Garnet-type solid electrolyte: Advances of ionic transport performance and its application in all-solid-state batteries

All-solid-state lithium batteries (ASSLBs), which use solid electrolytes instead of liquid ones, have become a hot research topic due to their high energy and power density, ability to solve battery safety issues, and capabilities to fulfill the increasing demand for energy storage in electric vehicles and smart grid applications. Garnet-type solid electrolytes have attracted considerable interest as they meet all the properties of an ideal solid electrolyte for ASSLBs. The garnet-type Li7La3Zr2O12 (LLZO) has excellent environmental stability; experiments and computational analyses showed that this solid electrolyte has a high lithium (Li) ionic conductivity (10−4–10−3 S·cm−1), an electrochemical window as wide as 6 V, stability against Li metal anode, and compatibility with most of the cathode materials. In this review, we present the fundamentals of garnet-type solid electrolytes, preparation methods, air stability, some strategies for improving the conductivity based on experimental and computational results, interfacial issues, and finally applications and challenges for future developments of LLZO solid electrolytes for ASSLBs.


Introduction 
Lithium-ion batteries (LiBs) have been broadly used in portable electronics and mobile communication because of their high energy density, high operating voltage, and long cycle life. However, current commercially available LiBs generally used liquid electrolytes. energy storage [3,4].
The development of electrochemical energy storage devices for these applications has become subject of extensive research. Studies have been directed to understand the performance, safety, energy density, and cost requirements for the batteries of the future: all-solid-state lithium batteries (ASSLBs). Compared with LiBs using liquid electrolytes, ASSLBs (Fig. 1) are simple in structure and small in size, and may solve the safety concerns by replacing the liquid with solid electrolytes which results in a longer life cycle, simplify the manufacturing process, and increase energy density [5]. The main working principle of ASSLBs is the solid electrolyte (SE), which functions not only as an ionic conductor but also a separator, allowing the transfer of ions from the anode (negative) to the cathode (positive) during the charging and discharging processes, and the electrons generated by the reaction are used to promote a load in the external circuit [6]. Therefore, SEs are promising candidates to improve the overall performance of batteries. Battery performance depends on the materials used, so the development of new SEs and electrode materials are important in battery technology development. Several critical issues need to be considered for real-world applications, including the improvement of low ionic conductivities of SEs at low temperatures, large interfacial resistance at the electrode-electrolyte solid interface, poor electrochemical compatibility with electrodes such as Li-metal anodes and high-voltage cathode materials, enhanced physical stability, and better understanding of the interfacial process after charge/discharge of the battery [7]. Even though there is a fast development in ASSLBs, the safety and manufacturing process are another important challenges to fulfill the requirements of safe solid storage systems [8].
There are two general classes of materials used for SEs in ASSLBs: inorganic and polymer electrolytes. The main difference between them is the mechanical properties that influence the battery design. Although polymers are easier to process and reduce the fabrication costs, they possess a low Li-ion conductivity for battery operation at room temperature (RT). However, polymer electrolytes can be combined with Li anodes and safely cycled at elevated temperatures [9]. Therefore, challenges remain to develop stable polymer electrolytes for applications with Li metal anode and cathode materials at acceptable C-rates.
The key functional features of solid electrolytes should include: high electrochemical decomposition voltage of 5 V vs. Li/Li + ; chemical stability against reaction with the electrodes to prevent the formation of any undesired products at the electrode-electrolyte interfaces; negligible electrode-electrolyte interface charge transfer resistance; low cost; environmental friendliness; and most importantly a high Li-ion conductivity of about 10 -4 S·cm -1 [10,11]. Inorganic solid electrolytes are more suitable for high temperatures or aggressive environments, and a wide range of inorganic electrolytes have been developed, due to their amorphous and crystalline structures [12]. These two structures enable ion transport, as they provide enough vacancies and coordination defects that enable the flux of a number of ions throughout the framework [13][14][15]. Crystalline materials offer the highest Li-ion conductivities in SEs. The main inorganic SEs being explored are NASICON-type, perovskite-type, LISICONtype, garnet-type, and sulfide-type materials. Research has shown that Li 1+x Al x Ti 2-x (PO 4 ) (LATP) and Li 1+x Al x Ge 2-x (PO 4 ) 3 (LAGP) exhibited high ionic conductivity (up to 10 -3 S·cm -1 ), but they are unstable against Li metal anodes due to facile Ti 4+ reduction [16,17]. LLTO presents a similar problem at the interface with metallic Li, but the synthesis methods also result in high Li 2 O losses [18]. L 7 P 3 S 11 , a sulfidetype material presents the highest ionic conductivity (10 -2 S·cm -1 ), though it is sensitive to moisture and has poor compatibility with cathode materials [19].
Nevertheless, the major drawback of all of them is their mechanical properties. Table 1 summarizes the main characteristics of inorganic SEs. Thus, it is a major challenge to determine which electrolyte is more suitable for solid battery development and application, material synthesis, and device integration.
Garnet-type Li 7 La 3 Zr 2 O 12 (LLZO) electrolytes stand out as the most promising SEs. LLZO presents high ionic conductivity at RT (10 -4 -10 -3 S·cm -1 ), wide electrochemical window range (0-5 V), and good stability against Li metal anode, all this according to the first-principals calculation and experimental results. LLZO is also simple environmental caring [20][21][22][23]. Figure 2 shows a brief chronology of the development of garnet-type solid electrolytes among ASSLBs. In 2003, Thangadurai and coworkers [24] reported Li 5  (M = Nb, Ta, Bi, Sb) with a bulk ion conductivity of 10 -6 S·cm -1 at RT. After this study, several Li garnettypes have attracted attention due to their desired physical and chemical properties for application as solid electrolytes in ASSLBs. The most notable composition of garnet-type is Li 7 La 3 Zr 2 O 12 (LLZO), which has two polymorphs: a cubic phase (c-LLZO) and a tetragonal phase (t-LLZO) [24][25][26]. The c-LLZO phase presents the highest values of ionic conductivity. Substitution is the primary method to improve the conductivity; commonly studied garnets typically contain five to seven Li atoms per formula unit, which allowed them to induce changes in the Li-ion distribution [27,28]. Further doping with tantalum (Ta) and gallium (Ga) elements resulted in an improvement of ionic conductivity, to 10 -3 S·cm -1 [29,30]. Experimental and computational simulation results suggest that the coordinated migration of Li-ions by the side of Li1 and Li2 sites, contributes to auto-diffusion in the cubic phase in LLZO; ab initio molecular dynamics method (AIMD) confirms that the Li transport mechanism is related to Li-ion concentration [31,32].
In 2010 early studies of the compatibility of LLZO SE for ASSLBs were focused on the usage of Li metal anode [35]. The composite solid electrolyte (CSE) in 2015, utilizing mainly polyethylene oxide (PEO) and some other polymer materials such as polyacrylonitrile (PAN), poly(vinylidene fluoride) (PVDF), and poly(methyl methacrylate) (PMMA), has been investigated [36]. The research showed that the morphology of LLZO has a significant influence on the overall performance of the SE, and the chemical and mechanical interactions between ceramic particles and polymer matrix enable fast Li + conduction and wide electrochemical window in the CSE [31].
Therefore, efforts have been directed to understand diverse methods of synthesis of garnet-type materials and development of diverse morphologies and structures such as bulk ceramics, thin films, and nanostructured LLZO [29]. However, challenges in producing effective ion transmission networks, enhance the need of engineering stable electrolyte-electrode interfaces with cathode and Li metal anode materials [35][36][37][38][39][40]. Current research is oriented towards the construction of high energy density Li batteries all over the world. The goal of the "future batteries" is to produce a battery pack of 500-800 Wh/kg. It will only be achieved by focusing on developing methods to increase our understanding of the multiple non-equilibrium process in batteries, exploring the use of Li batteries in large energy storage systems, with higher performance, low cost for applications, and low maintenance for grid storage.
We will review now the fundamentals of garnet-type electrolyte LLZO, and discuss the developments and strategies used by researchers to improve the performance of LLZO as an SE. We will describe the Li-ion conductivity, chemical composition, stability in the air, and interfacial properties between LLZO and electrodes, as well as the electrochemical stability, with different types of anode and cathode materials based on experimental and computational results. Then, we comment on the mechanisms by which the LLZObased and LLZOpolymer solid electrolytes operate in a full battery with different LLZO/electrode interfaces. Finally, we provide information about further application of solid electrolytes in ASSLBs, challenges, and perspectives.

