer, and sulfurizing agent) and organic solvent for realizing rapid and low-cost synthesis of sulfide SEs.

The liquid-phase synthesis method of sulfide solid electrolytes for all-solid-state batteries offers many advantages: low cost, high scalability, and short processing time. The liquid-phase synthesis method is classified into the suspension method and solution method. The chemical reaction in these methods is determined by the nature of the solvent. We summarized a guideline for solvent selection based on the donor number, dielectric permittivity, and solubility. The suspension method leads to low energy consumption during solvent removal; however, it involves a long processing time owing to the poor reaction kinetics in the precursor suspension. In contrast, the solution method provides higher scalability and a shorter processing time than the suspension method. Studies on the solvent effect induced by various solvent combinations have the potential to greatly advance large-scale solution synthesis technologies. Moreover, the addition of the solubilizer and reactant is an effective strategy for achieving enhanced reaction kinetics in wet-chemical synthesis. The addition of excess elemental sulfur plays a role as the solubilizer and reactant, offering the rapid solution synthesis of a variety of SEs. Given that the cost of ASLBs is still higher than conventional LIBs, the development of a synthesis method using inexpensive raw materials instead of Li₂S reduces barriers to the commercialization of ASLBs. Additionally, minimizing solvent removal in the synthesis process will increase the feasibility of its application into all-solid-state battery manufacturing. The chemical perspective in the liquid-phase synthesis gives significant insight into recycling SEs. The recent progress in the community of ASLBs has focused on their demonstration in lab-scale cells. There is a demand for the development of a large-scale manufacturing technology that takes into account environmental load owing to the use of organic solvent and cost.