Highly textured and crystalline materials for rechargeable Li‐ion batteries

To build an environment‐friendly energy‐based society, it is important to develop stable and high‐performance batteries as an energy storage system. However, there are still unresolved challenges associated with safety issues, slow kinetics, and lifetime. To overcome these problems, it is essential to understand the battery systems including cathode, electrolyte, and anode. Using a well‐controlled material system such as epitaxial films, textured films, and single crystals can be a powerful strategy to investigate the relationship between microstructural and electrochemical properties. In this review, we discuss the need for research with well‐controlled materials system and recent progress in the well‐controlled cathode, solid‐state‐electrolyte, and anode materials for Li‐ion batteries. Enhanced stability and electrochemical performance due to the facilitation of prolonged and endured Li‐ion transport in facet‐controlled battery materials are highlighted. Finally, the challenges and future directions utilizing the well‐controlled battery system for high‐performance battery are proposed.


| INTRODUCTION
As the energy crisis deepens, global trends toward decarbonization are attempting to shift primary energy sources away from traditional fossil fuels to renewable and environmentally friendly alternatives. 1 Efficient energy storage technologies allow intermittently produced green energy to transcend time and space restrictions. Electrochemical energy storage, such as rechargeable batteries, is the most practical and effective option for a wide range of small and large-scale storage applications. 2 Lithium-ion batteries (LIBs) have been a great pioneer in energy storage since being introduced to the market in 1991, and have continued to advance over recent decades. 3 Today, LIBs not only lead the small-size battery industry for portable electronic devices, but are also prominent in electric vehicles and stationary energy storage technologies. 4 Yet, traditional LIBs cannot satisfy the stringent requirements of future batteries, such as energy density, safety, and cost standards, where the ever-increasing diversity of applications necessitates greater electrochemical energy storage. Because of the electrolyte system, which utilizes extremely flammable organic liquid electrolytes or polymer electrolytes, typical LIBs may pose significant safety issues. 5 When batteries are overcharged and discharged, the poor thermal stability and low flash point of liquid electrolytes can easily result in serious accidents such as fire and explosion. 6 Furthermore, liquid electrolyte systems exacerbate performance degradation and electrolyte-electrode side reactions. The unstable electrolyte is reductively decomposed to form a solid electrolyte interface (SEI), which generates massive numbers of inactive dead Li. 7 The SEI layer becomes out of control and eventually breaks down, substantially degrading Coulombic efficiency and producing high cell impedance. 8 Meanwhile, the narrow electrochemical window of liquid electrolyte, which is incompatible with lithium metal anodes, easily promotes the formation of lithium dendrites, which lead to short-circuit risk and capacity loss in batteries. 9 Therefore, the standard liquid electrolyte is no longer suitable for the high-power battery of the future due to its limited operating temperature, breakdown under high voltage, and failure to inhibit lithium dendrite.
One next-generation battery strategy focuses on "allsolid-state batteries" (ASSBs), which have generated great promise in terms of operational safety, variable cell design, and high-energy density. 10 Recently, there has been a surge of interest in building a performancecompetitive ASSBs due to the discovery of novel solid electrolytes (SEs) with available conductivity at ambient condition. 11 The substitution of liquid electrolytes to SEs opens up the prospect of designing advanced battery chemistries in addition to solving the major challenges with liquid electrolytes. 12 SEs can enhance safety since they are inflammable, heat-resistant, and do not cause leakages or gas production within the cell. In parallel, SEs not only have remarkable potential to increase energy density, but are also less likely to cause adverse effects than liquid ones, which contribute to extending the life expectancy of ASSBs. 13 Given the tendency of portable electronics to become increasingly downsized and consume higher power, solid-state batteries will play a key role in future energy storage systems.
In contrast to semi-infinite diffusion in the liquid electric double layer, mass transfer across the entire ASSB exclusively occurs via the vacancy-diffusion process through the solid space charge layer, in which kinetics is sluggish. [14][15][16] To further shorten the migration pathway and lower the interfacial barrier, mass-charge diffusion channels should be aligned parallel to the electrical field vector. 17,18 Consequently, the specific crystallographic arrangement of each ASSB component becomes increasingly crucial for efficient vacancy hopping of lithium ions. In recent years, interfacial design and crystal engineering that control lattice orientation have attracted attention as a key strategy for improving the performance of ASSBs. [19][20][21][22][23][24][25][26] Building the textured, epitaxial, and single crystalline systems is able to refine and disclose the interfacial behaviors of electrochemical mechanisms as well as to improve the durability and energy density of solid-state batteries. 27,28 Furthermore, significant adverse events at interface, one of the major obstacles to ASSBs, can be successfully suppressed by optimizing the orientation order of electrodes and SEs. 29,30 In this review, we discuss the single crystal-, textured-, epitaxial growth of ASSB components, cathodes, SEs, and anodes, with specific examples (Figure 1). First, the Li-ion kinetics in LIB and existing issues in conventional ASSBs are introduced. Second, recent research progress on facet-controlled cathodes, solidstate-electrolytes, and anodes is reviewed. At the end of this review, challenges and future research directions with a well-defined battery system are proposed. This provides insights into the importance of facet control to achieve further performance improvement in the near future as traditional methodologies of battery technology development reach their limits.

| THE MERITS OF CRYSTALLINE GROWTH CONTROL IN LIB
LIB is basically composed of a cathode, anode, electrolyte, and separator. The cathode transfers Li-ions to the anode through the electrolyte, and Li-ions are stacked on the anode through a charging process. In this charging process, the chemical potential of the anode is higher than that of the cathode. The Li-ions stacked in an anode return to the cathode through a discharging process, and electrons are moved through a circuit to generate a current. The separator keeps the cathode and anode materials apart. 36,37 Battery performance is mainly evaluated by cycle performance stability and specific capacity. To achieve high cycle performance stability and specific capacity, Liions should move quickly in the electrode and the electrolyte, and charge transfer by solvation/desolvation of Li-ions should occur at the interface between the electrode and the electrolyte. 38 If the difference between the diffusion rate of Li-ions at the interface and the charge transfer rate in the electrode is large, polarization occurs in the electrode due to the distribution of nonuniform ions and electrons, causing a plating phenomenon of Li-ions. To improve electron conductivity, it is important to form an electron transport path in the electrode. The main factor in determining electron conductivity is electron mobility, where charge transfer by the hole is generally considered as well. The equation related to the conductivity of the carrier's mobility is as follows: This equation represents conductivity by the mobility of electrons and holes. n i and p i are the numbers of electrons and holes, μ e and μ h are the mobility of electrons and holes, and e is the absolute value of each carrier's charge. It is important to select a material with high mobility based on the equation. When an electrode in a thin film state is used, electron conductivity may be easily decreased due to a thin thickness, so it is more important to ensure the mobility of the electrode. It is possible to improve the mobility of the electrode by coating heat treatment, conductive agent, and conductive material. [39][40][41] The diffusion rate of Li-ions is diffusion coefficient D i and flux per unit concentration change. The diffusion rate equation of Li follows the Arrhenius equation.
where D 0 is the experimental predictor, ΔG is the energy potential, and K B is the Boltzmann constant. Raising the temperature based on the formula can increase the diffusion rate but can cause irreversible capacity loss due to side reactions at the interface. Raising the temperature may not help improve the diffusion rate of Li-ions. The diffusion rate of Li is actually affected by the lattice of the material. The lattice difference of each material serves as a diffusion barrier, and it is preferable to construct a battery by aligning the composition and topology structure of the electrode. Li-ion diffusion rate within the electrode may be described below.
where x is diffusion distance, and q is dimensionally constant, and depending on the sample situation, the values are 2, 4 and 6, corresponding to 1D, 2D, and 3D. To reduce the diffusion time (t) and secure a high dimension constant, a microstructure that can cause 3D diffusion should be created. In general, by reducing the material size to a nano size, the contact area is improved, and the moving distance between electrons and Li is reduced. If the electrode structure is an open pore structure, a conductive interconnection network can be created in the 3D direction, which increases the reaction rate by creating a transport path.