Crystal structure and phase transition of LLZO
LLZO is represented by the general chemical formula A 3 B 2 (CO 4 ) 3 , where the cations A, B, and C occupy crystallographic sites with 8-fold, 4-fold, and 6-fold oxygen coordination sites respectively, which usually crystallize in cubic structure with the space group 3 Ia d  [41]. The LLZO has two polymorphs with Li + ionic conductivities that differ by 2-3 orders of magnitude. One is the t-LLZO, space group 1 4 / I acd , with a garnet-type framework with two types of dodecahedral LaO 8 polyhedral (8b and 16e) and ZrO 6 octahedral (16c) (Fig. 3(b)). The Li ions occupy three distinct sites in the t-LLZO. The first one is the tetrahedral 8a site with distorted octahedral 16f and 32 sites (Fig. 3(f)) [42], which is thermodynamically stable at RT [43] and presents a total ionic conductivity of 10 -7 -10 -6 S·cm -1 and lattice constant of a = 13.134(4) Å, c = 12.663(8) Å, and c/a = 0.9641 [42,44]. The second one is the t-LLZO with a desired cubic structure (c-LLZO), space group 3 Ia d  , with a framework of 8-fold coordinated LaO 8 dodecahedra (24c) and 6-fold coordinated ZrO 6 octahedra (16a). This site possesses the higher ionic conductivity of 10 -4 S·cm -1 (Fig. 3(a)). In addition to the distinct value of ionic conductivity, the most significant difference between the two structures is the distribution of Li. In the cubic phase Li ions occupy two positions (Li1-tetrahedral void 24d and Li2-eccentric octahedral gap 96h) (Figs. 3(d) and 3(e)), while in the tetragonal phase Li is fully occupied (Li1 tetrahedral void 8a, Li2 octahedral gap 16f, and Li3-eccentric octahedral gap 32), which impedes the ion transmission between the adjacent positions [45]. However, as Li ions are unstable at Li3 sites, lithium ions primarily occupy the Li1 and Li2 sites, forming a Li-ion channel for t-LLZO ( Fig. 3(c)). Murugan et al. [33] were the first to report a c-LLZO (Li 7 La 3 Zr 2 O 12 ) synthesized at 1230 ℃ (a = 12.9682(6) Å; 3 Ia d  space group) with a face-centered cubic (FCC) anionic framework. Based on crystal lattice evidence of pure LLZO phases, the constant ranges from 12.95 to 12.97 Å [35,[46][47][48]. The phase stability of LLZO has become a challenge. LLZO can transform from cubic to tetragonal phase in a transition process called tetragonal distortion. The tetragonal distortion is a result of Li + redistribution that removes the short Li−Li interactions and disordered Li + . The 24th site in the cubic phase is transformed into fully occupied 8a sites and unoccupied 16e sites, while the 96h site is in the cubic phase. The cubic phase is then transformed into two 16f and 32g sites [43,49,50] (Fig. 4(a)). The lattice parameters documented in the literature range in a = 13.07-13.12 Å and in c = 12.67-12.72 Å in the tetragonal distortion. Several studies have shown that tetragonal distortions had an effect on the Li distribution and decreased the ionic conductivity over two orders of magnitude [33,42]. By density-functional theory (DFT) and molecular dynamics (MD), Bernstein et al. [43] argued that at low temperatures the tetragonal structure is formed and at higher temperatures, the 938 J Adv Ceram 2021, 10(5): 933-972 www.springer.com/journal/40145 cubic phase appears. Also, the transition temperature decreases with an increasing concentration of Li + vacancy, and the disrupted cubic structure has lower energy when the number of vacancy positions per formula unit is greater than about 0.4. Therefore, the tetragonal phase stability benefits from Li-ion sublattice ordering and volume preservation of tetragonal distortion, which generate the Li-Li gap and relieve the Coulomb repulsion [46].
Structural data obtained by in situ synchrotron X-ray diffraction (SXRD) of atomic layer deposition (ALD) showed that the phase transition of LLZO is related to an annealing temperature as low as 555 ℃ [52]. Rather low temperatures achieved the tetragonal phase of LLZO (< 650 ℃) (space group 1 4 / acd ) and ionic conductivity two to three orders of magnitude lower than the high-temperature cubic phase [47]. Experimental and molecular studies have reported the phase transition of Li 7 La 3 Zr 2 O 12 change through tetragonal to cubic phase between 177 and 750 ℃ [10]. Studies by Chen et al. [50] combined MD and DFT to simulate LLZO, and found two important factors of origin of the phase transition. The evolution of the Li migration pathway can be divided into four stages by temperature dependence of the occupancy of different sites and the effects on the Li distribution when supervalent elements are doped with LLZO. The four stages of Li migration pathways are oscillation, local migration, 2-dimensionallike migration, and 3-dimensional migration ( Fig. 4(b)) these combine occupancy and local structure by temperature. In the 2-dimensional-like migration, there are vacancies in 8a, 32g, and 16e sites, which are responsible of the formation of the Li migration pathways a and b (parallel), and c (perpendicular) dimensional directions. The a and b pathways intersect at 8 sites and Li-ions can be exchanged in these two directions, making the ionic conduction in the c direction relatively small and blocked by 16f sites [51]. When 16f sites participate in the Li conduction, the 3-dimensional migration is formed as a result of no blocking ions; in the meantime the gap of diffusivity between the ab plane and c direction is reducing, suggesting a blocking effect of 16f sites, and thus, reducing the temperature of the phase transition from the tetrahedral phase to the cubic phase [47]. These studies also revealed that Ta doping can stabilize the cubic phase at low temperatures by variating the site occupancy with different contents of Ta as a function of temperature [53,54].
Studies on phase stability of LLZO described the influence of the temperature in the phase stability. In addition to the high conductive cubic phase and the tetragonal phase, another phase was identified and called the "low-temperature cubic phase" [42][43][44][45][46][47]. Early studies suggest that the nature of the low-temperature cubic phase showed in Li garnet materials was caused by the sensitivity to humid conditions and the formation of a low-temperature cubic phase between the transition of cubic to tetragonal phase was cause by hydration mechanisms [55][56][57]. Recent studies of nuclear magnetic resonance spectroscopy (NMR) revealed that when we use Al-doped LLZO, aluminum (Al) cations slowly diffuse into the LLZO structure forming Li 5 AlO 4, which react with the tetragonal phase and transform into the cubic phase LLZO [58]. However, multiple Al environments may occur due to the disorder generated by the 24d Li site shown on the crystal structure, which results in the formation of the low-temperature cubic phase [58]. Recent confirmation of LLZO doping demonstrates the influence of Al and Ta elements on the formation of the cubic phase of LLZO at lower temperatures. In the structure of the high-temperature cubic phase (space group 3 Ia d  ) the Li sublattice is disordered with a partial site occupation, which influences the value of the Li + ionic conductivity to 10 -4 S·cm -1 at RT [28,59]. One main drawback of the high-temperature cubic phase is that it is not stable at RT and a high temperature of sintering is required (> 1150 ℃ ). Various chemical compositions are possible by doping at different sites; further doping with elements closer to ionic radius is used to stabilize the c-LLZO and improve the ionic conductivity, and this strategy is one of the most powerful approaches in considering the usability of SE [60,61].

Development process and synthesis of LLZO
As discussed above, the cubic phase exhibits the highest ionic conductivity and its stabilization is a critical issue. Furthermore, it has been shown that synthesis conditions, such as sintering temperatures, have a high effect on the crystal microstructure (e.g., crystal size, grain size, grain boundary, and bulk density, which can influence the ionic conductivity of LLZO), resulting in different ionic conductivity and electrochemical performance [62][63][64]. The total ionic conductivity of LLZO is affected by the grain www.springer.com/journal/40145 boundary resistance due to the impure phases at the boundary [65]. Therefore, increasing the density of the LLZO electrolyte is essential to improve ionic conductivity, as it not only reduces the grain boundary, but also increases the mechanical strength [66]. Adding sintering aids and applying different sintering methods, such as hot-pressing and discharge plasma sintering, can be an alternative to solve this issue. Various methods have been used to synthesize LLZO, being the conventional solid-state reaction the most commonly utilized in laboratories [67]. However, this method requires high temperatures and high energy consumption, which leads to reactivity with the reaction vessels. Other chemical techniques to synthesize dense LLZO are the sol-gel and co-precipitation methods [68]. These permit good ionic conductivity at low sintering temperatures and can regulate the microstructure, hot-press sintering [69], and field-assisted sintering [65]. Sophisticated sintering methods allow to optimize the performance of the bulk LLZO, and some examples include electrospinning [70,71], thin-film preparation technology [72], and spark plasma [73] ( Table 2). The most common method of synthesis yields a high ionic conductivity in the cubic phase of LLZO and, as a result, enhances Li ionic transport in the SE [8]. However, current research shows that existing preparation methods of LLZO still present some challenges and limitations, such as availability of commercially reagents, problems to densify in a thin format to reduce ohmic resistance, and incorporation into a battery with a suitable mixture of electrode materials. After all, LLZO has a strong potential to become a promising electrolyte in the development of future ASSLBs.

Challenges on LLZO air stability
Theoretical calculations and experiments revealed that garnet electrolytes are unstable in the air, to atmospheric moisture and carbon (IV) oxide (CO 2 ), leading to lithium carbonate (Li 2 CO 3 ) formation, ionic conductivity degradation, and high interfacial resistance [79,80]. Li 2 CO 3 easily forms a coat on the surface of LLZO because it can react with H 2 O and CO 2 in the air, as shown in Fig. 5. Cheng et al. [81,82] reported that a Li 2 CO 3 coating was formed on the surface of LLZO when LLZO pellets were exposed to air. Different mechanisms have been reported to describe the reaction between garnet electrolytes and air. The most satisfactory mechanism involves a two-step reaction.
The contaminating layer formed on the surface of LLZO decreases the ionic conductivity and induces an increase of interfacial resistance of the electrolyteelectrode interface. Studies from Kobi and Mukhopadhyay described that spontaneous cracking of LLZO occurs during storage of c-LLZO pellets in ambient air atmosphere for a few weeks possibly due to the formation of La 2 Zr 2 O 7 in the LLZO bulk [87]. X-ray diffraction studies indicate the formation of Li 2 CO 3 and LaAlO 3 just the 3rd day ahead upon exposure to air following by the formation of La 2 Zr 2 O 7 . The XRD pattern recorded in the pellet exposed to air for 6 days shows another additional small peak corresponding to cubic La 2 Zr 2 O 7 ( Fig. 6(b)). Importantly, after 21 days of exposure to air, the intensity of the La 2 Zr 2 O 7 peaks increased significantly to the extent that it appeared to be the primary phase, co-existing with the original cubic garnet LLZO phase, and other impurity phases ( Fig. 6(c)). 56 24 16  Researchers discovered some factors that affected air stability of garnet electrolytes, like humidity, grain size, the amount of grain boundary, and relative density [85]. Sharafi and co-workers [83] exposed LLZO in the air at relative humidity (RH)  50% and RH  0.5% for 240 h. The LLZO exposed to dry air (RH ≈ 0.5%) presents a lower carbonate peak intensity at 1100 cm -1 when compared to LLZO in ambient air (RH ≈ 50%) exposed for the same amount of time (Fig. 7). This data indicates that the moisture content during exposure plays a significant role in the formation rate of Li 2 CO 3 on the LLZO surface [84]. Also, the microstructure of LLZO: the pores, grain size, and grain boundaries influence the air stability. Xia et al. [88] prepared LLZO pellets in both Al 2 O 3 and Pt crucibles. The pellets sintered in Pt crucibles have large grains and reduced grain boundaries compared to those in Al 2 O 3 crucible. The formation of Li 2 CO 3 was shown by X-ray photoelectron spectroscopy (XPS) as a lower peak for LLZO pellets sintered on Pt crucible. Using Pt crucible is favorable for the air stability, enhancing the ionic conductivity and relative density. However, Cheng et al. [82] demonstrated that smaller grain size (≈ 20 µm) leads to lower formation of Li 2 CO 3 and higher air stability for LLZO pellets sintered on Al crucibles. By X-ray absorption spectroscopy (XAS) lower peaks of Li 2 CO 3 of the small grain size LLZO are observed. They conclude Fig. 7 Raman analysis of LLZO before and after exposure to ambient and dry air. The dotted line highlights the growth of the Li 2 CO 3 layer on LLZO as a function of exposure time and RH. Topographic analysis of LLZO exposed to air (RH = 50%) for 240 h. Reproduced with permission from Ref. [83], © The Royal Society of Chemistry 2017. this difference is due to that Al tends to segregate at grain boundaries for the small grain size LLZO, resulting in an Al-rich surface and the variation of the Al and Li distribution at the sample surface. More research is needed to clarify this topic.
Mechanical polishing [84], etching [89][90][91], and microstructure modification are effective methods to improve the air stability and interfacial issues of LLZO. The increase of interfacial resistance is caused by the Li 2 CO 3 insulating surface layer, which can be removed effectively after surface polishing. Both thermal and chemical etching, can remove Li 2 CO 3 . Ruan et al. [90] demonstrated that H 3 PO 4 can also react with the Li 2 CO 3 /LiOH passivation layer of the LLZT surface, and form a uniform Li 3 PO 4 modification layer. This procedure transformed the contamination layer into SEI film which not only promoted interfacial wettability but also suppressed Li 2+ dendrite penetration from the surface of LLZO [90]. For instance, Xia et al. [88] recovered 94.1% ionic conductivity of pellets stored in humid air for 6 weeks by removing the top reaction layer, that had originally caused a decrease in ionic conductivity to 55.9%, before surface polishing. Abdel-Basset et al. [91] demonstrated that thermally etching at 900 ℃ for 6 h partly removes the Li 2 CO 3 layer as well as chemical etching using 0.5 M HNO 3 to remove the Li 2 CO 3 layer for 15 min. Li et al. [89] developed a method to completely remove the Li 2 CO 3 layer. In this study, the garnet was incubated with carbon-treated at 700 ℃ for 10 h which resulted in pellets with ultralow interfacial resistance. The lack of air stability of the garnet-type electrolytes may contribute to the high interfacial resistance hampering its development. Material properties, synthetic conditions, and the humidity in ambient air play an important role in the air stability. Approaches to mitigate the formation of the Li 2 CO 3 need to be developed. affect the ionic conductivity of Li + [47]. Thus, in addition to the effect of sintering temperature, the control of chemical stoichiometry and Li + concentration in LLZO is important to improve the ionic conductivity [51]. Li + conductivity is determined by the concentration and mobility (π) of the divalent cation. The properties of LLZO and doping mechanism can be expressed by the n , e , and  represent Li + concentration, elementary load, and Li + mobility, respectively [43], and the elementary load is the Li + conductivity constant.
One way to improve the conductivity of LLZO is modifying the Li + concentration or the mobility of the cations [53,93]. This can be achieved by element doping or substituting ions following the specifications listed in Table 3. A considerable number of publications demonstrate that doping is increasingly being used in the synthesis of LLZO. Elements such as Fe, Ga, Al, Sr, Y, Ti, Ta, Sb, Mg, Sc, Zn, Ru, W, Nb, and Te [32,48,62,[92][93][94][95][96] have been demonstrated to promote the stabilization of the c-LLZO phase at RT by reducing the Li concentration or increasing Li vacancy content, which result in an improved Li 2+ -ion conductivity [97,98]. Based on various doping sites of LLZO, doping may be classified into lithium-, lanthanum-, or zirconiumsite. The lithium-site doping uses high-valent ions to replace Li from the sites, producing lithium vacancies that alter the concentration of Li 2+ [63]. Through lanthanum-and zirconium-site doping, the size of the LLZO framework structure can be changed to adjust the size of the Li 2+ ion migration path, ultimately changing its mobility [29,44]. In addition to affecting the size of the Li 2+ -ion migration path, the concentration of Li-ion can be also affected by the lanthanum-and  [114] www.springer.com/journal/40145 zirconium-site doping ions if these are non-equivalent [94,95]. Therefore, by modifying the framework structure of LLZO, the ion mobility and lithium-ion concentration can be modified, which will result in different electrochemical properties [93]. This will render a solid electrolyte with higher ion conductivity and longer battery cycle performance [40].

