Introduction

Lithium-ion batteries (LIB) have been societally omnipresent ever since their inception three decades ago1,2. However, current and future applications are pushing the performance requirements to the point where current LIB technology will soon be insufficient; improvements in capacity, longevity, and cost reductions are needed for particularly demanding technologies such as long-range electric vehicles3 and battery-powered flights4. The performance of LIBs is often limited by the cathode, which is typically the most expensive component while also storing far less reversible capacity per unit weight as compared to a typical graphite anode. Therefore, improvements to the cathode are an effective way to meet these requirements, and ternary layered oxide cathodes (TLOCs) are subject to the most research and development due to their relatively high energy density and low cost5. TLOCs are generally comprised of the composition LiMO2, where M is typically a combination of transition metals and other elements such as Ni, Mn, Co, Al6,7,8, B7,9, Ta7,10, and W5,7. Synthesis methods, performance modification methods, failure mechanisms, and recycling methods of TLOCs have been analyzed and summarized in detail in various reviews11,12,13,14,15; however, there is a lack of research focusing on the mechanisms of how the various cathode morphologies affect the performance of TLOCs, as morphology plays an important role in the properties of TLOCs7. Various microstructures have been demonstrated to either enhance reversible capacity at high charge/discharge rates or improve longevity by alleviating particle strain and resisting surface degradation.

Dense polycrystalline (PC) spherical TLOCs are widely applied in commercial production because of their high tap density, uniform particle size distribution, and high fluidity, leading to high energy density and electrode loading. However, the polycrystalline structure is also well known to develop intergranular cracks during charging at high voltage over extended cycles, which exposes additional surface areas to parasitic side reactions and subsequent secondary particle pulverization, resulting in shortened cycle life16,17,18. Meanwhile, the randomly ordered primary particles cause tortuous Li+ diffusion pathways inside the secondary particles, resulting in long diffusion distances19. The relevant parameters that can affect the Li+ diffusion and electrochemical stability of TLOCs are summarized in Fig. 1, and most of them can be improved by adjusting particle morphology. For Li+ diffusion, diffusion barrier, diffusion channels, crack generation, and stability can be optimized by controlling the morphology of TLOCs, while for stability, stress caused by phase transformation, crack generation, electrolyte side effects, and transition metal dissolution also can be controlled by optimizing the morphology. Thus, to alleviate these problems, morphological strategies, including surface modification, application of core-shell structures, and secondary and primary particle morphology tuning, have been investigated and developed to resist improved Li+ diffusion pathways or surface degradation7,20,21. Furthermore, single crystal (SC) structures are investigated as a suitable particle morphology for better long-term cyclability, due to the complete absence of internal grain boundaries11. The performance of SC TLOCs is also heavily dependent on their morphologies, especially in regard to enclosed facets and crystal size, which all affect Li+ diffusion and susceptibility to surface reconstruction22,23,24,25.

Fig. 1: Schematic depicting parameters that affect Li+ diffusion and stability.
figure 1

The two images illustrate the important parameters related to Li+ diffusion and stability, and the morphology of cathode materials can control Li+ diffusion and stability by influencing these parameters.

In this review, morphology includes secondary particle size, secondary particle shapes, inner morphology of secondary particles, primary particle size, primary particle shapes, single crystal particle size, and single crystal particle shapes. For secondary inner morphology, it includes primary particle orientation, pores distribution, and gradient structure. We focus on the mechanism of how morphology affects the electrochemical performance of TLOCs in the aspects of Li+ diffusion and stability. First, we discuss how morphology affects Li+ diffusion path and rate by tuning Li+ diffusion-related parameters (Fig. 1). Second, we overview the problems leading to degradation (Fig. 1) and analyze how to solve these problems by controlling the morphology of TLOCs. Last, we discuss the challenges and our perspectives on future morphology design in the field of fundamental research, sustainability of morphology, and scale-up.

Morphology-controlled Li+ diffusion

Li+ diffusion plays a critical role in the rate performance of TLOCs. The diffusion time of Li+ from the core to the surface of a particle can be influenced by the diffusion path and diffusion rate. This relationship is described using Einstein’s formula: \(\tau ={\lambda }^{2}/D\), where τ represents diffusion time, λ represents transverse length, and D represents the diffusivity26. Therefore, D and λ are critical factors that impact Li+ diffusion in cathode materials. As shown in Fig. 2a, within a secondary particle, there exist various Li+ diffusion pathways. These Li+ diffusion pathways have different diffusion paths and diffusion rates. Ideally, Li+ diffusion pathways should have both a short diffusion path and a fast diffusion rate to facilitate efficient Li+ migration. The morphology of TLOCs can significantly affect their Li+ diffusion path and rate, which will be discussed in this section.