To ensure a fast movement speed of electrons and a conversion speed of Li-ions, a material for a high Di value should be selected, a material size should be reduced, and a thin electrode should be configured to obtain a low x value. In addition, the high q value should be achieved by configuring the rational structure at the interface. 39 Figure 2 shows a schematic of issues for each component of the LIB. The ion mobility at the interface between the electrode and the electrolyte decreases due to the formation of improper interphase and resistive space charge layer. This causes high resistance internally, and dendrites of Li are formed at the interface between the electrode and the electrolyte or inside the electrolyte through plating and stripping of Li-ions during repeated cycle performance of the battery. 42,43 The dendrite is a protrusion of Li caused by an inhomogeneous electric field between the electrolyte and the molten Li interface. And it makes the ion flux nonuniform with a tip effect. Li  dendrite increases the resistance of the battery and causes extra strain energy due to the deformation of the electrolyte and the added interfacial energy. The driving force of the actual electrolyte is larger than the resistance force. A high electron conductivity moves from anode to electrolyte, thereby reducing the total electric potential applied to the electrolyte. As a result, during charging, the electrical potential of the electrolyte is lower than 0 V, which causes nucleation and growth of Li dendrite. 44,45 The volume change of active material forms cracks during a cycle and reduces the contact area at an effective electrolyte/electrode interface. 46,47 By reducing the specific resistance area at the interface, the charge transfer through the interface slows down and increases the electric potential applied to the electrolyte. An irreversible side reaction occurs when the electrolyte exceeds the electric potential that the electrolyte can withstand, and the resulting reaction interferes with the movement of Li-ions at the interface. The chemical potential between the electrode and the electrolyte limits the movement of Li-ions to create a space charge layer, which has a low Li-ions diffusion coefficient. 48,49 The high resistance of these interfaces leads to low coulombic efficiency, low power performance, and short cycling life. 36,50 Growing a thin film by controlling the crystal direction of the solid anode and the SE helps to study how mechanisms occur in atomic units during the EC process. Using transmission electron microscopy (TEM) and other neutron reflectivity equipment, it is possible to observe the movement of additional other atoms. 51 Physical/chemical analysis according to the crystallinity of each part of the battery will help understand the interfacial properties of LIBs and will be important base research to improve battery performance in the future.

| Cathode
As discussed in the previous part, research was conducted continuously to improve the cathode performance in a solid-state battery through the crystallization of cathodes. Compared to the conventional cathode, the cathode with high crystallinity exhibits excellent mechanical strength, structure or thermal stability, and cycling performance. 52 For a cathode without crystallinity, the Li-ions move to the electrolyte through grain boundaries (GBs) to increase the interface impedance between the electrolyte and the cathode. In contrast, a crystalline cathode has a continuous diffusion path of Li-ions and a lower interfacial resistance than a cathode with low crystallinity, thereby having a high diffusion momentum of Li-ions, and a cathode efficiency in a battery is higher than an amorphous state.
Suitable cathode materials should have low fermi levels and high potential energy. And it has a high reversibility to the movement of Li-ions so that Li-ions can be well intercalated/deintercalated. In addition, it should have a high ion diffusion rate, electron conductivity, and ion conductivity inside the cathode, and thermal stability and electrolyte compatibility should be good. Easy synthesis and low material prices are also F I G U R E 2 Schematic illustration of various issues in lithium-ion batteries (LIBs). required, and it is not easy to find cathode materials that satisfy all conditions at once. In general, lithiated transition metal oxide is mainly used, and the crystal structure of the cathode is divided into a layered-type structure, spinel-type structure, olivine-type structure, and silicate-type structure. Until now, the cathode material has been studied, focusing on the electrochemical performance of the cathode for the intercalation/ deintercalation of Li. 11 Studies on the performance analysis of LIB according to the crystallinity of the solid cathode have been actively conducted. The characteristics of the heterojunction interface between solids are determined by the crystallinity of each solid. The crystalline cathode has the following applicability to an industrial environment: high reaction homogeneity, small specific area, and high structural and thermal stability compared to a polycrystalline cathode. Improved battery performance may be determined by cycling performance. 53 Differences in the crystallinity of cathodes make differences in the physical and chemical properties of electrodes. An electrode in a single crystal state has superior mechanical strength, structure/thermal stability, and long cycling performance than in the conventional polycrystal structure. 54,55 When the same mechanical pressure (45 MPa) is applied to the cathode, the single-crystal cathode maintains its morphology better than the polycrystalline cathode. This means that morphological integrity is higher in single crystals than in polycrystalline so that structural stability can be better maintained during the calendar or charge/discharge process. 56 Under the 300cycle performance for measuring electrochemical stability, electrodes in polycrystalline structure, unlike single crystal structure, generated many cracks along the grain boundary. 57 The electrode in the single crystal cathode has high structural stability, and the specific surface area is maintained small during cycle performance, which generates less gas generated in the battery, thereby suppressing side reactions occurring at the interface. 53,56 3.1.1 | Layered structure (LCO, NC, CA, NCM, and NCA) The layered structure, expressed as LiMO 2 , is contained in the rock salt structure. MO 2 is located in the octahedral position and forms the MO 2 layer. The Li-ions are located between these solid layers. The disadvantages of a layered structure are that it actually has a capacity value of half lower than the theoretical capacity value, thermal stability is low, and the price of the cathode material is high. To compensate for this disadvantage, two or three elements are added to the transition metal, and representative materials include nickel, cobalt, and manganese. The ratio of each element is adjusted, the complementary point of the layered cathode is improved, and research is generally conducted in the direction of increasing the content of Ni and decreasing the content of the Co element. 58 Various experiments have been conducted on general trends. Mechanical strength, cycle stability, and thermal stability were measured for samples with adjusted ratios of Ni, Co, and Mn. An important factor in the problem of solidstate batteries is the negative interfacility kinetics at the interface between SE and electrode that interfere with charge/discharge. To analyze this, various variables (surface impurities, damage, space charge effect, interdiffusion) should be considered, and variables unrelated to crystallinity should be removed. [59][60][61] In recent years, crystallographic analysis has been conducted, excluding other factors for cathode materials. As shown in Figure 3, the crystallinity was controlled in the whole battery, or half battery was tested. Sayers et al. 62 confirmed that the 10-layer cubic perovskite superstructure Ba 1.7 Ca 2.4 Y 0.9 Fe 5 O 13 was epitaxially deposited, resulting in a lower grain boundary density than polycrystal, increasing ion conductivity. Figure 3A-C is a cross-section TEM image of Ba 1.7 Ca 2.4 Y 0.9 Fe 5 O 13 deposited with 10 layers, where stacking faults are generated by the SrTiO 3 (STO) substrate ( Figure 3C). In addition, the ion conductivity of each temperature was compared with the cathode form in the thin film and bulk state, and it was confirmed that the cathode in the thin film form manifested a higher ion conductivity ( Figure 3D,E). 62 In addition, a partial single crystal type LiMn 2 O 4 (LMO) cathode electrolyte was grown on an Li 0.33 La 0.56 TiO 3 SE part to measure cycle performance in each crystal direction of the cathode ( Figure 3F-H). 63 LiCoO 2 (LCO) is a cathode that has been studied steadily for many years, and there are many reports comparing battery performance according to the crystal direction by controlling the crystal plane of LCO. Experiments are continuously conducted to measure cycle performance by crystal direction through voltammogram curve analysis ( Figure 3I-L). 64 LCO also could be deposited epitaxially using PLD, making it possible to research the movement of Li-ions by the growth direction of cathode materials. Takeuchi et al. 65 deposited conductive SrRuO 3 (SRO) on the STO substrate to use it as a current collector to analyze electrochemical properties, and the STO substrate controlled the growth direction of LCO using substrates in the directions (111), (110), and (100), respectively. In each substrate, the LCO was deposited to the out-of-plane oriented (001), (110), and (104). When a surface oriented {104} plane between LCO and liquid electrolyte, the Li diffusion rate between the electrolyte and cathode was the fastest. 34,65 In addition, in a solidstate battery, the crystal direction of the cathode material affects battery performance more than in a liquid electrolyte. 66 The size of the battery will become smaller, and the demand for high energy density and repetitive stability will increase. Researching the performance of a solid-state battery according to crystallinity is essential because the interface between the electrolyte and the electrode will have a more dominant effect on the battery performance.