1 Control of Li-ion concentration
The concentration of Li 2+ can be adjusted individually by adding various amounts of lithium sources, or by combining different forms and quantities of highvalent elements at the lithium-site [62]. Usually, the Li contained in the LLZO structure is classified as Li3, Li5, Li6, and Li7 based on chemical composition [34]. Garnet-type materials can accommodate cations of different valence states and different sizes without any major change in the symmetry [95][96][97][98].
As described above, lithium doping uses high valence ions to replace lithium sites, to form lithium vacancy that enables to adjust the concentration of lithium-ion (it is easier to select a particular valance state between the dopant and the Li-ion) [97]. Ion substitution can maintain the chemical balance of oxygen and reduce the Li content at the same time, or increase the concentration of Li vacancy to stabilize the c-LLZO [98,99]. The exact number of doping required has not been determined, but is generally considered between 0.125 and 0.500 mol lithium vacancy per unit [43,62]. LLZO with a higher concentration of Li vacancy (0.4-0.5 mol) exhibit maximum ion conductivity at RT [115]. It has also been reported that the incorporation of some elements into the lithium-site will change the LLZO spatial group, thus changing its mobility [93]. Thus, it is important to understand the mechanisms of doping at different sites to improve the development of LLZO SEs with higher values of Li ionic conductivity.
Single c-LLZO exists when the Li content is between 5.63 and 6.24 mol combined with an Al concentration of 0.24 mol (1.5 wt%). Rangasamy and co-workers [62] reported that as the Li + content increased from 6.24 to 7.32 mol, a phase transformation from cubic to tetragonal LLZO occurred, where a sample with 6.24 mol of Li + exhibited the highest Li + conductivity of 4.0×10 −4 S·cm −1 . However, La 2 Zr 2 O 7 and LaAlO 3 were also present in the samples, contributing to the lower concentration of Li + than the actual value. Li + vacancy distribution in un-doped LLZO and Ga-doped LLZO are shown in Fig. 8. Gallium preferred to localize at the LiO 4 tetrahedron and introduced two more Li + vacancies around [97,116,121]. Xiang et al. [116] showed that doping with Ga and different Li + concentrations of LGLZO (Li 6.4 Ga 0.2 La 3 Zr 2 O 12 ) that revealed different amounts of Li + in the lattice and a change of Li + conductivity and activation energy (E a ). With the rise of Li + concentration, the Li + conductivity rises initially and then drops. Furthermore, when the amount of Li + is 6.55 mol, the lithium ionic conductivity reaches a maximum of 1.09×10 −3 S·cm −1 , while the sample with 6.91 mol Li + exhibits the minimum Li + conductivity of 4.3×10 −4 S·cm −1 . Correspondingly, the E a decreases initially but increases afterwards. When utilizing 6.55 mol of Li + the E a reaches a minimum of 0.22 eV, compared to the sample with 6.91 mol that shows a maximum of 0.28 eV [116]. This indicates a proper Li + concentration improves the conductivity of Li + [97]. Thus, the effect of the Li + concentration has been widely investigated as an important factor influencing the Li + conductivity of LLZO. By adding different amounts of Li 2 CO 3 , Liu and co-workers [117] studied the effect of an excess amount of Li 2 CO 3 (from 0 to 50 mol%) in starting materials when preparing LLZTO in alumina crucibles. XRD experiment rendered patterns showed that the excess of lithium salt is not crucial for achieving the cubic LLZTO, and that LiAlO 2 existed as the second phase in all the samples except for LLZTO-0%Li 2 CO 3. In this case, the interaction between excess Li 2 O 3 and alumina crucible formed the Li 2 O-Al 2 O 3 compounds [118]. Larger amounts of excess Li 2 CO 3 resulted in better densification of the pellets and increased the ionic conductivity, indicating ionic conductivity is influenced by density. Note that these two properties are critical factors in practical applications of garnet-type materials. Zhang et al. [65] used 5-20 wt% excess Li 2 O and prepared Al-doped LLZO samples with controlled lithium-ion concentrations and less grain boundary impurities by field-assisted sintering. The authors found that as the lithium-ion concentration gradually increased from 5.80 to 7.14 mol, the ionic conductivity initially increased, reaching the maximum at 6.35 mol, which was 5.56×10 -4 S·cm -1 ; after this peak, the conductivity decreased [65]. In silico experiments using density functional theory (DFT) and molecular dynamics (MD) calculations suggested that the introduction of 0.4 to 0.5 mol of lithium vacancies in Li 7 La 3 Zr 2 O 12 would be more conducive to the migration of lithium ions [43]. A refined mechanism utilizing Li 7 La 3 Zr 2 O 12 followed by neutron diffraction analyses demonstrated that that with the increase of lithium-ion concentration, the occupancy rate of lithium-ion in the octahedral vacancy gradually increased, while the occupancy rate of the tetrahedral vacancy decreased at first, followed to an increase of lithium content around 6.4 per formula unit ( Fig. 9(a)) [99]. Based on the higher octahedral vacancy rate and the higher degree of lithium-ion disorder in the tetrahedral vacancy (lower tetrahedral occupancy leads to higher disorder), migration of lithium ions is favorable, so the highest lithium-ion conductivity was achieved when the ion concentration was 6.4±0.1 mol. Aliovalent doping is commonly used to increase the Li content to optimize the ionic performance, for example in Li 7 La 3 Zr 2 O 12 , Li 3+x Nd 3 Te 2−x Sb x O 12 , and Li 5 La 3 (Ta/Nb) 2 O 12 [118].
A summary of the highest conductivities of un-doped, doped, and multi-doping of LLZO substitutions reported in the literature is presented in Fig. 9(b) [119]. Briefly, the highest Li-ion conductivity was achieved by the lattice parameters in the range of 12.91-12.98 Å, at Li contents between 6.1 and 6.8 per formula unit (highlighted on the red circle). Substitutions that result in lattice parameters outside this range show lower total Li-ion conductivity. The presence of an optimum lattice parameter was observed when Nb was replaced by Zr, and alkali earth metals of different ion sizes (Mg, Ca, Sr, and Ba) were replaced by La in the LLZO. The optimal lattice parameters of LLZO have been studied systematically by Kihira et al. [94]. The data show no change in the Li content in any component related to element doping; the group also described optimum lattice constants are between 12.94 and 12.96 µm, which are consistent with other reports for the garnet-type SEs, where the reported optimum lattice parameter ranged from 12.90 to 12.95 Å [120]. The highest conductivity obtained so far was 2.06×10 -3 S·cm -1 at RT for Li 6.55 Ga 0.15 La 3 Zr 2 O 12 [121], which is around an order of magnitude higher than the first LLZO synthesized in 2007 (3×10 -4 S·cm -1 at RT) [33]. These differences can be explained by the fact that simple lattice expansion may lead to the expansion of Li-ion hopping interstices, which eventually leads to high ion conductivity [116]. In fact, computational analyses have shown that the extended lattice provides slight improvements in conductivity [123]. Theoretically, the evidence suggests various Li-diffusion pathways to improve conductivity. For instance, in silico calculations and Li NMR suggest a mechanism by which the 24d Li-ions are immobile and Li migrates by hopping between the octahedral sites in LLZO with Li contents under 5 per formula unit [122][123][124]. Ab initio calculations support two pathways of Li-ion migration in cubic garnets that consider Li contents over 5 units per formula [34]. As shown in Fig. 9(d), in pathway A the Li migrates via the interstice between neighboring octahedral sites, by passing their common tetrahedral neighbor. In pathway B the Li-ions move through the famous shared triangular "neck" of octahedral and tetrahedral sites in the framework. The Li-ion migration in path A is favored when Li + content is lower (Li 5 La 3 Nb 2 O 12 ) and with an activation energy of 0.8 eV, while path B selects high Li + content (Li 7 LaZr 2 O 12 ) with the activation energy of 0.26 eV [25,31].
It has been postulated that the Li vacancy concentration is a crucial factor in the ionic conductivity of LLZO [118]. LLZO has two stable sites: a tetrahedrally coordinated one (Li1) and an octahedrally coordinated one (Li2). The migration between sites of the Li-ion is through a triangular "neck". The size of the octahedron/tetrahedron and the neck also affects the conductivity by changing the site's energies and/or the neck energy [97]. Studies on the channel size have attempted to elucidate whether a larger or smaller channel width is beneficial for ion transport [124]. This is an important consideration, as channel size has been defined as the correlation of the size of the lattice with the conductivity [120]. Therefore, large changes in the conductivity may indicate that separate control of the two structural aspects is essential for controlling Li-ion diffusion [125]. DFT calculations, and a variety of experimental techniques like synthesis methods, neutron diffraction, high resolution X-ray diffraction (XRD), Raman measurements demonstrated a relationship between structure and stoichiometry. These two features offer the possibility of controlling the ionic conductivity of doping LLZO, suggesting that co-doping optimizes the Li vacancy concentration and the lattice size simultaneously [126]. A study using neutron diffraction and theoretical calculations showed the influence of doping LLZO, on the two Li sites (Li1 and Li2) and the triangular "neck" window between them change [126]. The neck size increased when Ta doping increased. Neutron diffraction (ND) analyses showed the Ta ion pulls the neighboring O ions more tightly than the Zr ion. The TaO 6 octahedra and the LaO 8 contribute cooperatively to decrease the lattice parameter, while slightly increasing the volume of Li sites [126]. Doping LLZO with supervalent ions like Al, Ta, Nb, and Ga can create and balance the Li vacancies and thereby produce a cubic lattice improving the ionic conductivity [104][105][106]111]. It is accepted that dopants, such as Al 3+ , of the sintering process from the crucible can stabilize the cubic phase of LLZO [55]. Geiger et al. [59] studied the influence of Al doping and sintering temperature in the structure of Li 7 La 3 Zr 2 O 12 . The crystal structure of LLZO allowed identifying two Li binding sites without doping and low sintering temperature, which constituted the tetragonal phase. Kotobuki et al. [127] reported that Al 2 O 3 is an effective sintering additive, as the sintering temperature was reduced and the impurity formation of La 2 Zr 2 O 7 was inhibited by this compound. Doping with Al compounds can expand the lattice filling with tetrahedral Al 3+ ions, increasing the bottleneck size [55,58,62]. The position lattice was studied by density functional theory and NMR of the Al in the c-LLZO; the data showed Al 3+ could have a number of slightly different local 4-fold coordinations around 24d and 96h sites in cubic LLZO: these suggest the facility of garnet structure to create intersices and enhance the Li + diffusion [127][128][129]. Researchers have used computational and experimental results to understand the stoichiometry and crystal structure, and enhance the optimization of LLZO using elemental substitution. Study about the performance the Li-ion conductivity has been the main goal to develop garnet solid electrolytes commercially available for battery fabrication.