Fig. 2: The effects of morphology on the Li+ diffusion length.
figure 2

a Schematic depicting possible and unlikely Li+ diffusion pathways inside secondary particle: (I) shown Li+ diffusing tortuous path through a-b planes in grains; (II) shown that Li+ can not diffuse along grain boundaries; (III) shown that Li+ can diffuse across grain boundaries with tight contact; (IV) shown that Li+ can diffuse across twin grain boundaries; (V) shown that Li+ can diffuse in sharpe concentration change area by two-phase reaction with; (VI) shown that Li+ can diffuse through gradual concentration change area by solid-solution reaction with. b Scanning electron microscopy (SEM) image of a secondary particle with sheet morphology. c SEM image of a secondary particle with tube morphology and the inner is the TEM image of a secondary particle with tube morphology. d SEM image of a secondary particle with porous morphology. e Schematic depicting secondary particles with randomly arrogated block-like primary particles with the characterized Li+ diffusion pathways and surface. f Schematic depicting secondary particles with randomly arrogated mixed primary particles with the characterized Li+ diffusion pathways. g Schematic depicting secondary particles with ordered arregated platelet primary particles with the characterized Li+ diffusion pathways, active planes, and surface. h Schematic depicting secondary particles with ordered arregated rod primary particles with the characterized Li+ diffusion pathways. Panel a is adapted with permission from ref. 36, Copyright American Chemical Society, 2021. Panel b is adapted with permission from ref. 30, Copyright Royal Society of Chemistry, 2018. Panel c is adapted with permission from ref. 33, Copyright Royal Society of Chemistry, 2018. Panel d is adapted with permission from ref. 32, Copyright Springer Nature Limited, 2016. Panel e and g are adapted with permission from ref. 19, Copyright Wiley-VCH, 2019. Panel h is adapted with permission from ref. 41, Copyright Elsevier, 2021.

Li+ Diffusion path

Li+ diffusion path refers to the pathways for Li+ diffusing from the particle core to the surface, which includes two critical parameters: the Li+ diffusion distance and Li+ diffusion path. The diffusion distance represents the distance from the start point to the end point of diffusion, while the diffusion path means the actual path for Li+ diffusing from the core to the surface of the particle. When electrolytes can diffuse into the inner part of secondary particles, like porous morphology and hollow rods, the Li+ diffusion distance is the shortest Li+ diffusion path because Li+ just needs to transfer from the center to the surface of the primary particle and then diffuse into the electrolyte. When an electrolyte only can wet the surface, Li+ needs to diffuse from the center of the secondary particle, after crossing lots of primary particles, the Li+ can reach the surface and then diffuse into the electrolyte. The Li+ diffusion distance will depend on the particle size while the Li+ diffusion path will depend on particle size, path tortuosity, and the hindrances to Li+ diffusion. In TLOCs, including secondary particles of PC TLOCs, primary particles of PC TLOCs, and SC TLOCs, their morphologies play critical roles in their Li+ diffusion distance and length. Thus, in this section, the effects of their morphologies on their diffusion path will be discussed.

Morphology-controlled Li+ diffusion distance

To decrease the Li+ diffusion distance from bulk to surface, the most effective method is reducing the particle size, leading to a shorter diffusion distance. However, too small a particle size will cause some problems, like lower packing efficiency and tap density, leading to low energy density of the cell26,27. Therefore, balancing the Li+ diffusion distance and other properties is important during controlling particle size.

Meanwhile, altering the shapes of particles also can achieve short Li+ diffusion distance by allowing more electrolyte wetting of the particles28,29. Particles with 2D sheet morphology, as shown in Fig. 2b, have been studied to decrease the Li+ diffusion distance by allowing more particles at the surface, enabling ultrafast Li+ diffusion30. Hollow rods31 and porous morphologies (Fig. 2c, d) also can allow electrolytes to contact the inner particles and thus decrease the Li+ diffusion distance32,33.

Morphology-controlled Li+ diffusion path

To achieve shorter Li+ diffusion paths, the inner morphology of TLOCs plays a crucial role. As shown in Fig. 2e–h, it has been observed that particles with different morphologies can have different lengths of Li+ diffusion path, even when the diffusion distance is the same34,35,36. This is mainly caused by the different tortuosity of Li+ diffusion paths.

To reduce the tortuosity of Li+ diffusion paths, controlling the orientation of primary particles and limiting the hindrances to Li+ diffusion are effective strategies. Primary particles within the secondary particle can exhibit different crystal orientations, resulting in varying tortuosity of Li+ diffusion pathways. Controlling the morphologies of primary particles is an effective way to organize their orientation. As illustrated in Fig. 2e, f, randomly oriented primary particles, with block-like or mixed shapes, can lead to long and winding Li+ diffusion paths, particularly between grains with inconsistent crystal planes, due to migration across the grain boundaries19. On the other hand, as shown in Figs. 2g, h, primary particles with platelet34,37,38 and nanorod39,40,41 morphologies, which are aligned along (003) planes, tend to be arranged radially, providing nearly straight and the shortest Li+ diffusion paths from the secondary particle cores to their surfaces7,35,38,41,42,43,44. The longitudinal axis of nanorod primary particles is parallel to the Li+ diffusion path, and they exhibit only slight angular deviation, thereby enabling almost straight Li+ diffusion paths across adjacent primary particles and significantly enhancing the rate performance7,45,46,47.

Conversely, places that impede Li+ diffusion, such as loose contacted grain boundaries, cracks, and thick NiO-like layers on grain boundaries (Fig. 2a, e), can increase the tortuosity of Li+ diffusion pathways36. It is worth noting that these places typically develop during cycling, making the stability of particles crucial for Li+ diffusion. Therefore, in the next section, we will discuss the relationship between stability and morphology.