Lee et al. 67 adjusted the growth direction of LiNi 0.5 Mn 1.5 O 4 (LNMO) using the MgO (100), (110), and (111) substrate orientations and the LNMO in the (100) orientation has the highest surface energy. 68 As shown in the schematic diagram of each side of Figure 4A-C, the packing density is lowered in the order of (111), (100), and (110) orientation. The (111) oriented plane has the lowest surface energy because it has the highest surface density. The LNMO in the (100) direction has an exposed surface on a high proportion of (100) orientation, and the LNMO in the (100) orientation has the lowest stability compared to the crystals in which other sides are exposed. Accordingly, thick LiCO 3 is formed on the surface oriented (100) during the annealing process ( Figure 4D). In the galvanostatic charge-discharge (GCD) curves of the LNMO growth direction in Figure 4E-G, the initial charge curve has a larger capacity than the theoretical capacity value due to an irreversible oxidation reaction. 68 Figure 4H-J is a graph comparing GCD curves at the 1st, 50th, and 100th cycles per LNMO crystal direction. Capacities and coulombic efficiencies were compared, and until the initial 50 cycles, the LNMO in the (100) orientation had a higher capacity value compared to other directions. In Figure 4K, as the C-rate increases, there is a large decrease in the capacitor in the (111) direction. In 100  (100) and (110). (H) Stability test against cycle for each substrate orientation. Reproduced with permission. 63 Copyright 2022, American Chemical Society (I) θ-2θ measurement XRD pattern at ψ = 35°and (J) out-of-plane direction XRD pattern for thin film LCO on Pt (110) substrate. Volammograms for cycles of the LiCoO 2 /Pt (110) structure according to the orientation of LCO oriented (K) (11)(12)(13)(14)(15)(16)(17)(18)(19)(20) and (L) (10)(11)(12)(13)(14). Reproduced with permission. 64 Copyright 2016, Elsevier B.V. cycles, the largest decrease occurs at the cathode in the (100) orientation ( Figure 4L). The reason why the orientation (111) has the highest capacity stably in 100 cycles is that Li-ion is diffused by the exposed surface of the facet, and the orientation (111) is relatively densely packed than the orientation (001), making it difficult to diffuse. 69 In addition, cycling performance and thermal stability are important variables to actually commercialize a battery, and Fan et al. 70 tested LiNi 0.83 Co 0.11 Mn 0.06 O 2 to confirm that a high crystalline cathode has higher cycle stability than the amorphous cathode. At a specific temperature of the Li 1.2 Mn 0.567 Ni 0.167 Co 0.067 O 2 cathode material, thermal stability was measured in the amount of energy stored per gram, and a single crystal was measured to be higher than a noncrystalline cathode. 70 3.1.2 | Spinel structure (LiM 2 O 4 , M = V, Ti, Mn, Zn, and Ni) The LiM 2 O 4 formula indicates this, and the spinel structure corresponds to the Fd3m space group. In this structure, oxygen ions are located at the 32e position, which consists of cubic close packing. The transition metal, expressed as M, is located at six coordination sites in octahedral, and the Li-ions are located in tetrahedral vacant. Tetrahedrons and octahedrons provide 3D channels when Li-ions are diffused through co-planar and coedge. 33,37 The representative cathode material with the spinel structure is LiMn 2 O 4 . In the existing Li x CoO 2 , the x value is between 0.5 and 1, and the charge capacity was measured to be about 140 mAh/g. However, LiMn 2 O 4 has received a lot of attention as an alternative material for LCO cathodes because the cobalt used is expensive and toxic. The spinel structure LiMn 2 O 4 has a charge capacity (148 mAh/g) similar to that of LCO, is economical in material price, and is not harmful to the environment. 71, 72 Yu et al. 72 experimented with comparing the Li-ion storage of LiMn 2 O 4 in the single-crystal structure and the polycrystal structure. Generally, the single-crystal cathode has higher stability in crack formation and thus has superior cycle stability than the polycrystal structure. Yu et al. 72 compared and analyzed the specific capacity of the cathode according to the voltage range. In the 3 V range, single-crystal LiMn 2 O 4 has higher specific capacity and cycle performance stability than polycrystal LiMn 2 O 4 . However, unlike the Ni-rich layered cathode, Mn (Ⅱ) ion dissolution occurs in the LiMn 2 O 4 cathode at the 4 V region, resulting in a phase transition. Because the {110} side is exposed in the single crystal facet, Mn dispersion occurs more easily than in the densely packed polycrystal structure. It reduces the cycle and electrical stability of the single-crystal state cathode. As a result, the single crystal LiMn 2 O 4 has higher specific capacity and cycle performance stability than the cracking effect at polycrystal, depending on the voltage range. 72,73 There is also a study that analyzed the characteristics of cathode materials by crystal direction of LiMn 2 O 4 , and KC et al. 74 controlled the plane direction of LiMn 2 O 4 in the out-of-plane direction using substrates in the (100) and (111) directions of STO. The movement of active materials at the interface and the degree of secondary phase formation were observed by the crystal direction of LiMn 2 O 4 . In the case of LiMn 2 O 4 deposited in the direction (100), it was confirmed that the cycle performance of capacity decreased sharply, and this confirmed that the crystal direction of LiMn 2 O 4 was a factor affecting the performance of the cathode. 74 Kim et al. 75 manufactured a high-efficiency battery by forming a spinel structure LiMn 2 O 4 as a single crystal nanowire with a large surface part relative to volume. The nanorod form was a one-dimensional electron transfer path and strain relaxation while the battery was charged/discharged. Methods to synthesize LiMn 2 O 4 include combination, sol-gel, solution-phase, template, and PLD. The author used hydrothermal synthesis, which synthesized beta-MnO 3 nanorods into freestanding single-crystalline LiMn 2 O 4 through chemical changes. The single crystal nanorod cathode showed a higher capacity value than the charge storage of the commercialized 10 µm size particle and maintained 85% of the initial charge storage capacity during 100 cycles. As Li is intercalated, it causes a small lattice parameter change in the cubic phase, which causes a potential window of 3.5-4.3 V. Figure 5A-D shows the change in the lattice parameter of the spinel cubic structure due to the change in the composition ratio of Li by delithiation. With this change, the potential of the cathode material is divided into a two-phase domain state within 4.15 V. In Figure 5E, it can be divided into three stages of reaction for each voltage. The first step is to eliminate the oxygen vacancies remaining in the excess lithium and lattice up to 3.8 V, and the second step is to undergo a lithium de-intercalation reaction at voltages higher than 3.9 V. In the LiMn 2 O 4 cubic spinel phase, the amount of Li is changed to an insufficient phase in the same state. 76 In voltage near the 4.15 V plateau, two-phase mixings are present, and the reaction is completed by having a composition of Li 0.2 Mn 2 O 4 . The reaction terminals can be classified by the peak value shown in Figure 5F. Comparing the powder type used in the industrial field with the discharge value of the LiMn 2 O 4 type used in this paper, it can be confirmed that the single crystal nanorods type can transport one-dimensional electrons, and thus the kinetics of LiMn 2 O 4 is improved even at a high rate ( Figure 5G,H). 75

| Olivine (LiMPO 4 and Li 2 MSiO 4 )
The chemical formula for the olivine structure is LiMPO 4 , and the close-packed hexagonal structure is slightly distorted. Phosphorus ions are located in the tetrahedral vacant, and transition metal ions and Li-ions are located in the octahedral vacant. The framework of the crystal consists of MO 6 octahedrons and PO 4 tetrahedrons. Each PO 4 shares one point with MO 6 , and PO 4 is not interconnected. So far, LiFePO 4 is the most promising cathode material in the olivine structure and has already been commercialized in the LIB. In the case of LiFeO 4 , ions move slowly due to a compact structure but have high structural stability at a low voltage. This is because the volume change is within 6% when charged/discharged. However, since the FeO 6 octahedral network confines electrons between Fe-O-Fe, the oxygen-metal ion combination in hexagonal close packing reduces electron mobility. This drawback makes olivine batteries unsuitable for use in environments that require high rates. The main purpose to overcome this disadvantage is increasing the conductivity of LiFePO 4 by coating carbon or other conductive materials and reducing the particle size, doping ion, size reduction and morphology control. 31,77 Unlike the conventional method, the kinetics of the extraction/insertion process of Li-ions can be improved by controlling the orientation of LiCoPO 4 .