2 Regulation of ion mobility
Li-ion mobility is considered as one key requirement for a good electrolyte. Although high mobility of Li + in solid materials has been encountered at elevated temperatures, room-temperature mobility and chemical stability remain as significant challenges in the field [130]. These issues can be solved by element doping, at the three sites possible of LLZO: Li site, La site, and Zr site. As we mention in the previous section, lithium-site doping was only used to regulate the concentration of lithium-ion. Al 3+ is widely used as a common element to control lithium-ion concentration and stabilize cubic lattice. Huang and coworkers [131] used the Al-doped LLZO lithium site, and obtained an ionic conductivity of 3.6×10 -4 S·cm -1 at RT. However, in recent years, researchers have discovered that certain lithium-site doping elements can not only regulate the concentration of lithium ions and stabilize LLZO in their cubic phase, but also increase the mobility of Li ions by changing their space group, thereby improving the lithium-ion conductivity [131][132][133][134]. For instance, the compound Li 6.25 Ga 0.25 La 3 Zr 2 O 12 exhibited a conductivity of 1.46×10 -3 S·cm -1 at RT through Ga doped on the Li site of LLZO, resulting in a mobility for Li 7−3x Ga x La 3 Zr 2 O 12 of 10 −8 -10 −7 cm 2 ·V -1 ·s -1 at 60 ℃. This is significantly higher than the mobility observed for W-doped LLZO (10 −9 cm 2 ·V -1 ·s -1 ) [93]. Calculations and in silico simulations by Garcia Daza and co-workers [132] proposed that in the Al-doped LLZO, the Li ions near the Al ion are unable to migrate; conversely, the Li ions near the Ga ions in the Ga-doped LLZO are mobile. Further XRD analyses of Ga doped LLZO by single crystal particle, demonstrated that the space group of the cubic phase structure Ga doped LLZO is 43 , which is different from the traditional cubic phase LLZO, 3 Ia d  space group [133]. This structure and the cubic phase of LLZO 3 Ia d  space group are all formed by ZrO 6 octahedron and LaO 8 dodecahedron, and Li ions filled the vacancies [28]. The difference is that Li1 (24D) tetrahedral vacancy and Li2 (96h) octahedral vacancy is formed by the cubic phase LLZO of the 3 Ia d  space group, while Li1 (12a) and Li2 (12b) occupy the tetrahedral vacancy, and Li3 (48e) the octahedral vacancy forming the cubic phase of LLZO of the 43 space group (Fig. 10). Rettenwander and coworkers [101] developed the Li 6.4 Al 0.2-x Ga x La 3 Zr 2 O 12 compound and found that increasing the Ga content favored the ionic conductivity,  . On the other hand, Wagner et al. [100] found that doping with Fe renders a similar crystal-chemical behavior to Ga-doping. When Fe doping at 0.18 and 0.25 pfu the cubic LLZO space group obtained was also 43 I d  and the ionic conductivity of the sample of 0.18 pfu reached a maximum of 1.38×10 -3 S·cm -1 . Studies regarding the doping of LLZO lithium sites with Al, Fe, and Ga focused on the influence of the change of space group on the migration of LLZO lithium ions by neutron diffraction [134]. The analyses demonstrated that the cubic phase LLZO with space group 43 I d  has multiple Li-ion migration channels, and the Al ion with space group 3 Ia d  cubic phase LLZO will not block the Li ion migration channels, resulting in lower activation energy and improved Li ion migration (Fig.  11).

Change of bottleneck size on mobility
The structure of LLZO is formed by the common edge of ZrO 6 octahedron and LaO 8 dodecahedron, where Li ions filled the voids in the framework, allowing the size of the lithium migration channel to be modified by www.springer.com/journal/40145 doping at the La or Zr sites (Table 3) [57,60,61,75,93,95,[99][100][101][102][103][104][105][106][107][108][109][110][111][112][113][114]. There are few reports about lanthanum doping. The large size of lanthanum ions and the strong influence on the crystal structure of LLZO make it difficult to find suitable doping ions for La site [60,95]. However, some studies showed that doping with Sr at the La-site the ionic conductivity at RT of LLZO increased from 2.1×10 -4 to 4.95×10 -4 S·cm -1 [60]. The development of the Li 7 La 3 M 0.25 Zr 1.75 O 12 (M = Ge 4+ , Ti 4+ , Sn 4+ , Hf 4+ ) compound allowed to investigate the influence of the Zr doping ion radius on the bottleneck size and activation energy of the Li-ion migration channel by synchrotron radiation [50]. The data showed that by increasing the Zr doping ion radius, the M-O bond length increased, as well as the Li vacancy polyhedron with the same edge increased, and the bottleneck size (the circumscribed radius of the smallest triangle of the migration channel) [104][105][106][107][108][109][110][111]. Consequently, as the M-O bond length rises, the volume of the MO 6 octahedron increases, so do the cell parameters ( Fig. 12(a)). Another cation that has been investigated in La-site doping on LLZO is Ce 4+ . Ce 4+ concentrations over 0.2 mol stabilized the LLZO to the cubic phase at RT; however, its ionic conductivity was only 1×10 -4 S·cm -1 [95]. There are further reports on Zr-site doping, and there are several elements that can be integrated into it and enhance its ion conductivity. Doping the Zr-site of LLZO with alkaline earth ions (Mg, Ca, Sr, Ba) into the octahedral sites (Zr 4+ ) instead of La 3+ -sites resulted in a relaxed the crystal structure with enlarged the bottleneck, but improved ionic conductivity [135]. Bachman et al. [136] summarized published literature on findings by several groups and proposed a general rule applicable to all solid electrolytes. The proposal is that the size of the Li-ion migration channel and the cell parameter volume increase if the doping ion radius in the framework structure increases, while the activation energy decreases facilitating the lithium-ion migration. Consistently, Wang and coworkers [137] calculated the crystal framework and the diffusion coefficient of Li-ion by using the first principles [137]. They found that with the increase of framework volume, the ion diffusion coefficient increases. Simulations of a cubic phase LLZO model where the cell volume was changed by adjusting the external force at 1000 K, showed that the conductivity of Li-ion was enhanced as the cell parameters were increased [32]. However, variations of metal concentrations also rendered changes in sizes. Nemori et al. [104] [104,105]. When Ca and Ta were used to dope La-and Zr-sites to control the size of the Li-ion migration channels, it was observed a decrease in the lattice parameters with the increase of co/doping content, suggesting that the Li + migration path size of LLZO decreases ahead [138]. The study also showed an initial drop in the activation energy which gradually increases at a later time. The study concluded that the movement of Li ions is not conducive to the movement of too large or too small migration channels [135,138]. Xiang and co-workers [139] developed the Li 6.4 La 3 Zr 1.4 M 0.6 O 12 (where M represents Sb, Ta, Nb) compound, and investigated changes in particle sizes by neutron diffraction. The study showed that the pentavalent ions are doped by a stronger inter-particle Coulomb force produced in the reaction. As shown in Fig. 12(b) the radius of the hetero ions, the length of the M-O, and the distance between the lithium-ion and the migration channel increase, while the length of the La-O bond is compressed, and the bottleneck size and unit cell volume gradually decrease [139]. Thus, the activation energy is coordinated by the size of the bottleneck and the M-O bond length [139,140]. On the other hand, the bond length of M-Li2 became larger with the increase of radius of doping ions, consistently with the increase of M-O [139]. Figure 12(c) shows the structural representation of the bond M-O length indicating the difference between the M 5+ and Li + migration channel.