Li+ Diffusion rate

Morphology can affect not only the Li+ diffusion path but also the Li+ diffusion rate. The speed of Li+ movement in the cathode material is determined by the Li+ diffusivity (D). D can be represented through an Arrhenius relationship: D = D0 exp \((-\triangle G/{kT})\), where D0 represents the pre-factor determined empirically, ∆G represents the energy barrier of migration, k is the Boltzmann constant, and T is temperature48. Therefore, changing the energy barrier for Li+ diffusion is a critical strategy to enhance the Li+ diffusion rate. Morphology can affect the energy barrier for Li+ diffusion in several aspects, including grain boundaries, active area, and the distribution of crystal facets. These effects will be covered in this section.

Morphology controlled grain boundary effects on Li+ diffusion rate

Morphology can affect the Li+ diffusion rate by influencing grain boundary in the aspect of the grain misorientation angle. Inside TLOCs, the energy barrier for Li+ diffusion can be influenced by whether the diffusion is through the bulk or across grain boundaries. Generally, due to the high grain boundary energy, the asymmetric grain boundaries between the primary particles are obstacles to Li+ diffusion, which is the primary factor limiting Li+ diffusion in TLOCs47,49,50,51. As depicted in Fig. 3a, by decreasing the grain misorientation angle, Li+ is easier to diffuse across the grain boundary36,52. Optimal Li+ diffusion is achieved when the transport planes are parallel to the radial direction, as shown in Fig. 3b52. The grain misorientation angle can be controlled by the shape and orientation of primary particles, particularly with platelet and rod shapes, the ordered orientation of primary particles can be achieved as we mentioned before, resulting in a decrease in the grain misorientation angle36,52,53. Furthermore, as illustrated in Fig. 3c, SC TLOCs can provide a more continuous Li+ diffusion pathway than PC TLOCs since they don’t contain grain boundaries, and thus deliver a higher Li+ diffusion rate.

Fig. 3: The effects of morphology on Li+ diffusion rate.
figure 3

a Schematic depicting difficulties of Li+ across grain boundaries with different grain misorientation angles. b Schematic depicting the best case for Li+ across grain boundaries. c Schematic depicting the case for Li+ to diffuse inside an SC particle. d The rate performance of secondary particles with different particle(secondary particle) sizes and secondary particles with the same secondary particle size but different grain(primary particle) sizes. e The schematic of cathode–electrolyte interface layer (CEI) and Li+ diffusion rate on stable and unstable particle surface. f The schematic of the (001), (012), (104), and (010) facets. g Transmission electron microscopy image of primary particle exposing (012) and (104) plane on the surface of secondary particle. h The schematic of the presence and percentage of (012), (001), and (104) facets on the surface of SC particles with different morphologies. Panel a and b are adapted with permission from ref. 52, Copyright Elsevier, 2020. Panel c is adapted with permission from ref. 80, Copyright Elsevier, 2020. Panel d is adapted with permission from ref. 27, Copyright Elsevier, 2021. Panel e is adapted with permission from ref. 81, Copyright Elsevier, 2015. Panel f is adapted with permission from ref. 61, Copyright Elsevier, 2022. Panel g is adapted with permission from ref. 8, Copyright Elsevier, 2020. Panel h is adapted with permission from ref. 25, Copyright Royal Society of Chemistry, 2019.

Morphology controlled surface area effects on the Li+ diffusion rate

The surface area is the region of the cathode particle contacting the electrolyte. Through changing the surface area of TLOCs, Li+ diffusion rate can be controlled by morphology, because the surface area can have a significant impact on the Li+ insertion/extraction behavior through affecting the number of active sites for Li+ to insert and extract in a given amount of time54. Morphology can control the surface area by the surface structures, size, and shapes of TLOCs. Although theoretically large surface area causes a high Li+ diffusion coefficient, a very large surface area can lead to serious side reactions that can consume electrolytes and form an unstable cathode–electrolyte interface layer (CEI). These negative effects will impede the interfacial Li+ transfer between electrolyte and TLOCs, and thus affect the Li+ diffusion coefficient and rate performance55,56,57. For example, in Fig. 3d, TLOCs with large grains (primary particles) deliver the best rate performance among the different sizes of grains. The main reason is that TLOCs with large surface area have high surface energy, leading to serious side reactions with electrolytes and forming an unstable intimate coating layer from electrolyte decomposition56,57 (Fig. 3e). Thus, it is critical to have a balance between limiting the side reaction and promoting Li+ diffusion by controlling the surface area, and this can be achieved by modifying the morphology.