[010] oriented LiCoPO 4 nanoplates/nanoflakes structure and [100] or [001]-oriented nanorod and nanowire structures have a minimum shape of crystallite size in b-axis, which is the diffusion direction of Li-ions, and when Li-ions are certain in this crystal direction, the diffusion rate of Li-ions in the olivine structure can be increased. Here, the crystal direction of the cathode in the olivine structure is determined by the free surface energy of each facet in a thermodynamic equilibrium state. 78 The preference of the crystal facet is determined by the solvent properties, and the LiCoPO 4 may form a crystal direction in the direction (010) by using ethylene glycol as the solvent and solvothermal condition. [79][80][81] Wang et al. 82 made a high-rate LIB by adjusting the crystal direction of LiFePO 4 nanoplates. As mentioned above, olivine LiFePO 4 nanoplate is prepared using other precursors and other growth kinetics through a glycol-based solvothermal process, and the nanoplates may adjust the crystal direction with ac facet or bc facet. In Figure 6A,B, the difference in capacitance between the two directions was not significant at a low C-rate, but at a high rate, the difference was significant. As the C-rate increased, the reversible capacities of the ac-facet crystal orientation were higher than those of the bc-facet crystal orientation. The difference is that each crystal direction controls only rate capability, and the defect concentration controls the degree of chemical availability. Using this, only the rate capability of the battery can be adjusted without interfering with other elements. 82 In addition to the layered, spinel, and olivine structures mentioned above, there is a cathode with a silicate structure. Silicate with the orthorhombic structure is represented by the Li 2 MSiO 4 formula and has two Li-ions per metal to enable reversible de-intercalation. This is why the cathode material with the silicate structure is attracting attention. Tetrahedrons of SiO 4 and tetrahedrons of MoO 4 form a layered-like structure, and tetrahedrons of SiO 4 have a structure that is periodically repeated. 32,84 In 2021, Du et al. 83 measured charge (330 mAhg −1 ) and discharge (220 mAhg −1 ) values that were almost similar to theoretical values. In Figure 6C, the discharge capacity differs by 30 mAhg −1 depending on electrode loading for Li 2 CoSiO 4 cathode and Li 2 Co 1-x Mn x SiO 4 cathode. However, until now, few studies have analyzed the performance of the battery according to the crystalline state of the silicate cathode material or the characteristics of the interface according to the crystal direction. Due to the various applications of solid-state batteries, the demand for solid-state batteries will increase. As the battery is miniaturized and each part is solidified, battery performance measurement by the crystallinity of the cathode material is a part to be studied. 83 3.2 | Solid-state electrolyte (SE) SEs are one of the most vital components of ASSBs since the performance of solid-state batteries depends on the diffusion of ions inside the electrolytes. SEs necessitate not only high ionic conductivity but also extremely low electronic conductivity and substantial levels of chemical stability. Unlike cathode materials, few studies have focused on facet control in SEs; instead, the majority of research has concentrated to investigate novel compound and structure types that satisfy the requirements. 85,86 This is possibly because there has been far less research and development in the field of SEs since the period of attention on SEs is considerably shorter than that of cathode materials. 87 The SEs is a rapidly developing area that is still in the early stages of maturation. Fast ion conducting materials utilizing crystallographic advantages will be more actively developed and explored in the future.

| Garnet type-SEs
The traditional garnet structure has a standard chemical formula of A 3 B 2 (XO 4 ) 3 , where cation-filled A-B-X forms a face-centered cubic structure. The garnet structure possesses Li-ion conductivity when a Li-ion occupies the X position. 88 The crystal structure of cubic Li 7 La 3 Zr 2 O 12 (LLZO) is shown in Figure 7A. 89 Since the Li 7 La 3 Zr 2 O 12 stoichiometry contains seven Li atoms per garnet formula, it is called Li7 system. The fundamental framework of garnet structure is composed of dodecahedral LaO 8 and octahedral ZrO 6 . Within the interstices of the framework, the Li atoms occupy two different crystallographic positions, L1 and L2, with the Li2 site being vacant in the ideal case. Li-ion conduction is primarily governed by disorder and partial Li atom occupancy at Li2 sites. As illustrated in Figure 7B, the networks of tetrahedral and deformed octahedral, made of Li and Li vacancies, are interconnected across the entire framework. The three-dimensional pathway of Li-ion migration implies an isotropic nature of ionic conduction with little dependency on crystallographic orientation. 94 Total ionic conductivity therefore depends on the contribution of GBs, which has a significant effect on ion migration within the pellet. There will be a noticeable difference in ionic conductivity between single and polycrystalline pellets as the grain boundary impedance hinder the Li-ion transport in LLZO. 95  [110] direction in LLZO(001) and the  direction in LLZO (111). The insets illustrate the impedance spectra at different temperatures. Reproduced with permission. 93 Copyright 2013, Royal Society of Chemistry.
The influence of GBs on Li-ion transport inside the LLZO SEs was reported by Yu et al. 90 For three GB models with different crystallographic interfaces, the composition, energy, and transport characteristics were examined at the atomic level. Figure 7C schematically shows the three symmetric tilt GBs with coincident-site lattice classifications of Σ3 and Σ5. The unrelaxed structures are represented as a BCC Zr sublattice with Li, La, and O atoms omitted for convenience. The formula units of LLZO within the grains are 160 (Σ5(310)), 80 (Σ5(210)), and 96 (Σ3(112)), respectively. Every grain includes stoichiometric amounts of LLZO. The calculated Li-ion diffusivity across LLZO cells for the respective GB configurations is presented in Figure 7D. In all types of GB models, diffusivity is always higher in the bulk areas and reduces in the GB region. It was discovered that the loss in ionic diffusivity was caused by both a rapid rise in confined immobile Li-ions near GB region as well as a local degradation in ion mobility due to disruption of the interconnection of ion transport channel across GBs. Although diffusion inside bulk LLZO is isotropic, it may exhibit local anisotropy due to the unusual lattice geometry around the GBs. 96 As a result, the degree of diffusivity drop considerably differs amongst the distinct GBs. Figure 7E displays Arrhenius plots for Li-ion diffusivity inside the bulk LLZO and three GB models over a temperature range of 700-1100 K. The bulk LLZO has the lowest activation energy of 0.52 eV as extracted from the slope and has the greatest ionic diffusivity across the entire temperature range. The variation in ionic diffusivity is more pronounced at low temperatures, with the Σ5 boundary at 300 K having up to two orders of magnitude slower diffusivity than the bulk. Hayamizu and colleagues reached similar conclusions by comparing various garnet-type SEs in single crystal and powder form. 91 Arrhenius plots in Figure 7F give D Li for single crystal and powder samples of garnet-type SEs with various dopants. The evaluations included single crystal samples of sc-LLZO-Ta 97 and sc-LLZO-Nb 91 as well as powder samples of LLZO, 98 LLZO-Ta, 99 LLZO-Al-Ta, 100 and LLZO-Nb. 91 Over the entire measured temperature range, the single crystal samples clearly displayed larger D Li , regardless of the doping component. In garnet-type SEs with three-dimensional ion transportation, it is obvious that single-crystal materials with no GBs exhibit higher diffusivity without suffering from a mobility deterioration due to transboundary penetration. SEs demand extremely low electrical conductivity to prevent systems from self-discharging. The unintended production of Li dendrites may also be encouraged by a non-negligible electronic conductivity. Numerous external factors, including chemical inhomogeneity, impurity, and grain boundary, can determine macroscopic electronic conductivity. Therefore, high electronic conductivity values obtained in nonideal systems such as polycrystalline may originate from the large chemical inhomogeneities of the interfacial region at GBs. Wilkening's group analyzed the bulk electronic conductivity (σ eon ) and total conductivity (σ total ) for single crystal Li 6.4 Ga 0.2 La 3 Zr 2 O 12 (Ga-LLZO) with a variety of different polycrystalline garnet-type SEs. 92 The temperaturedependent electronic conductivity depicted in Figure 7G covered various single-and polycrystalline SEs with different dopants. The specific electronic conductivities have been demonstrated to vary by several orders of magnitude for polycrystalline materials even with the identical composition. The Ga-LLZO single crystal exhibits the lowest electronic conductivity in the given data measured at low voltage. While excluding any contributions from the GBs, it represents true bulk value. This supports the general observation that garnet-type SEs exhibit higher electronic conductivity in polycrystalline materials than in single crystalline. Figure 7H provides a summary of the total conductivities and activation energies for the samples displayed in Figure 7G at room temperature. Despite showing a low electronic conductivity, the single crystal Ga-LLZO has been demonstrated to have a rather high total conductivity, the sum of the ionic and electronic conductivities. Given the negligible contribution of electronic conductivity to the total, it can be inferred that the total conductivity is driven by ionic conduction. In other words, single-crystal Ga-LLZO is highly ionic conductive and ion transport is not coupled with electron transport. By minimizing extrinsic factors in the single crystal system, it was able to achieve both high ionic conductivity and low electronic conductivity at a level that does not lead to lithium dendrite formation.