Coordinated regulation of lithium-ion concentration and mobility
As discussed above, the ionic conductivity of LLZO can also be improved by simultaneously adjusting the Li-ion concentration and changing the skeleton and then adjusting the mobility. When using the heterovalent ion Ta to dope Zr-sites and concurrently modulating the Li-ion concentration and migration channel size, an inverse correlation was found. The increase in Ta doping (up to 0.6 mol) leads to a drop in the concentration of Li-ion and reduction of the migration channel size, and the activation energy decreased resulting in the ionic conductivity to be 1×10 -3 S·cm -1 at RT [111]. A cubic phase Li 6.10+2y Ga 0.3 La 3−y Rb y Zr 2 O 12 compound where the Rb concentration was increased, resulted in an increased concentration of Li-ion and migration channel size; the activation energy increases until the content of Rb doped is 0.05 mol. Under these conditions, the activation energy is the lowest, and the ion conductivity is high as 1.6×10 -3 S·cm -1 [112]. Buannic et al. [113] reported the cubic phase of Li 6.55+y Ga 0.15 La 3 Zr 2−y Sc y O 12 , where the increase of Scandium (Sc) doping increased the concentration of the Li-ion concentration and the migration channel size, similarly to other metals; in this case the activation energy exhibited an initial decreased followed by an increase. A concentration of 0.1 mol of Sc resulted in the lowest activation energy and an ion conductivity of 1.8×10 -3 S·cm -1 . Under these conditions, the LLZO structure is most stable when the occupancy rates of Li1 and Li2 are 46% and 47%, respectively [47]. Zeier [120] has shown that controlling the unit cell parameters in the Li 6 MLa 2 Ta 2 O 12 compund and substituting the La-site leads via chemical pressure of the Li-O polyhedra pathways, ultimately lead to a change of the Li + mobility. Thus, the evidence supports a mechanism by which co-doping can stabilize the cubic junction and can effectively promote the transport of Li ions. DFT was used to measure the possible locations of selected dopants in the LLZO, and was proven to be a good method for finding LLZO-based SE with excellent performance [141]. In this case the doping elements included Al, Fe, Ge, and Ga to replace Li, and Sr, Y, and Ce to replace Zr. All of these are amongst the most common elements that can be incorporated to LLZO (Table 3). Al 3+ promotes the lattice distortion allowing 3 Li vacancies and the migration in the tetrahedron site is more difficult for Li-ions [129]. Although the occupancy of the tetrahedron site with Al 3+ has a blocking effect on the Li-ion conduction, it rendered a consistent increase of the bottleneck size [74,156]. Furthermore, the introduction of Al 3+ in the crystal structure increased the conductivity [73][74][75]82]. Thus, it is essential to regulate the concentration rate during element doping, since these may alter the Li concentration [94,103].
Recent work targeted the substitution of Ga 3+ on the tetrahedral Li site [93]. This substitution also increased the ionic conductivity. The crystal structure of c-LLZO showed the local structure of the molecule with increased conductivity [116,121]. This was associated with the enhanced disorder, the transport route change from 96h→24d→96h to 48e→12a (12b) →48e and 48e↔48e and the Li content on the mobile octahedral site [116,142]. The amount of Li content of LLZO with the highest conductivity was 6.1-6.8 per formula unit, lower than the theoretical limit of 7.5 per unit [143]. The range of Li concentrations in the Li 6.55 Ga 0.15 La 3 Zr 2 O 12 compound represents the highest ionic conductivity of 2.06×10 -3 S·cm -1 at RT [120]. In addition, MD simulations calculated by Jalem et al. [144] suggested that the ionic conductivity of Li 7-3x Ga x La 3 Zr 2 O 12 (x = 0.02) was 6.08×10 -3 S·cm -1 , which is one of the highest values reported for LLZO electrolytes. Nevertheless, it is still lower than liquid Li electrolytes (∼1× 10 −2 S·cm −1 at RT) [136]. The size of the bottleneck in the study of the LLZO solid electrolyte structure [139], precisely refers to the circumscribed circle radius of the smallest triangle of the Li-ion migration channel. The radius of the circumscribed circle has a major effect on the migration of Li. As we know, ZrO6 octahedron and LaO8 dodecahedron form the framework structure of LLZO through the common edge, where Li ions fill the voids in the framework [140]. Therefore, modifying the shape of their structure and the size of bottleneck by doping at the La-or Zr-sites results in an altered Li-ion migration channel [138][139][140]144].

Applications of garnet-based LLZO in all-solid-state lithium batteries
In recent years, significant improvements have been made to the performance of ASSLBs using LLZO electrolytes [36,40,79]. Research has addressed problems associated with the electrode/electrolyte interface and increasing their energy and power densities (Fig. 13) [149][150][151][152][153][154][155][156]. The high interfacial resistance between electrolyte and electrode has become an urgent task to solve, as it possesses significant influence on the overall performance of the battery [8]. The study of electrochemical reactions is a major research area to understand the contact interphase of the solidsolid electrolyte-electrode interface, as insufficient contact between these leads to the formation of a solid electrolyte interface (SEI) [146]. Li ions diffuse from electrolyte to electrode through their interconnected region, and redox reactions with active materials and electrons happen at the electrolyte-electrode interface [145,150]. To ensure that the charge transfer reaction continues steadily, it is important to maintain an efficient solid-solid electrolyte-electrode interface in the battery all the time [149]. A major challenge to be solved is how to improve the ionic conductivity at the electrode-solid electrolyte interface [150,151]. Despite that, studies focusing on Li-ion migration, diffusion behavior across the interface, and understanding the interface at the sub-molecular and atomic-level, are still limited [152]. Another important factor affecting the ionic interphase is the mechanical properties of the SEs in terms of fragmentation of electrode materials [145]. The interfacial contact between the active electrode and the SE represents critical elements during battery design, assembly and manufacturing, and overall battery performance [40]. Table 4 lists the fundamental parameters for the performance of ASSLBs. There are two basic strategies to use LLZO as a solid electrolyte for ASSLBs. One is used LLZO-based solid electrolytes as bulk ceramic [119]. However, they usually exhibit larger interfacial resistance, poor battery performance due to weak contacts between the electrode-electrolyte, side reactions with the electrolytes, and moisture/CO 2 , which lead to low battery performance [149][150][151][152]. Nevertheless, developing a composite solid electrolyte (SCE) by combining LLZO with a solid polymer electrolyte (SPE), has become a research spotlight due to the advantages these represents [146][147][148]. For instance, SCEs exhibit a wide electrochemical window, favorable mechanical strength, stability, and remarkably improved ionic conductivity [157,158]. On the other hand, challenges exist on the interfacial resistance of each material. According to the previous development of the garnet-type LLZO solid electrolytes, it is well known that the high conductivity is a necessary but an insufficient property to design future ASSLBs.

1 LLZO-based all-solid-state lithium batteries
LLZO electrolytes have a high elastic modulus with small contact area between them and the electrodes [164]. It is possible that structural defects, such as voids and cracks at the interface, or even a third phase during the charge and discharge cycles impair the performance of the device [165]. The unideal interfacial contact leads to a high interface resistance, thus reducing the Li + migration dynamics at the interface [161]. One of the first studies with LLZO-based ASSLBs and Li showed that Li-ions move through LLZO-based electrolytes without Li accumulation on the surface of LLZO [165], where the interfacial issues result in a low discharge capacity of 15 μAh·cm −2 [35]. An alternative cell was assembled with LCO|LLZO-Nb|Li, and showed stable cycling performance and comparable interfacial impedances with traditional batteries; 98% of the initial capacity was retained after 100 cycles [166]. Another strategy to overcome interfacial issues was the development of a co-sintering model of LLZO with LiCO 2 not having side reactions [167]. The performance of the battery was successful; this device charged and discharged with specific capacities of 98 and 78 mAh·g -1 [167]. Using an interface-engineered Li 6 [151]. In an ASSLB assembled with a mixture of LiMn 2 O 4 and LLZO-B 2 O 3 , the performance of the specific capacity was 102.6 mAh·g -1 on the first cycle, and kept a capacity retention of 94% after 20 cycles [168]. Using LLZO nanoparticles, Yan and co-workers [169] fabricated an ultrathin solid electrolyte film and the Li|LLZO|LiFePO 4 battery presented a discharge capacity of 160.4 and 136.8 mAh·g -1 in the first cycle and 100th cycle at RT, respectively [169]. This work demonstrated a high performance of SE layer with micrometer thickness that is suitable for applications of ASSLBs. From experimental and computational analyses, it was concluded that LLZO is favorable for ASSLBs with high energy and power densities [153,155]. However, interfacial issues and long resistances with anode and cathode electrodes are the bottleneck headache; especially the grain boundary chemistry of LLZO needs to be studied for further application of LLZO-based solid electrolytes.

