Recent progress in synthesis and surface modification of nickel-rich layered oxide cathode materials for lithium-ion batteries

Nickel-rich layered oxides have been identified as the most promising commercial cathode materials for lithium-ion batteries (LIBs) for their high theoretical specific capacity. However, the poor cycling stability of nickel-rich cathode materials is one of the major barriers for the large-scale usage of LIBs. The existing obstructions that suppress the capacity degradation of nickel-rich cathode materials are as a result of phase transition, mechanical instability, intergranular cracks, side reaction, oxygen loss, and thermal instability during cycling. Core–shell structures, oxidating precursors, electrolyte additives, doping/coating and synthesizing single crystals have been identified as effective methods to improve cycling stability of nickel-rich cathode materials. Herein, recent progress of surface modification, e.g. coating and doping, in nickel-rich cathode materials are summarized based on Periodic table to provide a clear understanding. Electrochemical performances and mechanisms of modified structure are discussed in detail. It is hoped that an overview of synthesis and surface modification can be presented and a perspective of nickel-rich materials in LIBs can be given.

Double-metal co-doping is also investigated to improve stability of LiNiO 2 . Enlightened by improvement performance of LiNi 0.8 Fe x Al 1−x O 2 (x ⩽ 0.2) [43,44], LiNi 0.90 Co 0.07 Mg 0.03 O 2 [45], Li[Ni 0.9 Co 0.09 W 0.01 ]O 2 [46] and LiNi 0.92 Co 0.08 O 2 -Zr 2.0% [47], the Li[Ni x Co y Mn 1−x−y ]O 2 (NCM) and Li[Ni x Co y Al 1−x−y ]O 2 (NCA) were developed, and they are easy synthesis and high-energy density [5,6]. There is high initial discharge capacity with increasing Ni content in nickel-rich cathodes, however, its mechanical stability, thermal stability and air stability are substantially reduced during cycling. To realize their full potential, strategies including control of the particle and crystallite microstructures, intuitive bulk and surface stabilization are often applied [48][49][50]. Among them, core-shell structure has been considered as ideal method for suppressing the lattice instability of NCM/NCA during cycling [51,52]. Otherwise, the doping of foreign element endows the NCM/NCA a reasonable electronic transport pathway [53,54]. Huang et al prepared LiNi 0.8 Co 0.1 Mn 0.1 O 2 microspheres through rapid co-precipitation and spray drying method. Benefiting from the optimized particle structure, the 94.7% of its initial capacity was achieved after 200 cycles at 1.0 • C in a NCM811/graphite full cell [55]. A core-shell structured Li [59].
Similarly, many nickel-rich cathode materials also suffer from severe lattice change induced by the surface phase transition and side reactions with electrolyte, and many efforts have been initiated to restrain them. For instance, Zhang et al constructed LiNi 0.8 Co 0.1 Mn 0.1 O 2 primary particles agglomerated with tight grain boundaries by a hydrothermal method [60]. At 18.0 mA g −1 rate, NCM-750 • C delivered an initial discharge capacity of 203.7 mAh g −1 , 93.25% and 86.19% retentions were retained after 100th and 200th cycles at 180 mA g −1 , respectively. Yan et al constituted Li 3 PO 4infused LiNi 0.76 Co 0.14 Mn 0.10 O 2 within the surface and grain boundaries of secondary particles [61]. When cycled at 200/3 mA g −1 after three formation cycles 20.0 mA g −1 , the composite maintained 91.6% retention after 200 cycles. In one word, increasing attentions have been committed to improve performance of nickel-rich cathodes for LIBs and promising materials have been developed.
Many review articles focus on the particles structure optimizing, capacity fading mechanisms [62,63], and stabilization techniques for their development prospects [3,64,65]. In this review, recent remarkable research of nickel-rich cathode materials in the synthesis, surface modification, characteristics, degeneration mechanisms, strategies to enhance their stability for commercialization [66][67][68], as well as particles optimizing, e.g. the fabrication of single crystals [69] are introduced in details. Specifically for the surface modification methods and mechanisms for improving the nickel rich cathode materials performance, including doping or/and coating strategies et al, are summarized based on Periodic table. This review not only presents clear discussion about previous studies, but mirrors future development of nickel-rich cathode materials.