Kim et al. 93 successfully synthesized Al-doped LLZO epitaxial thin films on Gd 3 Gd 5 O 12 (GGG) substrates by pulsed laser deposition and investigated the intragrain Li-ion diffusion mechanism. The epitaxial LLZO thin films are deposited in two directions depending on the orientation of the GGG substrate: LLZO(001)/GGG(001) and LLZO(111)/GGG (111). The schematic crystal structure in Figure 7I illustrates the epitaxial growth of Al-LLZO films in two orientations. Figure 7J shows the temperature-dependent conductivities of the epitaxial LLZO thin films obtained from 25°C to 80°C. The ionic conductivity was measured along the in-plane pathway in the [110] direction for the LLZO(001) and the  direction for the LLZO(111) thin film, respectively. The distinct semicircles in the impedance spectra depicted in the inset can be considered as the parallel coupling of a capacitance and a resistance. In comparison to the LLZO(001) film, the epitaxial LLZO (111) film displayed a higher ionic conductivity as well as a lower activation energy. Since Li-ions migrate isotropically across the three-dimensional channel of cubic LLZO, the disparity in ionic conductivity might result from compositional or structural differences between the films. The composition of the two films is comparable, while the (111) film was found to have less lattice distortion than the (001) film. Hence, the lattice mismatch between the substrate and film caused the lattice to distort, which led to a discrepancy in conductivities. The researchers demonstrated the practical effects of lattice distortions directly through an epitaxial film model system.

| Perovskite type-SEs
One representative ceramic SE is of the perovskite type which has ABO 3 crystals structure. 101 Lithium lanthanum titanate (LLTO), which has a double perovskite structure where La 3+ and Li + occupy the A site of the titanate compound (ATiO 3 ), 102 is a promising candidate for ASSBs applications due to fast ionic conductivity and good chemical stability. 103 Figure 8A displays the perspective views of Li 3x La 2/3−x TiO 3 (LLTO) crystal structure. 104 In ABO 3 perovskite structure, a portion of the A-sites is occupied by Li-ions, La-ions, and vacancies, while Ti atoms fill the B-sites coordinated by six O atoms in the TiO 6 octahedral configuration. Due to its deficient nature, La 2/3 TiO 3 contains many inherent defects, almost 1/3 of the vacancies, in the A site. At room temperature, these La vacancies are not randomly arranged; instead, they stack along the c-axis and are divided into La-rich and La-poor layers, forming a partially arranged superlattice structure. In the LLTO framework, Li-ions horizontally travel along the La-poor layer, which contains most of the vacancies and the Li-ions, via the ion-vacancy transition process. 108 During migration, Li-ions transmit through a quadrilateral window composed of four oxygen, and the concentration of transport window directly influences the mobility of Li-ions.
The impact of the transmission window area on the ionic conductivity in strain-controlled epitaxial Li 0.33 La 0.56 TiO 3 films was examined by Wei et al. 106 The schematic structures of epitaxial films with tensile and compressive strain are illustrated in Figure 8B. Depending on the lattice mismatch with the substrate, the LLTO layer is subjected to either compressive or tensile stress, which causes lattice distortion. At this time, the area of the transmission window extends under tensile strain and reduces under compressive strain. Figure 8C  have been shown to grow in different orientations because each substrate induces a respective lattice mismatch and corresponding strain. For STO substrates with a sub > c LLTO/2 , the applied tensile strain induced the ac-plane to deform less along the in-plane direction, contributing to the h00 orientation. On the other hand, for the LSAT and NGO substrates where a sub < a LLTO , the compressive strain imposed on the LLTO lattice allowed the ab-plane to have less strain, resulting in 00l orientation. Figure 8D displays the Arrhenius plot of compressed LLTO thin film on NGO substrate. The anisotropic in-plane ionic conduction across ac-and bcplanes was observed. The compressive strain of the orthogonal lattice is stronger along the a-axis than the baxis. Therefore, the transmission window in the ac-plane is narrower than that in the bc-plane, suggesting that greater compressive strains result in restricted ionic conduction requiring higher activation energy due to the deepened bottleneck. 109 Further, the dependence of ionic conductivity on domain orientation was explored by Kuwabara's group in a single crystal LLTO system. 105 In contrast to garnet-type LLZO, LLTO exhibits anisotropic behavior in ionic conduction with the presence of abplanes, the La-poor layer, serving as the primary pathway for Li-ions. 110 Domain orientation, hence, affects the macroscopic Li-ion conductivity. Figure 8E shows a bright-field scanning transmission electron microscopy (BF STEM) image of a single crystal LLTO. There are two types of lattice defects in LLTO crystal: 90°domain boundaries (DBs) and antiphase boundaries (APBs). At 90°DBs, each perovskite unit in one domain is joined to units in the neighboring domain by a La-rich layer. Lithium transmission over DBs requires higher activation energy, which significantly impacts ionic conductivity. APBs presumably exert a lesser impact on Li-ion conduction than 90°DB since each domain orientation aligns with the conduction direction and the separation in between La-poor layers is only c/2 unit cell. Figure 8F,G is the cross-sectional scanning electron microscopy (SEM) images and electron backscatter diffraction (EBSD) measurements of region A and region B, respectively, with different crystal orientations. The α, β, and γ regions of the three different domains were demarcated by trisegmented colors. Schematic diagrams of the crystal orientation models for regions α, β, and γ are depicted in Figure 8H. All combinations of three domians, α and β, β and γ, and γ and α, exhibit orientation relationships forming a 90°DBs. To figure out the relationship between crystal orientation and Li-ion conduction, it is necessary to consider the degree to which each domain has been rotated perpendicularly along the impedance measurement direction. When projected from the impedance measurement axis, the ab-plane is tilted by about 18°C in the γ region, whereas it is tilted by 45°in the α and β regions. This implies that region γ carries out more favorable Li-ion conduction than regions α and β. Area calculation revealed that region A contains 70% γ domain of the total, while region B has 61% γ domain. The difference in occupancy of region γ between regions A and B is expected to lead to a disparity in ionic conductivity. According to the Nyquist plots evaluated at 300 K in Figure 8I, regions A and B were estimated to be have ionic conductivities of 1.0 × 10 −3 and 0.8 × 10 −3 S/cm, respectively. Thus, even for single-crystal materials, it has been demonstrated that the ionic conductivity changes depending on the microcrystal orientation owing to the anisotropic conduction nature.