Interfacial issues between LLZO/anode
Li-metal is the most attractive anode material possess a high theoretical specific capacity (3860 mAh·g -1 ), the lowest redox potential (3.040 V vs. standard hydrogen electrode), and low density (0.534 g·cm -3 ) [170]. Insufficient contact between electrolyte and electrodes leads to form an SEI on Li metal anode, decreasing the capacity and low coulombic efficiency (CE) [171]. Also, irregular deposition of Li metal can form Li dendrites, go through the current collectors, and mix with the products of reactions happening in the SEI during the cycling, resulting in active material losses, increasing impedance, and with the possibility of short circuit, affecting the performance of the ASSLBs [180]. The area-specific resistance (ASR) is used to define impedance, which depends on the area of ionic contact between the solid electrolyte and the active material (electrodes) [189][190][191]. In situ computational analyses of LLZO suggest an excellent electrochemical stability with Li metal [21]; however, experimental evidence reports high ASR, and thus, many strategies have been developed to improve the contact of LLZO and Li [177]. Atomic layer deposition (ALD), sputtering, heat or wet treatment, evaporation, polishing, among others [149,153,164,[171][172][173][174] point towards the challenges and requirements for long-term stability, lifetime (in terms of the number of discharge-charge cycles of the battery), minimizing the Li dendrites, and reducing the costs of processing technologies to be consider at the Li|LLZO interface.
Research revealed a slight thermodynamic stability of Li 7 La 3 Zr 2 O 12 although at very low potentials (0.05 V vs. Li 0 ) and with low reaction energies (20 meV per atom) [21,175]. So, the observed stability of LLZO electrolytes against Li 0 is likely the result of the kinetic stabilization [175]. One of the simplest ways to improve the interface contact is Li heating and melting; after heating, the ASR is lower to 25-28 Ω·cm 2 at RT [176]. The formation of the Li 2 CO 3 passivation layer inhibits the adhesion of Li to the surface of LLZO [89]. The interfacial instability of the Li interface, which is determined by in-situ electron microscopy when LLZO is in direct contact with Li metal [177], showed that upon contact with Li, a layer of t-LLZO is formed in the interphase (~5 unit cells), and prevents full degradation of c-LLZO while promoting the Li + transport.
A strategy using heat or wet treatment by Zhou et al. [237] incorporated a wet process technology to treat Zn. A (NO 3 ) 2 solution is uniformly distributed on the surface of LLZO and thermal decomposition results in the formation of ZnO; thus, the ASR between Li and LLZO is also reduced effectively. Sharafi and co-workers [178] used the wet method before heating the Li metal supplemented by glycol-based additives, as strategies for polishing the LLZO surface. As a result the interface resistance of Li LLZO was reduced to 2 Ω·cm 2 . Polishing is another method used to improve the interface contact (Figs. 16(a) and 16(b)). Fu and co-workers [179] polished and ground LLZO   [179]. LiF can also be used to increase stability of LLZTO and to suppress the formation of Li 2 CO 3 . The LLZT-2LiF shows a lower interfacial resistance of 385 Ω·cm 2 , also with high coulombic efficiency 93% of its capacity after 100 cycles [161]. Another strategy to promote better interface contact is to introduce carbon as active electrodes and cycling them against the Li metal [180,181]. Mixing LLZO with carbon improves the electronic contacts and facilitates the intercalation of the mobile cations into and out the solid electrolyte on its reduction-oxidation process [182]. Mathematical analyses were used to understand the increase in wettability between Li-C composite and garnet electrolyte (Fig. 14(c)). The study showed that the Li-C composite and garnet present good interface stability upon electrochemical cycling [153]. The calculations favor a reaction between Li-C composite and garnet rather than Li-garnet. However, attention needs to be given to the surface contamination layer of Li 2 CO 3 , as this may affect the wettability of Li on LLZO [150,178]. Using composite materials is an alternative strategy to improve the interfacial contact [36]. A composite anode electrode was fabricated using Li-C and showed an improvement in the contact boundary between Li and solid electrolyte [153]. The composite is continuously stirred and melted on a hot plate at 250 ℃, and Li was prepared by adding graphite powder. The ASR with pure Li electrodes calculated is 381 Ω·cm 2 and the ASR obtained between Li-C/LLZO/Li-C was decreased to 11 Ω·cm 2 , one of the lowest values reported thus far [153]. A recent design of a composite polymer electrolyte (CPE) consisted of SHP and LLZGO NPs, called "hybrid electrolyte" [183]. This compound presents strong adhesion and a seal-healing functionality design for stabilizing the Li anode [183]. In this model, the CPE membrane was soaked with liquid electrolyte (LiPF 6 ) to the lithium, and SEM characterization showed a good interfacial contact between them. The study only reports the resistance associated with the passivation layer impedance on the Li surface of 92.1 Ω after 500 cycles, and the hybrid electrolyte only 6.9 Ω after 500 cycles [183]. Apart from inhibiting the dendritic growth of Li and the improved the electrochemical performance, the use of a hybrid electrolyte can enhance the high energy density and extend the cycling life, with also a better safety performance for ASSLBs. Composite material coating of metal layers is another strategy to solve the interfacial contact [36,86]. Tsai and co-workers [69] deposited a layer of about 20 nm of Au on the surface of LLZO using an ion sputtering coater. This strategy reduced the ASR of the Li anode to 58 Ω·cm 2 [69]. Xiang and co-workers [184] deposited a Cu film of about 100 nm on the surface of LLZO by magnetron sputtering and allowed to react with the Li negative electrode at 250 ℃ . This process resulted in an alloy at the interlayer with reduced ASR (from 677 to 29 Ω·cm 2 ). This lithium symmetrical battery can run more than 800 cycles at a current density of 0.2 mA/cm 2 [184]. To modify the Li|LLZO interface a Ge layer was coupled to the Li|LLZO|LiFePO 4 system [149]. This combination rendered a stable cycling performance at RT for the device, and delivered an ASR of 115 Ω·cm 2 , in addition to good stability. Wetting at the lithiathed Ge/LLZO interface was confirmed by mathematical modeling and fitting of the experimental data [149]. A recent report used a limited amount of Li anode, coupled to the garnet-type Li 6.4 La 3 Zr 1.4 Ta 6 O 12 (LLZTO) pellet [185]. In this case, the LLTO-cathode interface layer of solid-state plastic crystal electrolyte (PCE) and an anode-LLZTO interface layer of a gold thin film were prepared. The EIS measurements revealed that the ASR of Li|LLZTO ranged from 115 to 75 Ω·cm 2 after loading 20 MPa pressure, suggesting that the pressure can optimize the Li|LLZTO interface contact. However, for the Li|LLZTO|Li symmetry cell with Au thin film on LLZTO surface, the ASR of Li metal-LLZTO was 25 Ω·cm 2 , thus presenting an increased stability [185]. This study suggested a mechanism for Li loss, where the LLZTO SE first reacts with Li to form a stable SEI, and then a quite high Li loss is associated to the irreversibly trapped lithium in the copper current collector, as well as the contact loss of deposited lithium during the stripping process. It is well known that LLZTO reacts with Li to form Li 2 CO 3 and Li 2 O compounds on the SEI layer [182]. Au reacts with LLZTO and promotes enhanced molten Li diffusion into the garnet surface, improving the interface stability [183]. The study shows Li-Au alloy can be used as a Li-ion conductor between Li metal and garnet, which allows more uniform Li + flux and improves Li + transport. However, the formation of Li dendrites at high current densities (0.30 mA·cm -2 ) is inevitable [185]. ALD of Al 2 O 3 has become one effective method to reduce the ASR from 1710 to 1 Ω·cm 2 [144]. While the molten metal Li is on the surface of LLZO treated by ALD, the existence of the AL 2 O 3 layer increases the interface contact area and inhibits the formation of Li 2 CO 3 [144].
A chemical reaction between Zn(NO 3 ) 2 and Li produced a new solid-state interface that bonds together the Li 6.4 La 3 Zr 1.4 Ta 0.6 O 12 (LLZTO) electrolyte and the Li metal ion (denoted as the ZNR interface) [38]. This interface allows efficient conductive pathways for Li + transport through the interface without dendrite formation or side reactions with the electrolyte observed after 1000 h of continuous charging and discharging. The ASSLB Li||LiFePO 4 (Li||LFP) also showed a stable capacity of 150 mAh·g -1 at 1 C rate for nearly 400 cycles, which represents the highest performance reported to date [38]. Han and coworkers [171] deposited a layer of 6 nm Al 2 O 3 on the surface of Li 7 La 2.75 Ca 0.25 Zr 1.75 Nb 0.25 O 12 by ALD and heated it with the lithium anode at 250 ℃ for 1 h to form an alloy intermediate layer. The ASR of the lithium negative electrode was reduced from 1720 to 1 Ω·cm 2 at RT. Experimental and computational analyses revealed that the oxide coating enables wetting of metallic lithium in contact with the garnet electrolyte surface, and that the lithiated-alumina interface allows effective lithium-ion transport between the Li metal anode and garnet electrolyte [171]. Although the high contact area ALD is a very effective method to improve the interface anode contact, simple polishing becomes more attractive for practical applications.
(1) Li dendrite on LLZO Lithium dendrites can still be found in LLZO SEs; these tend to form along grain boundaries and SEs with voids [37,186]. There are two major mechanisms for the formation of Li-dendrites. The first one initiates with the nucleation of Li dendrite at the Li anode/ LLZO electrolyte interface, which then passes through the interstitial space of LLZO electrolyte particles [187]. The second mechanism proposes the formation of dendrites within the solid electrolyte [188]. At high currents, the formation of dendrites causes the Li/LLZO/Li cell to short circuit. The critical current density, at which the cell will be shorted, is less than 0.9 mA·cm −2 at RT [69,189,190], while the critical current density of liquid electrolytes can reach 4-10 mA·cm -2 at RT [191]. Evidence suggested that doped LLZO becomes partially lithiated when interacting with Li [170]. This lithiation causes a phase transition from the cubic to the tetragonal phase [177,186]. The macroscopic ionic conductivity of the tetragonal phase is lower but barely limits conduction as a nanometric film, constituting ideal interphase [187]. Han and co-workers [188] proposed that short-circuit formation in garnet SSEs is caused by internal Li deposition led by high electronic conductivity. Since the electronic conductivities of garnet SSEs increase with temperature, the evidence showed that Li dendrites are easier to form at high temperatures. When the temperature increases (from 20 to 100 ℃), the ionic conductivity of garnet SSE increases by one order of magnitude, and the electronic conductivity increases by two orders of magnitude, from 6 × 10 −8 to 1 × 10 −7 S·cm -1 [188]. At higher temperatures, the electronic leakage current to ionic current ratio should be lower, resulting in less internal Li deposition. However, according to other sources, the lithium dendrites form in LLZO, regardless of the dopants used to stabilize the cubic structure [191]. Also, the low relative density of SEs was thought to be the primary cause of Li dendrite growth, but there is no connection between dendrite suppression and the relative density LLZO [69]. Indeed, Li dendrites keep growing in a dense LLZO (relative density > 97%) [189,192], and further the formation of dendrites has been related to low lithium-ion diffusivity at grain boundaries [193,194].
As discussed above, increasing the ionic conductivity of grain boundaries does not increase dendrite suppression capability [196,197]. Dendrite growth was also thought to be primary reason of pre-existing defects in surface and bulk SEs, such as cracks, since dendrite growth is favored in defective or cracked areas. The stress that induces dendrites can further extend the crack, promoting Li dendrite propagation [198,199]. However, this theory lacks of evidence of the mechanisms by which the dendrites form at a similar current density in LLZO with different surface roughness and why the dendrites still form in single-crystalline LLZO [198]. Dendrite formation has also been attributed to inhomogeneous Li plating caused by inadequate interfacial interaction between Li and SEs [189]. This can be explained by the fact that the critical current density increases in an inverse manner with the decrease in interfacial resistance between Li and SEs [178]. Even if the ASR of the Li/electrolyte interface has been optimized to be comparable to or even lower than in the liquid electrolyte, dendrites still grow at a low current density (0.9-1 mA·cm −2 ) [200]. To date, there is a gap on our understanding on this process, and thus, there is no clear road to prevent lithium dendrite formation in LLZO at the current density used (10 mA·cm −2 ) required for the fast-charge goal, which may prevent their effective integration with Li anodes [201]. However, high mechanical strength can inhibit dendritic Li growth [159]. According to computational analyses, garnet SSEs should have enough mechanical strength to prevent Li dendrite growth [197,202]. At high current densities, however, Li dendrites can develop into single crystal garnet SSEs, according to a recent experimental report [193,199]. Despite that, experimental evidence suggests that Li dendrites can grow into single crystal garnet SSEs at high current densities [198]. Good contact between electrodes and SSEs can also enhance the cycling stability, due to an improved interface and uniformly distributed Li-ion transport. An alloy interface layer, on the other hand, can provide a secure interface during cycling as well [198][199][200].
Some strategies have been developed to improve the SEIs, and to inhibit the formation of lithium dendrites between the Li metal anode and garnet-type SE. For example, designing 3D ion-conductive frameworks (Fig. 16(c)) have been proposed to effectively suppress the growth of Li dendrites and avoid the occurrence of cell short-circuit [195]. In this model, 3D solid electrolytes that have plenty pores and voids allow Li to expand smoothly through the structure host without causing Li dendrites formation [195]. Furthermore, using electrolyte additives, external pressure, highly concentrated electrolyte, composite materials, surface coating, polishing layers, and unlimited Li sources, are other alternatives strategies to overcome these issues [199][200][201]. Hence, research needs to be directed to characterize novel compounds' structures at the atomic level and to utilize in situ modeling techniques to understand the mechanism of Li anode-solid electrolytes and improve the stability among them.
(2) Electrochemical stability of LLZO With both the Li anode and the cathode materials, the LLZO SEs are among the most stable SEs [6,9,21,22,29]. The wide electrochemical window of garnet SEs (> 6 V vs. Li/Li + from CV studies and 3 V from computational analyses) allows for high voltage batteries [203]. Stable battery efficiency also involves electrochemical stability [4,5,79,204]. The current density, mechanical properties, interfaces, and ionic and electrical conductivities all play a role in electrochemical stability [193]. In addition to theoretical calculations for understanding lithium-ion conductivity in garnet electrolytes, the intrinsic electrochemical stability of Li 7 La 3 Zr 2 O 12 can be calculated in situ by first-principals analyses (Fig. 17) [204]. A semi-blocking electrode is applied to test the electrochemical stability window by using a conventional method [17,204]. In this case, the c-LLZO electrolyte presented a wide electrochemical window that ranged between 0 and 6 V [205]. Despite this, the value derived from first-principals analyses was significantly lower than the experimental results. For instance, at 2.91 V, the LLZO electrolyte was found to oxidize to La 2 O 3 , Li 6 [134]. Additionally, Zr is produced by reducing Zr 3 O at a much lower potential. The LLZO electrolyte was not thermally stable, according the first-principals analyses. Nonetheless, deposition of lithium ions interfered with LLZO reduction at low potential, making the reduction of LLZO difficult [204]. The reduction of Zr 3 O to Zr occurred thermodynamically, without taking into account the effect of DFT measurement bias [204,205]. With the same calculation method, the electrochemical stability of LLZO doped with Al, Ta, and Nb was also investigated [112,139,176,196]. The analyses showed that elemental doping had no effect on electrochemical stability, and the small amount of element doping had a negligible effect [17,134]. The electrochemical stability of SSEs and Li metal anodes has been reported to be improved by a Li-Al or Li-Mg alloy interface [70]. Because of the lack of driving force, Li can deposit on the alloy layer and drive it away from the interface rather than moving through the metallically conductive alloy interface layer [159]. Better understanding of the electrochemical stability of the LLZO solid electrolytes with electrode materials is of great importance, which can provide guidance to avoid forming unwanted compounds and decrease the interfacial resistance and help to optimize the solid-state interphase at the battery performance.