Although nickel-rich cathode materials possess a series of outstanding properties, their challenges must be addressed. Firstly, anisotropic volume variations caused by phase transition destabilize the internal microcracks, and these microcracks propagate to the surface that provided channels for electrolyte penetration and subsequent degradation [79][80][81][82][83]. A large number of Ni 2+ ions are occupied at Li + during the discharge process which contributes to the large Li + /Ni 2+ mixing [84][85][86], anisotropic volume change and subsequent phase transition on the particle surface as well as microcrack propagation in the bulk [87]. Capacity retention is closely related to the extent of microcracking within the secondary particles; primary particles within the secondary particles are exposed to electrolyte attack and the structural damage is accelerated [88]. Especially the mechanical integrity of the secondary particles is undermined with increasing cutoff voltage [89,90], and the range and limit of the depth of discharge (DOD) [91,92]. A kinetic barrier will be established with rapid degradation of the internal exposed surfaces and increased surface impedance on the primary particle level [93,94]. For this issue, doping external elements is also a critical method to hinder the Li + /Ni 2+ mixing of nickel-rich materials, such as Al doping [95,96]. Second, side reactions between electrode and electrolyte induce transition metal dissolution, originating from the hydrogen fluoride (HF) initially presented in decomposition of the LiPF 6 salt during the charging/discharging process [97][98][99]. The dissolved transition metal ions are deposited on the anode side forming the organic and inorganic solid electrolyte interphase (SEI) layer [100][101][102], and consequently deteriorating both cathode and anode integrity [103,104]. To overcome this challenge, many studies have shown that precise treatment of the surface with a robust coating layer can effectively suppress side reactions against continuous cathode degradation [105,106]. Investigations reveal that TiO 2 [107,108], Li 3 PO 4 , and a pyrazinelinked 2D sheet [109], etc, as a protective layer can prevent the nickel-rich materials from direct contact with the electrolyte, and thus suppress unwanted side reactions [47]. Additionally, electrolyte additive such as diethyl phenylphosphonite can improve cycling stability of nickel-rich cathode through shielding HF and constructing a protective interphase [99,110]. Third, unwished side reactions are often exothermal, causing complicated electro-chemo-mechanical effect at specific temperatures [111]. Thermal behavior of NMC-811 during charging was performed by in-situ heating x-ray diffraction (XRD), surface-sensitive x-ray absorbtion spectra (XAS) and bulk sensitive x-ray Raman spectroscopy (XRS) data [112]. The results indicate that bulk phase transition occurs at temperatures above 150 • C for 75% Li + -deintercalation samples or 250 • C for 50% Li + -deintercalation samples, subtle changes involving evolution of oxygen and metal redox can occur at temperatures well below the detectable point of phase transition, and thermal decomposition starts at particles surface and diffuses into the bulk, as well as small particles have much less thermal stability than large particles. Surface reaction and structural transition (e.g. oxygen loss, reconstruction, metaldissolution) occur quickly and deeply at high temperatures. Even at lower temperatures, the degradation occurs rapidly and eventually matches the degradation at high temperatures [113]. Thermal stability of LiNi x Mn y Co z O 2 (NCM433, 532, 622, 811) during charging was investigated by in-situ XRD and mass spectroscopy techniques [114]. The results reveal that there is more oxygen release and lower initial temperature of phase transition with increasing of Ni contents. The effect between Li + vacancies and Ni 2+ raised Ni 3+ during Li + -deintercalation process can intensify oxygen loss, and form more NiO compounds in local coordination structure unit [115]. It has been reported that the allaying effect of surface reactivity is effective in improving thermal stability, such as sophisticated synthesis methods that reduce the amount of Ni on particle surfaces [6]. Suitable surface modification is an effective method to form strong metal-oxygen bonding. Development of these strategies require deep understanding of the mechanism about metal-oxygen interaction. Finally, gas generation caused by parasitic side effects is one of the fundamental problems of the most advanced LIBs, because bubbles can block parts of the electrode surface, blocking the transport of lithium ion and resulting in uneven distribution of current. During the production, storage, and application of nickel-rich cathode materials, Li 2 CO 3 decomposition was induced by the electrochemical oxidation of the, resulting in the increase in the Li + /Ni 2+ disorder and polarization resistance [116][117][118]. In addition, Li 2 CO 3 decomposition contributed to CO 2 evolution, especially during the first charging [119]. At the same time, the greater contribution of electrolyte decomposition was indicated by the large amount of 12 CO 2 observed [120]. Highly reactive singlet oxygen is released when charging NCM111 and NCM811 to a state-of-charge beyond 80%. On-line mass spectrometry reveals that the evolution of CO and CO 2 occurs once singlet oxygen is detected, providing significant evidence for the reaction between singlet oxygen and electrolyte to be a chemical reaction [121]. Otherwise, the operando detection shows that surface reconstruction (e.g. LiMO 2 → MO + 1 /2O 2 ) is a main contributor to electrolyte decomposition based on the correlation between CO 2 and O 2 release [122]. In order to resolve the drawbacks, a cycling operation of a 3.0-4.5 V galvanostatic at initial cycles has been identified as an effective method to regenerate the subsequent 3.0-4.3 V battery performances [116,123]. Meanwhile, the irreversible oxygen loss on the surface of the particles leads to the deterioration of the surface structure and ultimately to the degradation of the cycling performance. Therefore, inhibiting the irreversible oxygen loss of these compounds under high voltage is an important way to develop high-capacity nickelrich layered oxide cathode materials.