One way to adjust the surface area is by controlling the surface structure of TLOCs32. For instance, as shown in Fig. 2b, c, TLOCs with porous morphology can increase the surface area by generating pores on the surface or applying a hierarchical structure. The surface area can also be controlled by the size and shapes of both the whole particle (PC and SC TLOCs) and the primary particle. Commonly reported SC TLOCs are ~1–5 µm in size, while for PC TLOCs, secondary particle sizes can range from several microns to over 10 microns12. A larger particle size results in a smaller specific surface area. With the same size, PC TLOCs with smaller primary particles deliver higher specific surface area because their surface contains more primary particles which can provide more exposed area9,56,57. Due to the same reason, PC TLOCs show higher specific surface area than SC TLOCs of the same size27,58. As shown in Fig. 3d, small particles deliver better rate performance than baseline and large particles since small particles have higher surface area. Additionally, compared to spherical shapes, TLOCs with other shapes, like rods, sheets, and dumbbells, can also increase the surface area, as the spherical shape has the smallest surface area compared to other shapes with the same volume59,60. However, a large surface area may cause low capacity due to the serious side reaction. For example, although particles with sheet morphology can deliver a larger surface area than particles with spherical morphology, their capacity of them is lower, as shown in Table 1.

Table 1 Electrochemical performance and Li+ diffusion coefficient of PC TLOCs with different morphologies.

Morphology controlled distribution of crystal facets effects on Li+ diffusion rate

The morphology of TLOCs can control the Li+ diffusion rate by changing the distribution of crystal facets because different crystal facets have different Li+ diffusion channels61 (Fig. 3f). TLOCs have the α-NaFeO2 layered structure, and the only channel for Li+ diffusion is the 2D lithium layer plane. In PC TLOCs, primary particles normally have random block-like shapes, exposing (001) plane7,19,35,62. The Li+ insertion/extraction can be impeded when (001) facets, without Li+ diffusion channels, are randomly exposed, leading to a low Li+ diffusion rate61,63. Compared to (001) facets, other facets contain more open Li+ diffusion channels19,34,64. Therefore, to improve the Li+ diffusion rate, more crystal facets with open Li+ diffusion channels are expected to be exposed by controlling the morphology of primary particles. For example, nanorod primary particles can make secondary particles expose (104) crystal facets on the surface, as shown in Fig. 3g, and thus can provide more Li+ diffusion channels8. Similarly, as shown in Fig. 2g and Table 1, controlling the surface with the electrochemically active (010) planes can result in radially oriented primary particles, and thus 3D Li+ diffusion paths can be created, enabling Li+ diffuse along the active planes and straightforwardly from the center to the surface without crossing grain boundaries, significantly improving the Li+ diffusion rate19.

For SC TLOCs, the morphologies include nanorods45,65,66,67,68, platelets69,70,71,72,73, octahedron72,74,75, tetradecahedron76, and truncated octahedron25. Their morphologies also play important roles in controlling Li+ diffusion rate since particles with different morphologies have different distributions of crystal facets. According to experiment and density functional theory calculation, (001), (012), and (104) crystal facets are crucial in determining the equilibrium particle shapes25,77. As shown in Fig. 3h, SC particles with different morphologies exhibit distinctive distributions of these crystal facets. Therefore, altering the morphologies of SC particles can modify the distribution of these crystal facets, thereby influencing the Li+ diffusion and rate performance, as shown in Table 2. For example, tetradecahedron SC NMC811 with (104) crystal facets delivers 169 mAh g−1 discharge capacity, 10 mAh g−1 and 36 mAh g−1 higher than octahedron NMC811 with (012) crystal facets and PC NMC811, respectively76. Similar results can also be found in LiNiO2, and other types of cathode materials, so more research about modification of the distribution of crystal facets by controlling morphology is expected78,79.

Table 2 Electrochemical performance and Li+ diffusion coefficient of SC TLOCs with different morphologies.

Morphology controlled stability

The morphology of TLOCs is a critical factor that has significant effects on stability. One key consideration is the particle size and shape of primary, secondary, and SC TLOCs, as these characteristics can influence the mechanical stress and strain experienced by TLOCs during charging and discharging cycles. Variations in particle size and shape can result in different levels of cracking and structural degradation of TLOCs, which in turn can impact the capacity, energy density, and cycle life. Additionally, the surface area and porosity of the TLOCs can impact Li+ diffusion within TLOCs, which has been discussed in the section “Morphology controlled Li+ diffusion”. Although the surface area and porosity can affect the overall electrochemical performance, it is important to note that high surface area and porosity can also increase the likelihood of unwanted side reactions between the TLOCs and electrolytes, which can negatively impact the stability during cycling. Thus, optimizing the morphology of TLOCs is crucial for achieving high-performance and stable TLOCs. In this part, the discussion of stability is focused on cycling stability, which can be affected by morphology and structure degradation. Morphology degradation is related to stress release, and tolerance, while structure degradation involves phase degradation caused by side reaction between cathode surface and electrolyte.

Stress release and tolerance

Cracks, including intergranular fracture and intragranular fracture, play a significant role in the capacity fading of TLOCs. Such cracking phenomenon has been widely discussed in the literature, and Yin et al.82 summarized the mechanism of microcrack formation and evolution recently. In brief, it is widely accepted that cracks are caused by phase transformation, cation mixing, lattice oxygen loss, lattice collapse, surface reconstruction, and heterogeneous lithiation/delithiation, leading to the disintegration, pulverization of particles83 and decreased thermal stability, structural stability and cyclic stability of TLOCs82. Hence, many works focused on coating, doping, and grain boundary modification to suppress the generation of microcracks. However, the effects of morphology design have been underestimated. Morphology design can play a significant role in reducing the impact of cracks by facilitating stress release and increasing strain tolerance. This is because the stress change caused by volume fluctuations during charging and discharging can be addressed through careful optimization of the particle size, shape, porosity, and element distribution. As such, morphology design can significantly contribute to reducing the negative impact of cracks on the durability and reliability of TLOCs.