Interfacial issues between LLZO/cathode
There are a large variety of materials used for the cathode electrodes. Figure 18 shows the reaction energy of LLMO (M = Zr, Ta), the most used cathode materials, to study the thermodynamic stability of the cathode-LLZO interface [141], being the most common LiFePO 4 (LFP) and LiCoO 2 (LCO). LiFePO 4 has a long-term life cycle and high-power capacity in traditional lithium-ion batteries [206]. On the other hand, theoretical calculations indicate that LCO|LLZO has the most stable interphase due to a small driving force for LLZO decomposition in the charging state. LiMnO 2 (LMO) and LFP react very strongly with LLZO-based electrolytes [141,207]. X-ray diffraction (XRD), Raman spectroscopy, and X-ray photoelectron spectroscopy (XPS), have been used to study the chemical compatibilities with LLZO [208]. These studies showed LMO and LFP can react with LLZTO at temperatures as low as 500 ℃ and the reaction products between LLZTO and NCM are complex. LCO have received a lot of attention, due to its high electrochemical and chemical stability against a Li-garnet electrolyte, as well as its compatibility with LLZO SE [207,209]. Understanding the mechanism and stability of LLZO SE against mixing and sintering with major commercial oxide cathodes may help to avoid unwanted compounds, lower the interfacial resistance, and improve the composition of coating layers [141,[207][208][209]. First-principals studies based on thermodynamic DFT estimated the oxidation potential of LLZO to be 2.9 V, with an equilibrium potential of 3.3 V, which corresponds to totally de-lithiated LLZO [204]. This oxidation potential refers to the oxygen evolution reaction, which is considered to be kinetically difficult, and is predicted to have a high overpotential, which may be the cause of LLZO garnet's strong oxidation stability [204].
After all, the high ARS between LLZO and cathode materials represents a challenge that needs to be solved [212]. Undesired reactions occurred during battery usage that impaired their performance [211,212]. For instance, the formation of a tetragonal LLZO phase at the LCO|LLZO interface, due to the cross-diffusions of elements at high temperature, decreased the initial CE and life cycle [209]. The secondary phase on LLZO forms in a humid atmosphere, leading to form high ASR and adversely affecting the capacity at RT. Therefore, the reaction of LLZO with the cathode, CO 2 /H 2 O, and diffusion layers with cathode-LLZO interface, reduce the electrochemical performance of ASSLBs [214]. In addition, high interfacial resistance is due to the formation of the Li 2 CO 3 layer and weak physical contact [82].
To mitigate interface concerns, several approaches have been developed, where employing coating layers on the interface contributes to lowering the resistance and enhancing the lithium mobility at the interface [208][209][210][211][212][213][214][215]. Pulsed laser deposition technology (PLD) has been used to coating the cathode-electrolyte interface to address interface concerns [166,167,184]. A study used PLD to deposited LiCoO 2 and assembled the LiCoO 2 |Li 6.75 La 3 Zr 1.75 Nb 0.25 O 12 ASSLB, and determined the electrochemical performance and ASR [166]. Under these conditions, the ASR between LCO and LLZO-Nb was 170 Ω·cm 2 at RT [166]. The battery showed a good charge-discharge behavior and a stable cycle efficiency, as well as a theoretical electrochemical capacity for LiCoO 2 of 137 mAh·g -1 , for 0.5 Li per CoO 2 [166].
Studies have also focused on the usage of mixtures of different cathode materials to enhance the interfacial stability with LLZO and electrochemical and chemical properties of combinations like LCO/LLZO and LCO-LiBO 3 /LLZO cathode interfaces [210,214]. For the LCO/LLZO cathode interface, an irreversible electrochemical reaction at 3.0-3.8 V vs. Li + /Li was determined; this was detrimental to reversible capacity retention of the ASSLB [210]. Apart from that, during the high temperature cathode formation, time-of-flight secondary ion mass spectroscopy (TOF-SIMS) and XRD confirmed c-LLZO reacts with LCO and forms the t-LLZO phase at the interface [209]. For the LCO-LiBO 3 /LLZO cathode interface, it was determined that the LiBO 3 blocks undesirable chemical reactions at high temperature and the charge capacity was 79.9 mAh·g -1 , and discharge capacity was 67.2 mAh·g -1 with a CE of 84.1% [209], making this combination the most efficient and with higher capacity of all the samples tested [209]. Wet coating and hot-pressing methods were used to develop a simple low-cost method to construct ASSLB (Fig. 19(a)) [162]. The LFP cathode layer was uniformly coated on the Al foil; Figs. 19(b) and 19(c) indicate the not so obvious interface between the cathode layer and the electrolyte [162]. The interfacial resistance of the LFP/LLZO was reduced from 248 to 62 Ω·cm 2 [162].
Screen-printing is another method for enhancing the cathode interface [210]. An ASSLB was assembled with Li 3 BO 3 (LBO) lithium-ion conductor as a buffer layer between LCO active cathode material and Nb-doped Li 7 La 3 Zr 2 O 12 (LLZO-Nb) solid electrolyte [210]. Sufficient contact between the cathode layer and LLZO-Nb solid electrolyte was easily achieved by sintering LBO into the cathode layer by the annealing process (700 ℃) [210]. XRD and SEM results showed a good chemical stability and interfacial contact between the cathode and solid electrolyte materials and the ASR of the battery was 80 Ω·cm 2 at RT [210]. By using V 2 O 5 as cathode material, a high temperature ASSLB was assembled with Li 7 La 2.75 CaZr 1.75 Nb 0.25 O 12 (LLCZNO) particles in a reaction carried out at 100 ℃ [211]. The resulting compound had reliable safety parameters and stable cycling performance [211]. A thermal annealing technique to treat the cathode and garnet interface was also employed to ensure no parasitical reactions associated. In this case, the ASR was significantly decreased, from 2.5 × 10 4 to 71 Ω·cm 2 at RT, and the battery presented an interfacial resistance of 45 Ω·cm 2 with a CE of 97% and a stable discharge capacity at 100 ℃ [211].
The compatibility of Li 6.75 La 3 Zr 1.75 Ta 0.25 O 12 (Ta-LLZO) with LCO and Li(NiCoMn) 1/3 O 2 (NCM) commercial cathode materials was investigated [212]. First-principles DFT analyses indicated that the NCM/Ta-LLZO interface is more stable, with poor cycling stability which may be due to the formation of LaNiO 3 [212]. Figure 20 summarizes of the process for ball-milling and sintering treatments. In this case, Mn is the most stable element in NCM, while Ni is an active element and exchange with both Li and La at the interface  A brief schematic representation for the formation process and the structure of the LLZO surface layer on large NCM particles during the ball milling and co-sintering processes. Reproduced with permission from Ref. [212], © American Chemical Society 2018. [212]. Increasing Ni-Li/Ni-La may improve the interfacial stability between cathode and LLZO SE at high temperatures [212]. A novel mixture of cathode material composed by LiNi 0.7 Co 0.15 Mn 0.15 O 2 and LFP was developed [213]. This cathode material exhibited oxidation and reduction peaks at 3.58 V (cathodic peak) and 3.38 V (anodic peak), which correspond to the two-phase transition of the Fe 2+ /Fe 3+ redox couple [213]. This work proved that mixing cathode materials could improve the electrochemical properties, higher interface contact between cathode/LLZO, and better battery performance [213].
Coating materials are one promising solution to improve the cathode/LLZO interface, under different processing conditions [208][209][210][211]. LiNbO 3 and LiTaO 3 are recommended ternary metal oxides for mixing with LLZO, due to the improved the oxidation stability of LLZO, as these oxides do not have high oxidation limit [214]. LiTaO 3 /NCM and LiNbO 3 /LCO interfaces have no driving force to interact with the oxide cathodes in both fully lithiated and half lithiated stages [214]. Thus, to improve the interface contact between LLZO and the positive electrode, a composite positive electrode material with a solid PEO electrolyte and assembled LiCoO 2 -PEO|Ta-LLZO|Li battery was developed [155]. In this case, the ASR was large (6200 Ω·cm 2 ), but the battery can only run 10 cycles at 60 ℃ and 0.2 C [155].
An alternative ceramic processing strategy to assemble an oxide-based cathode composite based on LFP and LCO was developed (red square in Fig. 21). Here, the LCO-LLZO and LFP-LLZO were prepared through direct synthesis form metal salts to the oxide cathode in a porous LLZO scaffold. This allowed a good mechanical contact, with no adverse reactions at the interphase, and at low temperature 700 ℃ for synthesis [215]. The LCO-LLZO composite cathode showed a promising discharge capacity of 118 mAh·g −1 (3-4.05 V), with low interfacial resistance of 62 Ω·cm 2 [215]. This preparation method provides a solution to address the adverse interphase reactions based on chemistry and ceramic thermal processing records, as well as increasing the number of reaction sites for enhanced the performance of composite cathodes for LLZO ASSLBs.