Co-precipitation.
Co-precipitation is the most frequently used method for nickel-rich layered oxide cathode materials. As shown in figure 1(A), reaction process is conducted in an environmental of Ar, aqueous solutions of NiSO 4 · 6H 2 O, CoSO 4 · 7H 2 O, MnSO 4 · H 2 O (molar ratio = 8:1:1) are pumped into the reactor together with definite concentration of NaOH and NH 3 • H 2 O solutions. The mixture solution is stirred at a speed of 1000 rpm and the reaction temperature is controlled at 50 • C, the pH value is carefully monitored at 11 ± 0.2. After reaction for about 12 h, the obtained Ni 0.8 Co 0.1 Mn 0.1 (OH) 2 precursors are washed for several times and dried in an oven overnight [55,124,125]. As shown in figure 1(B), scanning electron microscopy (SEM) images show that Ni 0.8 Co 0.1 Mn 0.1 (OH) 2 primary particles stack inside the denser surface of spherical secondary particles with diameter of ∼8 µm [124]. Porous Ni 0.8 Co 0.1 Mn 0.1 (OH) 2 secondary particles (10 µm) are also obtained through controlling the flow speed of argon at 300 ml min −1 [126]. Whereafter, the resulted precursors and LiOH · H 2 O are mixed, the mixture is preheated first and subsequently annealed at 750 • C in a pure O 2 condition to obtain the spherical LiNi 0.8 Co 0.1 Mn 0.1 O 2 (NCM811) powders [125]. By comparison of the aqueous solutions used in the co-precipitation process, e.g. sulfate, acetate and nitrate, it is found that sulfate materials formed the best layered LiNi 0.6 Co 0.2 Mn 0.2 O 2 structure, showing the best cycling performance [127]. Gradual variation of pH value from 12.0 to 10.6 during the traditional co-precipitation has been performed to a prepared hollow microsphere of Ni 0.6 Co 0.2 Mn 0.2 (OH) 2 [128]. The self-organization process in the hollow LiNi 0.6 Co 0.2 Mn 0.2 O 2 is induced by the Kirkendall effect, resulting in the 4 µm sized spherical precursors with a core-shell structure (nanoplate-like surface and fine nano-seed core region). Regulating the reaction time of LiOH and Ni 0.58 Co 0.25 Mn 0.17 (OH) 2 is beneficial for enhancing electrochemical performance of nickel-rich cathodes [129]. The samples are collected at three different reaction times (3, 13 and 32 h), SEM of LiNi 0.58 Co 0.25 Mn 0.17 O 2 shows that the average sidewall thicknesses are counted to be 0.08, 0.16 and 0.27 µm, and the average edge lengths are 0.17, 0.47 and 0.74 µm, respectively. The sample of 32 h reaction shows the best electrochemical performance. In addition, calcination temperature is also an important factor for reaction of LiOH and Ni 0.76 Mn 0.14 Co 0.10 (OH) 2 [130]. By increasing calcination temperatures ranging from 725 • C to 900 • C for 20 h in air, it was found that calcination temperature (⩾800 • C) resulted in a significant growth of primary particle size, as well as extended lithium-ion diffusion pathways and inferior rate performance. The preparation process of single-crystal Ni 0.83 Co 0.10 Mn 0.07 (OH) 2 and Ni 0.6 Co 0.2 Mn 0.2 (OH) 2 precursors is similar with that of traditional secondary particles [131,132]. The formation process of a single-crystal LiNi 0.6 Mn 0.2 Co 0.2 O 2 cathode material was measured by Qian et al, and the ex-situ SEM images showed evolution of particle morphology and size [133]. Single crystal LiNi 0.92 Co 0.06 Mn 0.01 Al 0.01 O 2 displays less aggregation, better homogeneity and a smaller size (∼4 µm) compared with that of the secondary particles (∼10 µm) [134]. Furthermore, cracks, nanopores and other defects are great suppressed in these single-crystal nickel rich cathode materials.