Lv et al. 107 introduced highly oriented 2D LLTO crystals and substantially increased ionic conductivity as well as enhanced battery performance. A thin laminar inorganic solid electrolyte (LISE) of perfectly aligned LLTO flakes was applied as an interlayer channel inside the laminar architecture between vermiculite (Vr) nanosheets. Based on the confinement effect that induces oriented growth in a cramped environment, LLTO crystals were synthesized in continuous and regular 2D arrangement without crystal defects. 111,112 Interestingly, the LLTO LISE preferentially grew in the c-axis, the fastest path for Li + migration, along the interlayer channel, affording considerably improved structural stability, ionic conductivity, and battery performances. The schematic diagrams of Li-ion transport channel in Vr-LLTO LISE and LLTO Pellet are seen in Figure 8J. In contrast to Vr-LLTO LISE, the LLTO pellets produced in open space exhibit a complex structure with evident crystal defects due to unrestricted growth in nonuniform directions. The TEM image in Figure 8K shows an ordered Vr-LLTO LISE with a layered 2D structure. Vr nanosheets and LLTO crystals are represented by the bright and dark regions, respectively. Importantly, structural defects and voids are not detected even at higher resolutions. Arrhenius plots in Figure 8L compared the conductivity of Vr-LLTO LISE and LLTO Pellet. It is speculated that the higher conductivity of Vr-LLTO LISE originates from the arranged orientations that suppress the inherent structural defects of the LLTO crystal, thereby eliminating the grain boundary impedance. Indeed, the calculated activation energies of 0.336 eV for Vr-LLTO LISE and 0.411 eV for LLTO Pellet approximate those of bulk and grain boundary, respectively. 113,114 In addition to the ionic conductivity, the cycling performances of Vr-LLTO LISE and LLTO Pellet were further evaluated in the LFP/Li batteries. After accounting for the comprehensive consideration of the period for each charge/discharge operation and battery cycling life, the cycling performance was carried out with 0.5 C at 60°C. As demonstrated in Figure 8M, the battery made with LLTO pellets underwent a short circuit after 47 cycles due to its catastrophic structural flaws. On the other hand, the batteries assembled with Vr-LLTO LISE operated without breakdown for 150 cycles, although there was a slight variance in capacity reduction based on the thickness. Consequently, the exceptional performances of Vr-LLTO LISE suggest a strategic design for ultrafast Li + conduction devices with great potential for ASSB application.
So far, several types of research have been introduced that investigated the effect of facet control on ionic conductivity in two representative SEs, LLZO and LLTO. Although not actively researched yet, as the progress through composition engineering reaches its limit, it will become more important to enhance the performance of SEs by utilizing crystallographic properties such as single crystal, grain, and epitaxial growth.

| Anode
Anode has played a crucial role in battery performance. The factors governing the performance of anode depend on where it occurs. Charge transfer ability and the kinetics of Li-ions within SEI are usually more important at the surface of the anode. On the other hand, the diffusion of Li-ions such as interfaction/deintercalation, and alloying/dealloying are crucial to determine the battery performance in the volume. To design a safe, high-rate, and high-capacity anode, understanding the anode using a well-controlled material system is important. In this section, recent progresses for three different anode material systems including carbon-based materials, IVA and VA group-based materials, and transition metal compounds using facet control will be discussed.

| Carbon-based materials
Among the carbon-based materials, graphite is the first material used as an anode for LIBs and also was the first commercialized material. 115 Graphite has a layered structure with 0.335 nm of spacing between layers and charging/discharging in graphite-based anode occurs through the Li-ion intercalation/deintercalation. Although graphite has several advantages such as high capacity (372 mA·h · g −1 ), low cost, low overpotential, and cycle stability of graphite, there are some limitations that need to be overcome. Due to the small layer spacing, graphite has sluggish Li-ion intercalation kinetics and large Li-ion diffusion resistance. High current over 1 C induces lithium plating, which is the deposition of a dead Li layer on the anode surface, resulting in an increase in internal resistance and a decrease in battery capacity. More Li metal deposition can induce Li dendrite growth, which can cause safety issues due to short circuits and heat generation. To solve these problems, well-controlled graphite-based anodes have been studied.
As part of efforts, graphite anodes with alignment control have been explored. To control the orientation of graphite, magnetic field has been applied to the graphite that is functionalized by magnetic field responding materials. Billaud et al. 116 showed the electrochemical properties of a battery with an out-of-plane ordered anode. Graphite was coated with superparamagnetic iron oxide particles and the alignment of graphite coated with iron oxide was controlled by the magnetic field during the synthesis of anode. Battery with an aligned anode perpendicular to the current collector showed higher lithium storage capacity compared to the randomly oriented anode due to the shorter Li + diffusion pathway. Aligned anodes showed a lower overpotential and up to three times higher specific charge at higher rates. Zhang et al. 117 investigated the battery performance according to the alignment angle of the graphite anode. First, to change the orientation, graphite flakes were coated with ferrofluid of different concentrations and then an external magnetic field was applied. Figure 9A shows the cross-sectional SEM images for graphite with different orientation angles including 64°, 71°, 85°, 72°, and 66°, respectively. The alignment angles in SEM images show a volcano-like trend with increasing concentrations of ferrofluid. XRD patterns in Figure 9B exhibit that the (002) peak, which represents the orientation perpendicular to the hexagonal plane of graphite, first decreases and then increases. This implies that the graphite flakes functionalized by the ferrofluid with a concentration of 37.5 μL·g −1 are aligned almost vertically. This trend of alignment angles is shown in Figure 9C. The authors demonstrate that the different ferrofluid concentrations cause different magnetic susceptibility, resulting in different alignment angles. Figure 9D exhibits that the specific capacities increase as the alignment angle increases at a rate of 2C, indicating that the larger angle derives the shorter ion diffusion path. Furthermore, the aligned anode has high cycling stability at a high current rate. The authors also observed the effect of the rotating speed of the magnetic field on alignment at the same ferrofluid concentration of 37.5 μL·g −1 . Graphite anode prepared under a rotating speed of 350 r·min −1 shows the highest angle of alignment and highest specific capacity.
In addition to graphene, graphene oxide is another good carbon-based anode material. Zhang et al. 119 prepared an aligned Fe 3 O 4 nanoparticle-decorated graphene oxide anode using the ice-templating method. The synthesis method is as follows: (1) Dispersing the graphene oxide decorated with graphene oxides that are aligned. (6) Removing the ice crystal by reducing pressure. Pore size and wall thickness can be controlled by cooling rates and freezing temperatures. The rate performance of aligned anode (724 mA·h·g −1 ) is much higher than the anode synthesized via the slurry-casting method (162 mA·h·g −1 ) at 2 A·g −1 . The aligned anode also shows excellent areal capacity (3.6 mA·h·cm −2 under 10 mA·cm −2 ). The authors demonstrate that the high areal capacity is resulted from the improved transport kinetics, the presence of pores that buffer volume expansion, and decoration of conversion-based active material Fe 3 O 4 . Liu et al. 118 used pulse freezing method to prepare the vertically aligned graphene oxide films. Highly ordered graphene oxide is obtained through the following process: (1) Introducing the suspensions of graphene oxide into the two-step polydimethylsiloxane microfluidic channel with various flow rates. (2) Graphene oxide sheets are aligned in different directions depending on the channel region. (3) The graphene oxide suspension freezes rapidly ("Pulse Freezing") by immersing into a dry-ice ethanol bath. (4) Frozen graphene oxide film is pumped out of the channel after the interface part melts. (5) Transferring the extracted graphene oxide film back into a dry-ice ethanol bath to freeze again. Multiple freezing steps can maintain the vertically aligned structure of graphene oxide sheets and yield a porous structure. Figure 9E shows the schematic illustration of randomly oriented and vertically aligned graphene oxide films showing Li-ion transmission and mobility. Crosssectional SEM images in Figure 9F,G exhibits highly ordered graphene oxide sheets prepared at different flow rates of 0.2 and 0.4 mL/min, respectively. Graphene oxide sheets prepared using graphene oxide concentration of 9 mg·mL −1 and flow rate of 0.2 mL·min −1 have a horizontal orientation. On the other hand, graphene oxide sheets prepared using graphene oxide concentration of 9 mg·mL −1 and flow rate of 0.4 mL·min −1 have vertical orientation. Electrochemical and half-cell performance were investigated for pristine graphene oxide, aligned graphene oxide, freeze-dried graphene oxide, and graphite. Figure 9H shows that the half-cell with aligned graphene oxide sheets exhibits 190% higher specific capacity (440 mA·h·g −1 ) after 150 cycles compared to the half-cell with freeze-dried graphene oxide sheets.