2 LLZO/polymer composite solid electrolyte and its all-solid-state lithium batteries
Composite solid electrolytes (CSEs) are mixtures of inorganic fillers and polymer electrolytes [36,148]. Latest research demonstrates that CSEs can effectively minimize the crystallinity of polymer, thereby increasing the ionic conductivity of the final CSEs [158]. These have exceptional properties that could enhance interface problems between electrode-electrolyte, ion transport process across the interfaces, simple architecture on ASSLBs solving the wetting problems, and improve CE [71,216,217]. The most studied solid polymer electrolyte is PEO because of its low cost, ability to dissolve a wide variety of lithium salts, and relatively high chemical/electrochemical stability [36,71]. However, the PEO has a low ionic conductivity (10 -8 -10 -6 S·cm -2 ) at RT [218] and its mechanical properties at high temperatures result in lithium dendrites and safety hazards [219,220]. PEO/LLZO composite electrolytes melt when heated to high temperatures (near to PEO Fig. 21 Overview of all-oxide Li-garnet SSBs based on LiFePO 4 -and LiCoO 2 -based cathode composite design and ceramic processing options: processing temperature vs. chemical and mechanical stability. Reproduced with permission from Ref. [215], © The Royal Society of Chemistry 2020. www.springer.com/journal/40145 melting point), enhancing the ion transport [168,228]. Ion transport in the PEO matrix can be a slow process at low temperatures; here the formation and breaking of lithium-oxygen (Li-O) bonds, which involves the ether oxygen atoms on a segmental PEO chain coordinated with Li ions, leads to a rapid conductivity decrease [218]. Though PEO solidifies at low temperatures, the chain's segmental motion becomes slow; at the same time, the solid PEO blocks Li ions from moving between LLZO particles [158,162]. However, the integration of LLZO particles or fillers into polymer improves the ionic conductivity and also the mechanical strength and stability of the SE [36,155]. Combining experimental with computational analyses, researchers have investigated the Li transport properties of PEO. For the case of Ga-LLZO CSE, the improved ionic conductivity can be ascribed to the ionic conduction in the space charge regions and the percolation of the space charge regions. While the Ga-LLZO content exceeds the percolation threshold value, a continuous pathway is formed in charged regions that behaves as the fast channel for Li + ion transportation, as illustrated in Fig. 22. The PEO:Ga-LLZO composite shows great potential in ASSLBs with an electrochemical window 4.6 V [228]. This combination is stable with Li-anode, with a discharge capacity of 145 mAh·g -1 at 0.1 C [228]. To further develop LLZO CSEs with enhanced properties, other types of polymers, such as poly(propylene carbonate) (PPC) [163,219,221,222], poly(ethylene carbonate) (PEC) [223], polyacrylonitrile (PAN) [70], poly(methyl methacrylate) (PMMA) [224], cross-linked polyethylene glycol [225], poly(ethylene glycol) diacrylate (PEGDA) [226], and mixed polymers [218,227] have been used as substrates for constructing novel and more efficient LLZO-based/ polymer CSEs. Our group fabricated a CSE of LLZO/PEO, by mixing different amounts of LLZO ceramic powders into the PEO matrix [158]. We found that the addition of LLZO can effectively inhibit PEO's crystallization [228]. A correlation between an increase in the crystallinity content and the increase in LLZO content was determined [229], consistent with the fact that excessive LLZO particles are difficult to disperse in PEO matrix and easily form clusters [228,229]. Follow up studies on CSE composed by LLZTO, PEO-LiTFSI, and SN demonstrated that the CSE 60%LLZTO-10%SN exhibits a maximum conductivity of 1.22×10 -4 S·cm -1 [158]. The electrochemical stability of this compound ranged from 4.7 to 5.5 V vs. Li/Li + and a transfer number of 0.410 [158]. The LFP|LLZO/PEO-SN|Li cell shown in Fig. 23 exhibited the maximum discharge capacity of 151.1 mAh·g -1 after 200 cycles under 60 ℃ [158].
Another CSE (PEO-0.5LLZNO) was prepared by incorporating 33.3 wt% of LLZNO (LLZNO = Li 6.4 La 3 Zr 1.4 Nb 0.6 O 12 ) powder into the PEO matrix [230]. This CSE exhibited a high ionic conductivity of 1.4 × 10 -3 S·cm -1 and a wide electrochemical window range of 0-5.2 V at 60 ℃ [230]. Also, an LFP|PEO-0.5LLZNO|Li cell was able to deliver a high specific discharge capacity of 153 mAh·g -1 and excellent capacity retention of 97% after 150 cycles at 0.5 C under 60 ℃ [230]. Another example is the PEO-LiBOB-LLZO composite electrolyte prepared by Guo and collaborators [231], which presented an ionic conductivity of 2.4 × 10 -4 S·cm -1 at 20 ℃ . The assembled LFP|LLZO-LiBOB|Li battery was cycled at 0.1 C for 100 cycles at RT [231] and rendered a rate for capacity retention of 84.6% [231]. Karthik and Murugan [232] prepared a CSE (GCPEM-20) by solutioncasting method, and consisted of a PEO with a large molecular weight (~5×10 6 ), LCO, and Li 6.28 Al 0.24 La 3 Zr 2 O 12 (Al-LLZO) in a mass ratio of 8:1:2. The fabricated CSE All these results showed CSE fabricated with LLZO and PEO can deliver good ionic conductivity, suitable Li + transfer number, and compatible electrode/electrolyte interfacial resistance. Even though ASSLBs show acceptable performance at higher temperatures, polymer solid electrolyte achieves a low ionic conductivity at RT, inhibiting good performance for the battery at ambient conditions. A three-dimensional LLZO framework was developed by filling polytetrafluoroethylene (PTFE) as a plasticizer and Ta doped LLZO; this hybrid LLZTO solid electrolyte composite showed a 1.2 × 10 -4 S·cm -1 of ionic conductivity at 25 ℃ [233]. The assembled LFP|LLZTO-SN|Li handled a discharge capacity of 153 mAh·g -1 , at 0.2 C in the first cycle, and the capacity retention rate is high 90.3% after 200 cycles [233].
Particle size, morphology, and amount of LLZO solid electrolytes are key elements in the system of the cell. These can optimize ionic conductivity and solve interfacial issues, enhancing the performance of the ASSLBs [82,86]. Regarding particle size, a study of the effect of Li 6.4 La 3 Zr 1.4 Ta 0.6 O 12 (LLZTO) where different particle sizes (~40 nm, ~400 nm, and ~10 μm) were incorporated into the PEO matrix to form the LLZTO/PEO CSEs in the absence of lithium salts showed that the ionic conductivity of the CSEs inversely correlated the particle size of LLZTO [234]. The highest ionic conductivity was 2.1×10 -4 S·cm -1 at 30 ℃ and ~40 nm LLZTO. Due to the high ionic conductivity, two batteries were assembled LFP|LLZTO/ PEO|Li and LiFe 0.15 Mn 0.85 PO 4 |LLZTO/PEO|Li [234]. Both displayed superior electrochemical performance, with the specific capacities of 153.3 and 132.1 mAh·g -1 at 0.05 C, respectively, and capacity retention of 90% after 200 cycles at 0.1 C under 60 ℃ [234]. The authors proposed that a small particle size of LLZTO favors to achieve a relatively larger specific surface area, which improves the Li pathways and facilitates percolation, thus showing a wide electrochemical window up to 4.7 V [234]. However, instability problems occurred when operating for prolonged period of time at high voltage [234]. Therefore, a novel type of CSE composed of silane-modified Li 6.28 La 3 Al 0.24 Zr 2 O 12 (s@LLAZO) nanofibers and poly-ethylene glycol diacrylate (PEGDA) was developed [226]. This combination presented excellent cycling stability and wextraordinary high rate capability with LFO and Li[Ni 1/3 Mn 1/3 Co 1/3 ]O 2 cathodes (Fig. 24) [226]. Additionally, with Li-ion batteries, LLZO solid electrolytes are remarkable applications in otherwise rechargeable lithium batteries, such as lithium-sulfur batteries (Li-S), oxygen battery (Li-O 2 ), and Li-V 2 O 5 (Li-V) battery [40,86]. A solid-state Li-S battery was reported with LLZO-PEO electrolytes and sulfur composite cathodes based on LLZO@C, which operates successfully at 37 ℃ with a capacity of > 900 mAh·g -1 at 0.5 mA·cm -2 for 80 cycles [235]. Tao et al. [236] used a CSE containing PEO as the host and Al 3+ /Nb 5+ co-doped cubic Li 7 La 3 Zr 2 O 12 (15 wt%) as the filler and a modified sulfur cathode to assemble a Li-S battery successfully. Inside the assemble Li-S battery, the sulfur-based cathode is formed by using LLZO@C matrix and PEO binders, which successfully reduce the interfacial resistance between the sulfur and the ion/ electron conductive matrix [236]. Thus, taking advantage of the composite cathode and CSE, this Li-S battery can deliver good specific capacities of more than 900, 1210, and 1556 mAh·g -1 at 37, 50, and 70 ℃, respectively, and capacity retention of 98.7% after 90 cycles.

Conclusions and future outlooks
Garnet-type solid electrolytes are focus of intensive research and interesting due to their high ionic conductivity, wide electrochemical window, and chemical stability against Li ions. These are one of the most promising solid electrolyte materials to be used in the future ASSLBs. ASSLBs using LLZO-based solid electrolyte are expected to solve the challenges presented in terms of energy density and safety of power battery. However, after years of development, the ionic conductivity of LLZO at RT is still lower than liquid electrolytes. Though ASSLBs have many advantages during research and development progress, there are still two main problems: the low performance of ionic conductivity of solid electrolyte at RT, and the interface contact with the positive and negative electrodes. Besides, there is a significant gap in our understanding of the mechanisms of Li-ion transport and related electrochemical and chemical reactions, as well as structural and morphological properties, and thermodynamic behavior at ASSLB interfaces. Thus, research must be broadened by innovative in-situ characterization techniques and theoretical calculations. Despite LLZO/polymer CSE composite solid electrolyte, SCEs also have attracted attention due to their novel properties, such as suitable ionic conductivity and good thermal and electrochemical stability, as well as enhanced interface performance and mechanical properties. However, there are still many limitations to be solved in the development of suitable solid electrolytes for the future. The processing routes and synthesis techniques are not scalable and mass manufacturable for the large-production of such CSE and composite cathodes, representing a limitation in production processes. Until now, the cost of a solidstate lithium battery is still significantly greater than the cost of a liquid lithium battery. LLZO solid electrolyte ASSLBS also faces many material and technology issues, such as expansion during charge/ discharge and the relation between mechanical properties of the solid electrolyte and its performance. The problems of manufacturing technologies have to be addressed in commercial development, design, and optimization of garnet-LLZO solid electrolytes.