Morphology effects on heterogeneous stress

Heterogeneous stress, generated from the electrochemical expansion and contraction during charging/discharging, is the main reason for forming inter and intragranular cracking of particles84. In particular, large particle size and random grain orientations are thought to exacerbate these cracks because large particle size can cause heterogeneous stress, which is generated from inhomogeneous delithiation/lithiation during the charging and discharging process, and random grain orientations may bring huge internal stress between primary particles. Normally, small particle size experiences low stress, because the Li concentration is more uniform in smaller particles during lithiation and delithiation85. In particular, Cheng et al.86 found that by decreasing particle size, tensile stress may be reduced in magnitude or even be reverted to compressive stress. However, very small particles compromise ionic partial conductivity and ionic percolation on the cathode level. Considering the more reaction between small particles and electrolytes due to the large surface area, there must be an optimum particle size87.

Then, to further improve stress release and tolerance, researchers turned to modifying the morphology of primary particles, as shown in Fig. 4a and Table 1. Lin et al.47 clarified primary particles’ morphologies into two types: equiaxed and radially elongated morphologies. Compared to equiaxed primary particles, radially elongated primary particles have better stability and mechanical properties because they can uniformly dissipate the mechanical strain caused by internal structural changes via stacked in order along the radial direction. This radial morphology, including platelet and nanorod morphologies, can be obtained by adjusting coprecipitation reaction conditions, like pH, ammonia concentration35, surfactant35,43, and atomic doping47. Radially aligned primary particles can reduce the cation mixing and alleviate the volume change caused by stress concentration during reversible Li+ de/intercalation, leading to better stability47. Xu et al.19 synthesized NMC811 spherical secondary particles with radially aligned primary particles by a modified coprecipitation reaction. The radial primary particles with consistent crystal orientation can alleviate the volume-change-induced intergranular stress by coordinated expansion and contraction, as shown in Fig. 2e, leading to less pulverization of secondary particles and improved cycling stability (without any visible degradation after 300 cycles). Introducing different doping elements, such as B7,8, Ta7,88, Sb89, W7,90,91,92,93, Sn41, Mo94, and more95, can also obtain radially aligned primary particles. Although all of these research claimed the improvement in stability, they all used equiaxed primary particles as a control group. There are no studies on how these doping strategies improve the stability of pristine spherical secondary particles with radially aligned primary particles.

Fig. 4: The effects of morphology on stress release and tolerance.
figure 4

a The schematic of a build-up of local stress concentrations and stress distribution during charging depending on primary particle morphology. b The schematic of different degrees of sustained damage because of the different morphology. c The schematic of concentration gradient designs with Co-rich or Mn-rich surface and their mechanical stability. d The schematic of porous morphology. e SEM image of hollow microspheres. f The finite element modeling of the equivalent stress within the normal and hollow NMC particles upon completion of charging. g Cross-sectional image of the Li+ concentration at the completion of each discharging and charging process in a spherical cathode with 0%, 20%, and 50% porosity, respectively(C-rate is 40 C). h The schematic of crack evolution and the internal morphological difference for PC and SC particles during prolonged cycling. Panel a is adapted with permission from ref. 9, Copyright Wiley-VCH, 2020. Panel b is adapted with permission from ref. 142, Copyright Wiley-VCH, 2019. Panel c is adapted with permission from ref. 97, Copyright Springer Nature Limited, 2021. Panel e is adapted with permission from ref. 143, Copyright Wiley-VCH, 2019. Panel f is adapted with permission from ref. 144, Copyright Wiley-VCH, 2019. Panel g is adapted with permission from ref. 110, Copyright Springer Nature Limited, 2017. Panel h is adapted with permission from ref. 145, Copyright Elsevier, 2020.

Another essential factor affecting stability is the Ni content, which inherent structural instability caused by anisotropic lattice contraction in the deeply charged state generates local stress concentrations along the grain boundaries17,96, when the fraction of Ni is increased. Herein, core-shell morphology has been studied in abundant research articles, since it can reduce the internal stress concentration, as shown in Fig. 4b and Table 1. By employing core-shell morphology, the Mn/Co-rich shell could improve the mechanical properties97 and provide pressure to the inner core, which will be balanced by the tensile stress82. For example, core-shell TLOCs with a [Ni0.8Co0.1Mn0.1](OH)2 core and [Ni0.5Mn0.5](OH)2 shell are designed to simultaneously ensure structural and thermal stability98. In comparison with NMC811, the core-shell structure exhibited a lower discharge capacity but enhanced cyclability, retaining 17% more of its capacity than that of NMC811 (1 C, 3.0-4.3 V) after 500 cycles98. Besides, Liu et al. 97 studied the effects of Co and Mn on the mechanical properties of Ni-rich TLOCs, which provide guidelines for the design of core-shell TLOCs. As shown in Fig. 4c, their work revealed that the Co-enriched surface design benefits from its low stiffness, which can suppress the generation of particle cracks. Meanwhile, the Mn-enriched core limits internal expansion and improves structural integrity.