Based on the studies mentioned above, vertical alignment of carbon-based anode can shorten the Li-ion transportation path. Although there are several reports showing that such a short Li-ion path can greatly improve the battery performance, there is a lack of studies showing the impact of the other factors such as doping on orientation controlled carbon-based anode. These kinds of studies can provide a good platform to systematically investigate the impact of strategies such as doping and changing the spacing.
Carbon-based anode materials are excellent anode candidates for LIBs due to their low cost, structural tunability, and widespread availability. However, even if various strategies (e.g., doping) are used, there is a limit to further improvement of performance due to theoretically low capacity (e.g., graphite) and poor initial columbic efficiency. To this end, it is necessary to develop other anode materials with excellent performance.

| VA and IVA-based materials
VA and IVA-based anode materials includes semimetallic elements such as Sn, Sb, and Bi, and metallic elements such as Si and Ge. These materials store the multiple Li-ions by forming the alloy (e.g., Li 4.4 Si, Li 4.4 Sn, Li 4.4 Ge, and Li 3 Sb), leading to higher specific capacities compared to the conventional carbon-based anode materials. The theoretical specific capacities for Si, Sn, Ge, and Sb are 4200, 994, 1625, and 665 mA·h·g −1 , respectively. Despite their advantage of excellent specific capacities, the large volume change during charging/ discharging causes several problems such as irreversible deformation of anode material, low cycling stability, limited rate capability, and loss of electrical contact. Research on orientation-controlled VA and IVA-based anode materials can provide in-depth understanding to overcome these challenges.
Si has been regarded as a promising anode material due to its high specific capacity, abundant reserves, and environmentally friendly characteristics. Despite the high specific capacity of Si, huge volume expansion (~440%) and low conductivity (1.56 × 10 −3 S·cm −1 ) limit the battery performance using Si anode. Nugroho et al. 120 fabricated vertically aligned n-type Si nanowire array using n-type Si wafers via photolithography and cryogenic inductively coupled plasma-reactive ion etching process. XRD patterns displayed in Figure 10A show the <100> orientation of pristine Si wafer and aligned n-type Si nanowire arrays. Inset figure in Figure 10A exhibits cross-sectional SEM image of n-Si nanowire arrays. The diameter of the Si nanowire is~996 nm and the distance between the closest wires is~10.2 μm. Figure 10B,C shows the electrochemical performance of pristine n-Si wafer and n-Si nanowire arrays. Significant decrease of specific capacities was observed for the half-cell with pristine n-Si wafer (0.21, 0.15, 0.13 mA·h·cm −2 at 2nd, 50th, and 100th cycles) because of increase in internal stress resulted from the local volume expansion. While half-cell with n-Si nanowire arrays show higher specific capacities and stability (0.50, 0.42, and 0.43 mA·h·cm −2 at 2nd, 50th, and 100th cycles) than the that with pristine n-Si wafer. This is due to the larger surface area for lithiation, and shorter Li-ion diffusion path compared to the pristine n-Si wafer. Shi et al. 121 used the Si single crystal to reveal the fundamental fracture mechanisms during charge/discharge cycles. The damage to the Si anode surface occurs gradually and cumulatively. As shown on the left side of Figure 10D, SEM images show the orthogonal surface cracks, which can be observed after 30 cycles. The authors proposed that the orthogonal surface cracks originated from the orientation-dependent lithiation. The mechanical and electrochemical responses after long-term cycling were investigated for the Si single-crystal anode via finite element method (left side of Figure 10D). Because the lithiation in the (110) direction is 6.4 times faster than in the (100) direction, 123 the volume expansion is anisotropic, leading to high shear stress and crack. Since the volume expansion of unit cells is completely constrained by the surrounding material, a large plastic deformation occurs along the (100) edges. This forms the normal stress bands along the <100> directions. This study not only revealed the fracture mechanisms of Si single crystalline anode, but also presented a strategy to improve fracture resistance by healing electrode cracks using electrolyte additives based on physical understanding.
Sn is also a representative alloy-type anode material same as Si. The lithiation process of Sn anode is as follows: (1) The formation of Li 2 Sn 5 phase that remain crystallinity with no visible phase boundaries. (2) The formation of Li 4.4 Sn after further lithiation with a volume expansion of 260%. Like Si anode, the large volume change of Sn anode is also a barrier to overcome for high-performance LIB. Juan et al. 122 investigated the electrochemical properties of single crystalline β-Sn nanorods with controlled aspect ratios of (001) and (100) facets. Figure 10E shows the TEM image of β-Sn nanorods with different morphologies. The spherical β-Sn nanoparticles with diameter of 4.5 nm were obtained with the synthesis temperature of 26°C. β-Sn nanorods with aspect ratios of 2.5 and 11.1 were formed when the synthesis was performed at 10°C and 0°C, respectively. The authors demonstrate that the change in aspect ratio is because of different reduction rates and mobility of the particles in solution. Selected area diffraction patterns in Figure 10F exhibits that β-Sn nanorods grow along the [001] direction, indicating the tip of β-Sn nanorod is (001) facet. As shown in Figure 10G, cyclic voltammetry was measured for β-Sn nanorods with high and low aspect ratio. Cyclic voltammetry curves for both high and low aspect ratio β-Sn nanorods have the reversible and overlapping curves for the second and third cycles, indicating the stable cycling activity. However, β-Sn nanorods with high aspect ratio has a larger total current, implying the smaller ohmic resistance and polarization of high aspect ratio β-Sn nanorods. Furthermore, β-Sn nanorods with high aspect ratio exhibit the improved cycling stability in Figure 10H.
Research on VA and IVA-based anode materials using facet control is lacking. To develop the anodes that take advantage of their large specific capacities, it is necessary to deepen our understanding through well-controlled VA and IVA-based anode material systems.

| Transition metal compound-based materials
Transition metal compounds have been one of the most studied materials for anode. They can be divided into several categories according to the lithium storage mechanisms: (1) intercalation and (2) conversion. Spinel Li 4 Ti 5 O 12 is a well-studied intercalation-type anode material. During lithiation, the phase transition occurs from spinel (Li 4  were grown on (100), (110), and (111)-oriented Nb-doped SrTiO 3 single crystal substrates via pulsed laser deposition method. XRD patterns in Figure 11A shows that outof-plane crystal orientation of Li 4 Ti 5 O 12 films were coherently aligned with the substrate orientation. All Li 4 Ti 5 O 12 films have surfaces with <111> crystal facets, as this is the surface in the lowest energy state ( Figure 11B). As shown in Figure 11C, (100)oriented Li 4 Ti 5 O 12 films shows the highest capacities (313 mA·h·g −1 ), excellent rate performance, and good cyclability with improved cycle life and doubling of reversible capacities compared to the polycrystalline Li 4 Ti 5 O 12 in previous studies. Also, the authors observed that the lithiation waves are traveling more rapidly along the GBs before intercalating into the grains using a phase-field model. Speulmanns et al. 125 fabricated textured Li 4 Ti 5 O 12 films via atomic layer deposition followed by rapid thermal annealing. The pulse time of lithium hexamethyldisilazide precursor is an important factor to obtain the highly textured Li 4 Ti 5 O 12 film. Also, annealing time after deposition is crucial for crystallization. XRD patterns show that a highly (111)-textured film can be synthesized using pulse time of 5 s and annealing time of 8 min. The effect of orientation control on electrochemical performance was investigated. At an extreme rate of 200 C, highly (111)-textured Li 4 Ti 5 O 12 film shows capacities of 278 mA·h·cm −3 . Also, the textured film exhibits 97.9% capacity retention after 1000 cycles at 100 C.