However, the compositional mismatch between core and shell is difficult to prevent, especially during intercalation, and the interface is susceptible to crack and reduction of the lithium-ion transport efficiency from the core99. To overcome these challenges, the core-shell morphology evolved to core-shell concentration gradient100, then to full concentration gradient40,101,102. Combining them with primary particle morphology modification can further improve the stability100,102,103,104. Nevertheless, due to the morphological similarity between the full gradient strategy and the conventional spherical secondary particle, an extensive discussion of the former will not be pursued at this time. Although core-shell and full gradient morphology exhibit impressive performance and stability, they are still limited by their repeatability and consistency for scaled-up105.

Morphology effects on intergranular stress

Handling intergranular stress is also a potential strategy to improve the mechanical properties of TLOCs from morphology, which can be achieved by reserving space to accommodate volume change or removing grain boundaries. Thus, porous32, hollow106, and SC morphologies have been developed, and they show better cycle stability than unmodified TLOCs, as shown in Table 1.

Porous and hollow morphologies can be obtained by coprecipitation reaction107 and sacrificial templates method32,108. Porous NMC111 exhibited excellent rate and cycle performance at high voltage32,109. Porous morphology increases the surface area and decreases the length of the diffusion pathway110, which makes lithiation and delithiation more efficient, leading to a homogeneous delithiation and uniform stress generation during the charging/discharging process for suppressing cracks formation (Fig. 4d–g). Due to the same reason, as shown in Fig. 4e, hollow microspheres can also deliver a homogeneous delithiation and thus improve the rate and cycle performance; at 5 C, the capacity difference between hollow spherical NMC111 and normal spherical NMC111 was over 60 times. A similarly produced hollow NMC111 exhibited a capacity retention improvement of almost 30% after 150 cycles (1 C, 2.8–4.45 V) (ref. 111). Furthermore, Liu et al.112 analyzed the advantages of hollow morphology on stress reduction and Ma et al.113 showed that larger voids can alleviate stress concentration and more porous can lower tensile hoop stress (Fig. 4g). However, porous and hollow morphology will lead to a low tap density, resulting in low volume energy density.

SC morphology is an important advance in improving stability through morphology engineering. SC morphology eliminates the intergranular cracks due to the grain boundaries-free structure (Fig. 4h). Compared to intergranular cracks, the intragranular cracks have less influence on stability, thus the cycle stability of SC NMC622 is over 30% better than that of the PC NMC62269. Meanwhile, the intergranular cracks could be reduced by adjusting the orientation of the main crystal plane in SC TLOCs. Zhu et al. found that it could enhance stability by replacing the active (012) crystal face with inactive (001) or (104) planes25. In addition, SC particles exhibit higher mechanical strength than PC particles. Cha et al. found that SC particles can maintain their original morphology from 15 to 45 MPa and they have homogeneous reactivity during the charging and discharging process114. Overall, SC morphology eliminates the intragranular cracks and provides high mechanical properties and high cycle stability as shown in Table 2, however, it is limited by the control of morphology and size.

Other morphology designs effects on stability

There are other shapes of TLOCs, like rods31,115,116, wires29, sheet28,30,117, fibers118, and ellipsoidal119. They are attractive morphologies since they allow both Li+ ions and electrons to be inserted feasibly, facilitating high ionically and electronic conductivities simultaneously, but also large surface-to-volume ratios. Mixed metal oxalate micro rods can be regarded as precursors via a two-step coprecipitation reaction to prepare microrod LiNi0.8Co0.15Al0.05O2 (ref. 116.). The obtained microrod TLOCs exhibit a 6% better capacity retention after 100 cycles at 1 C. Hollow micro rods of NMC811 and NMC111 with high performance can be synthesized by adjusting the coprecipitation reaction31,33. In one work, CoC2O4 was chosen to precipitate first because its growth preference into nanorods can be utilized as a template during the precipitation of the other transition metal (TM) elements. In another work, nanofiber NMC622 with good performance was fabricated by ultrasound-triggered cation chelation and reassembly route29. However, these results lack the comparison with commercial PC TLOCs. And few studies have reported their mechanical properties. Stein et al. revealed the distributions of stresses in ellipsoidal particles, which are lower than spherical particles because the stresses were equilibrated by deformations along the semi-major axis119. Deshpande et al. only studied the diffusion-induced stress in nanowire morphology without comparison with other morphologies120. Therefore, there is still abundant work for other shapes development. Furthermore, the scale-up, repeatability, and consistency are also important for these morphologies.

Surface stability

Since the reaction between TLOCs and electrolytes at high states of charge leads to mechanical instability, reducing the surface area exposed to electrolytes and preventing parasitic electrolyte attack is important to reduce the accumulation of NiO-like impurity layers at the cathode–electrolyte interface82, which could be achieved by morphology modification, as shown in Fig. 5a. However, the good contact between cathode and electrolyte could benefit the Li diffusion, leading to better rate performance and less heterogeneous stress. Thus, there should be a balance for the contact between the cathode and electrolyte.

Fig. 5: The effects of morphology on surface stability.
figure 5

a The schematic of the effects of primary particle shapes on the chemical stability of the cathode surface during charge and discharge cycling. b Areal fraction of microcracks of secondary particles with different morphologies of primary particles in the charged state after different holding times at 4.3 V(The bars represent the statistical distribution of values). c Cross-sectional image of the Li+ concentration in the particle with 40% porosity and an uneven distribution. Panel a is adapted with permission from ref. 88, Copyright Wiley-VCH, 2018. Panel b is adapted with permission from ref. 123, Copyright Wiley-VCH, 2022. Panel c is adapted with permission from ref. 110, Copyright Springer Nature Limited, 2017.

Morphology-controlled surface area effects on surface stability

The particle size can directly affect the surface area of cathode particles. As reducing the particle size of TLOCs, the increased surface area is more prone to surface reaction and particle dissolution in the electrolyte, resulting in a severe reduction in the cyclic life of TLOCs121. Li et al. reported that large LiNi0.7Co0.15Mn0.15O2 cathode particles had an enhanced cycle performance benefited by smaller surface area and thick CEI layer122.

The modification of primary particles can also affect the stability of TLOCs by influencing the contact between the cathode and electrolyte. As shown in Table 1, this method can significantly improve the stability. The tight accumulation of elongated primary particles could reduce the contact with the electrolyte47. Namkoong et al. investigated the effect of exposure time to the electrolyte on the TLOCs at a high state of charge. In this work, as shown in Fig. 5b, the TLOCs with radially elongated primary particles show less exposure time owing to the suppression cracks generation by radially primary particle morphology, which reduces the contact between the new exposure surface and electrolyte123.

Morphology-controlled surface protection

Core-shell morphology with full concentration gradient reduced the contact between Ni4+ ions and electrolytes by reducing the content of Ni on the surface, as shown in Fig. 4b. The high content of Ni2+ and Mn-rich surface suppresses the reaction from Ni2+ to Ni4+, resulting in less reaction between Ni4+ and electrolyte100,102 and the improved stability. In addition, since SC morphology eliminates the intergranular fractures, the amount of new surface exposure to the electrolyte is significantly reduced, as shown in Fig. 4h, which prevents contact between the new surface and electrolyte82. Furthermore, porous, hollow, and other morphologies always provide a high surface area. Hence, it was of great importance to determine the optimal porosity. Song et al. investigated the effects of pore distribution in the spherical TLOCs110, and they stated that if the pores are not evenly distributed, the performance could decrease with increasing porosity, as shown in Fig. 5c. Thus, how to balance both stability and surface area by morphology control have not been fully studied. Controlling both particle size and percentage of different crystal facets at the same time might be a possible solution, and more theoretical calculations, as well as experiments in this field, are needed.

Summary and perspective

The morphological configuration of TLOCs plays a critical role in determining the performance of LIBs. In this paper, we discuss the effects of morphological design on Li+ diffusion and mechanical properties which exhibit the significant potential to affect the capacity and stability of TLOCs. By modifying the morphology, the Li+ diffusion path and rate can be altered, leading to distinct rate performance characteristics. Furthermore, an optimal morphology can effectively alleviate and withstand the stress induced by volume changes during the charging and discharging processes. Nevertheless, it is widely acknowledged that a trade-off relationship exists between rate and cycle performance, which extends to the domain of morphology design. Consequently, researchers must strike a delicate balance in their morphology design efforts to achieve desirable rate and cycle performance characteristics.

Hence, we believe that future morphology design could be further developed from the following aspects: (1) fundamental research; (2) sustainable morphology; and (3) scale-up. The relevant concepts are summarized in Fig. 6. First of all, the function of morphology should be clarified in fundamental research. For example, morphology parameters are always ignored in the study of coating and doping for Ni-rich TLOCs. The absence of essential morphological properties, including particle size, BET surface area, and porosity, in various articles could potentially explain the observed discrepancies in performance despite employing similar coating or doping strategies. Moreover, the effects of primary particles’ morphology are neglected. In many published articles, researchers utilize pristine particles featuring block-shaped primary particles as the starting material for coatings or doping experiments. As a result, the morphology of the primary particles undergoes a transformation from block-shaped to rod-shaped, which has been shown to enhance both rate and cycle performance. Surprisingly, there is a noticeable absence of studies comparing the performance improvements achieved by modifying pristine particles with rod-shaped primary particles. Consequently, it remains unclear whether the changes in primary particle morphology or the effects of the coating or doping elements primarily drive the observed performance enhancements. Therefore, it is crucial to address this knowledge gap and elucidate the relative contributions of primary particle morphology and the presence of coating or doping elements to the overall performance improvements.