Orthorhombic Nb 2 O 5 (T-Nb 2 O 5 ) is another intercalation-type anode material. The multielectron redox reaction enables a high theoretical capacity (200 mA·h·g −1 ). 130 Like Li 4 Ti 5 O 12 , high potential (1.65 V vs. Li + /Li) of Nb 2 O 5 prevents the formation of lithium dendrites and anode/electrolyte interface layers, which is highly associated with the safety of LIB. Low electronic conductivity is a hindrance to improving the performance of batteries with Nb 2 O 5 anode. Liu et al. 131 reported F-doped Nb 2 O 5 microflowers with nearly 97% exposed (100) facets. Exposed (100) facets of Nb 2 O 5 anode enhance the intercalation performance through the Li-ion transport along the loosely packed 4 g atomic layers. F-doped Nb 2 O 5 anode exhibits the improved the Li +/electron transfer kinetics and pseudocapacitive performance. The F-doped Nb 2 O 5 anode shows capacities of 210.8 and 164.9 mA·h·g −1 at 1 and 10 C, respectively and 85.1% of capacity retention after 2000 cycles at 10 C.
MoS 2 is the transition metal dichalcogenides-based intercalation type anode material. MoS 2 has a layered structure with a layer spacing of 6.2 Å, twice as large as graphite. Weak Van der Waals forces between the layers facilitates the rapid insertion/extraction of Li-ions. It has been considered a good anode material due to its long cycle life, high power density, and twice larger theoretical specific capacity (~800 mA·h·g −1 ). 130 However, there is also a disadvantage that the large volume change is large during charge/discharge. Shokhen et al. 132 investigated the electrochemical performance of vertically aligned and planarly oriented MoS 2 as an anode for LIB. Vertically aligned MoS 2 was synthesized through the chemical vapor deposition method. Planarly oriented MoS 2 was prepared by overlaying a slurry of 70% of MoS 2 powder, 20% of polyvinylidene fluoride, and 10% of carbon black onto the Cu foil via hand coater. Cross-section TEM images and XRD patterns indicate that the MoS 2 layers grow vertically along the (100) orientation. Both vertically aligned MoS 2 and planarly oriented MoS 2 show high specific capacity reaching the theoretical value. However, vertically aligned MoS 2 has higher cycling stability and colombolic efficiency than planarly oriented MoS 2 . Zhang et al. 133 synthesized polyaniline intercalated MoS 2 nanosheets, which are vertically grown on reduced graphene oxide (rGO). By intercalating the polyaniline, interlayer between MoS 2 layers is extended, facilitating the Li-ion insertion/extraction. XRD   928 mA·h·g −1 , 133 respectively. He et al. 129 investigated the electrochemical performance of Fe 2 O 3 hexagonal single crystals with an exposed (100) facet. XRD patterns in Figure 11D exhibits that the (110)/(104) peak intensity ratio increases as the concentration of Al 3+ increases, indicating the a-direction preferred orientation of α-Fe 2 O 3 . TEM images shown in Figure 11E suggest that hexagonal α-Fe 2 O 3 single crystal has the (001) facet dominant. Single crystalline Fe 2 O 3 shows a high specific capacity of 1261.3 mA·h·g −1 at 200 mA·g −1 , 133 outstanding cycle stability with capacity of~900 mA·h·g −1 at 500 cycles ( Figure 11F), and high-rate capability with a reversible capacity of 605 mA·h·g −1 at 10 A·g −1 .
In addition to the anode materials mentioned above, there are several new emerging anode materials such as phosphorus and phosphides. Phosphorus has a suitable potential (~0.8 V vs. Li + /Li) extremely high capacity (2596 mA·h·g −1 ). Also, rock-salt-type structure Li 3+x V 2 O 5 is considered a promising anode material for LIB. To solve the facing problems, existing anode materials as well as newly emerging anode materials are required to be studied in well-controlled material systems.

| SUMMARY AND PERSPECTIVES
As the importance of electric vehicles and renewable energy increases due to climate change, the development of batteries with superior performance and stability is becoming more crucial. Understanding battery systems is critical to improving performance and stability. In this paper, the overall operating process and critical issues of LIB are presented. And then, we review the current state of facet-controlled cathode, solid-state-electrolyte, and anode materials for LIBs, including epitaxial film, textured film, and single crystals. The outline of this review mainly focused on is as follows: (1) Cathode: Well-controlled cathode materials with various structure including layered structure, spinel structure, olivine, and silicate were discussed. LiCoO 2 is the most studied cathode materials with layered structure. Cahode materials with layered structures like LiCoO 2 shows different electrochemical performance depending on the crystal orientation. Recently, further improved cathode materials with layered structure has been studied (e.g., LiCoO 2 , LiNi 1/3 Co 1/3 Mn 1/3 O 2 , and LiNi 0.8 Co 0.15 Al 0.05 O 2 ). Spinel LiMn 2 O 4 single crystal shows higher electrochemical performance than polycrystalline LiMn 2 O 4 . The electrochemical performance of (010)-oriented LiCoPO 4 with olivine structure is higher than that of (100)-and (001)-oriented LiCoPO 4 due to higher Li-ion diffusion. (2) Solid-state-electrolyte: Recent studies of facetcontrolled solid-state-electrolyte materials were reviewed for garnet-type Li 7 La 3 Zr 2 O 12 (LLZO) and perovskite-type Li 3x La 2/3x TiO 3 (LLTO). LLZO has three-dimensional pathway for Li-ion transport, implying an isotropic nature of ionic conduction. By using the well-controlled LLZO, influence of grain boundary on Li-ion transport, lattice distortion effect on conductivity, and electronic conductivity difference between single crystalline and polycrystalline LLZO were investigated. LLTO has been considered as promising solid-state-electrolyte material due to its good ionic conductivity and excellent chemical stability. The ionic conductivity change depending on the area of Li-ion transmission window and the domain orientation are discussed. Additionally, the improved performance of battery systems with well-aligned SEs was addressed. (3) Anode: Three different material groups are mainly discussed for anode including carbon-based materials, VA and IVA-based materials, and transition metal compound-based materials. Carbon-based anode materials, especially graphite, is the most used for commercial anode of LIB due to its cyclic stability and high specific capacity. Vertical alignment facilitates the Li-ion transport, leading to enhancement of electrochemical performance. Apart from the carbon-based anode materials, VA and IVAbased materials have attracted great attention due to their extremely high specific capacities. Facet dependent battery performance and failure mechanism have been studied with single-crystalline system. For the transition metal compound-based anode materials, they mainly consist of intercalation-type material (e.g., Li 4  In all fields of cathode, anode, and electrolyte, research on the electrochemical performance of batteries in welldefined systems using epitaxial films, single crystals, and textured films is in the early stage. This may be because most of the studies so far have been conducted on a bulk CHOI ET AL. | 21 of 26 scale. However, as the need for a battery with a small size and high-performance increases, research for fundamental understanding is greatly required.
Significant advances have been made in improving battery performance. However, there are still many issues and challenges that need to be addressed. Fundamental understanding on the battery system can provide important insights into the development of stable and high-performance batteries. In this regard, the use of epitaxial films, single crystals, and wellaligned layered materials is a powerful strategy to disclose the unknown mechanisms for battery system. This is because these are well-defined and wellcontrolled system that can clarify which factors have a major influence. Also, interface dynamics at cathodeelectrolyte and anode-electrolyte interfaces can be investigated in well-defined and well-controlled system using various engineering approaches such as strain engineering and composition control. To make feasible to study the well-defined battery system, it is also necessary to study more synthetic methods. In addition to building a well-controlled material system, utilizing advanced characterization techniques can lead to many advances in understanding of battery system. Multiscale and high-resolution capabilities of characterization tools can help observe the Li-ion transport. In situ measurement technologies including in situ transmission electron microscopy, in situ secondary ion mass spectroscopy, in-situ Raman spectroscopy, in-situ X-ray diffraction, and X-ray photoelectron spectroscopy can also play a key role in revealing the Li-ion insertion/ extraction process and formation dynamics of the cathode-electrolyte and anode-electrolyte interface layers. Additionally, combination of experimental study with well-controlled material system and theoretical study can deepen our understanding of battery systems. Fundamental understandings on battery systems can provide insights that can lead to innovations and guidelines for designing new battery systems.
This review takes an overview of state-of-the-art LIB system using well-defined materials system such as epitaxial films, textured films, and single crystals. It is hoped that this review will provide new breakthroughs for advanced batteries in the future.