Fig. 6: The summary and perspective of the effects of morphology.
figure 6

a The schematic of challenge in studying the effects of morphology. b The schematic of preferred single crystal morphology for sustainability (crack evolution and the internal morphological difference for PC and SC particles during prolonged cycling). c SEM image of morphology of pristine, cycled, and recycled cathode materials from the hydrometallurgical process, solid-state reaction, hydrothermal reaction, and Ionothermal lithiation. d The challenges for scale-up. e The methods for scale-up. Panel b is adapted with permission from ref. 145, Elsevier, 2020. Panel c is adapted with permission from ref. 113, Copyright Elsevier, 2021, ref. 127, Copyright American Chemical Society, 2022, ref. 146, Copyright Elsevier, 2022, and ref. 147, Copyright Wiley-VCH, 2020.

In addition, compared to the study of PC morphologies, SC morphology is a new strategy to improve the stability of TLOCs, especially for Ni-rich TLOCs. Thus, many morphological effects of SC TLOCs need to be explored. Up to date, all SC TLOCs are mixed with several different morphologies, which may lead to problems in electrode manufacturing and research about the function of morphology. Therefore, it is notable to study how to synthesize SC TLOCs with single morphology and whether the performance of SC TLOCs with single morphology is better than that of SC TLOCs with mixed morphologies. In addition, for SC particles synthesized from precursors, the effects of primary particles in precursor on SC morphologies have not been comprehensively investigated. As coating/doping elements have an impact on primary particles’ morphology in secondary particles, their effects on SC morphology or crystal facets have not been reported. Thus, there are many puzzles that need to be addressed for SC morphology.

Besides, morphology optimization can also improve the interface between the cathode and solid-state electrolyte in all-solid-state batteries (ASSBs). The morphological effects, such as particle size, have been investigated124,125. Especially, SC morphology is a potential candidate of TLOCs for ASSBs, which can provide a better interface reaction between the cathode and solid-state electrolyte compared with PC126. However, other morphologies of PC TLOCs, like primary particles’ morphology and core-shell morphology, have not been deeply studied, which have the potential to improve the Li+ diffusion in the interface of cathode and solid-state electrolytes. Additionally, due to various types of solid-state electrolytes, more studies are necessary to pair different morphologies of SC and PC particles and different solid-state electrolytes.

Secondly, to achieve the target of net zero, the sustainability of cathode materials needs to be considered for reducing the cost, energy cost, and greenhouse gas emissions in the recycling processes. After cycling, the morphology of the cathode materials could be damaged, which affects the performance, quality, and consistency of electrodes, and its recovery of the morphology is often overlooked despite its importance. Different morphologies pose challenges in recycling: PC TLOCs struggle to maintain their original morphology due to cracks, while SC TLOCs are easier to recover. Based on present technologies, direct recycling, and hydrometallurgical processes offer solutions127,128,129,130 but each has its limitations. Direct recycling can easily recover the electrochemical performance as same as the commercial cathode materials by recovering Li content and structure with the hydrothermal131,132, solid-state reaction133,134, ionothermal lithiation135, and electrochemical relithiation processes136, but the morphology destroyed by large cracks is not able to be recovered back to perfect spheres137,138. The recovered cathode exhibits an abundance of small particles (less than 5um) with irregular shape139 or totally irregular shape of recovered partilces140, which may lead to a lower tap density and casting problems in electrode manufacturing. Meanwhile, the heat treatment of direct recycling may lead to larger primary particles, even changing shapes of primary particles, resulting in worse capacity and stability. Thus, we believe that complex morphologies, such as spheres, hollows, core shells, and so on, are better suited for hydrometallurgical processes due to the resynthesis process included in this type of recycling process. Through the hydrometallurgical process, all defects in spent cathode materials can be recovered. Moreover, the recycled materials enabled better performance than the commercial cathodes113,141.

On the other hand, SC morphology fits both the direct and hydrometallurgical recycling processes because of its simple morphology. SC TLOCs can be simply recovered by direct recycling because SC morphology is easier to maintain than PC morphology during cycling128. Although the SC particles will have cracks or break into small particles as the PC TLOCs after cycling, the particle size and facets can be recovered by controlling sintering temperature, time, and different lithium salts. Also, the hydrometallurgical recycling process can recover the SC TLOCs, but the process is more complicated compared to the direct recycling process.

Finally, although there has been a growing interest in exploring new morphologies for cathode materials, such as SC morphology, the academic and industrial sectors still predominantly favor traditional spherical PC particles. This preference can be attributed to the well-established synthesis methods and electrode fabrication processes for PC materials, which have reached a high level of maturity. The hesitance to adopt advanced morphologies can be attributed to several factors, including the complexity of manufacturing processes, limited repeatability and consistency, and challenges associated with scaling up production. In light of these considerations, we would suggest that future endeavors in morphological design should concentrate on modifying the morphology of primary particles, achieving full gradient structures, and exploring SC morphologies. These morphologies can be potentially obtained through the optimization of coprecipitation reactions, a widely utilized technique in industry. To facilitate progress in simplifying manufacturing processes, it is imperative for researchers to delve into the formation mechanisms underlying these advanced morphologies. By gaining a deeper understanding of the factors influencing the formation of advanced morphologies, it becomes possible to devise innovative approaches that enable their reliable and scalable production. This, in turn, would pave the way for the adoption of advanced morphologies in practical applications, thereby leveraging their potential benefits for enhancing the performance of lithium-ion batteries.