Moreover, it is found that a core-shell structure is an effective method to improve electrochemical performance of nickel-rich cathodes, and the most studied core-shell materials are LiNi 0. 85 [138], and their synthesis process involves two-time co-precipitation. Thermaldriven degeneration frequently occurs during the sintering process [139], thus, an additional pre-oxidization process was employed to regulate the Li + /Ni 2+ cation ordering, various pre-oxidization reagents have been studied such as polyvinylpyrrolidone (PVP) [140] and Na 2 S 2 O 8 [141]. Tang et al prepared surface-oxidized LiNi 0.815 Co 0.15 Al 0.035 O 2 via dissolving NCA precursors in mixed aqueous solution containing Na 2 S 2 O 8 and NaOH [142]. TEM analysis indicated that there is an well-organized β-NiOOH layer (4-11 nm), whereas a higher peak intensity ratio of (001)/(101) was found on the precursors surface. After 100 cycles at 90 mA g −1 within 2.75-4.5 V, pre-oxidated NCA delivered 80.3% of its initial capacity in contrast to 67   As shown in figure 1(C), rapid co-precipitation was observed when aqueous solution of ammonia and NaOH were pre-added into the reactor, afterwards, aqueous solutions of metal sulfates, NaOH and NH 4 OH were continuously infused into reactivator, and this process lasted for 1 h. After filtering and washing, the precursor was dried under vacuum at 110 • C for 12 h. Huang et al prepared the Ni 0.8 Co 0.1 Mn 0.1 (OH) 2 precursors by this rapid co-precipitation, and spherical secondary particles (NCM811) have a diameter ranging from 2 to 50 µm ( figure 1(D)). For comparison, some precipitates were ball milled for 3 h, and the obtained powders were spray dried for 5 h (spray dried NCM811 (SD-NCM), figure 1(E)), with an average particle diameter of NCM811 was ∼12 µm as shown in figure 1(F) [55]. The SD-NCM displays superior cycling stability under an upper cut-off voltage of 4.5 V. Extensive studies about the facile co-precipitation approach remain interesting, for example, precursor was prepared through dissolving chlorate in absolute ethanol [146]. Then, the mixture was mixed with ammonia aqueous solutions following vigorous stirring by peristaltic pump for 2 h. The precipitations were orderly separated, washed, dried and annealed to obtain spheroidal precursors. Well-ordered NCM811 layered oxides have been obtained by calcination with the precursors mixed with Li 2 CO 3 at 800 • C, and the NCM811 powders exhibited suppressed capacity fading. Carbonate coprecipitation has been investigated to improve the electrochemical property of secondary particles, such as core-shell structured Ni 0. 8 [78]. To form an homogeneous dispersoid, the solution was stirred vigorously at 80 • C for 12 h, and then the mixture was dried and calcined in flowing oxygen. The verage diameter of the porous microspheres was ∼3 µm. Ma et al synthesized MnCo(CO 3 ) 2 precursors by a similar carbonate co-precipitation using NH 4 HCO 3 as a chelating agent, and the obtained secondary spherical particles had an average diameter of 1.5-4 µm [148]. Similarly, (Ni 0.8 Co 0.15 Al 0.05 )(CO 3 ) x (OH) 2-2x precursors were synthesized via a carbonate co-precipitation route using NaOH and (NH 4 ) 2 CO 3 as a chelating agent [149]. This process has a faster particle growth rate than hydroxide process using NH 4 OH as a chelating agent, and the fabricated spherical precursors have an average diameter of 10 µm. Liu [151]. Firstly, spherical Ni 0.7 Co 0.15 Mn 0.15 (OH) 2 particles were obtained via a co-precipitation method. Then, the core-shell structured composition was obtained via a solid state ration method using Ni 0.7 Co 0.15 Mn 0.15 (OH) 2 precursors, and MnCO 3 and LiOH · H 2 O as raw materials. All samples exhibited a spherical or elliptical shape and a uniform size distributed with a diameter of around 9 µm. Moreover, Zhou et al synthesized the Ni 0.75 Co 0.15 Al 0.1 (OH) 2.1 precursors via an improved co-precipitation using nano Al 2 O 3 particles as the precipitate nucleus [152]. Perfect spherical LiNi 0.75 Co 0.15 Al 0.1 O 2 powders with in diameter of about 10 µm were obtained, and they were composed of primary particles with a diameter of about 1 µm.

Hydrothermal method.
Hydrothermal methods have been used to synthesize nickel rich cathode materials. As shown in figure 1(G), core-shell structured LiNi 0.8 Co 0.1 Mn 0.1 O 2 was prepared via a combination of coprecipitation and hydrothermal methods [153]. The oxalate precursors Ni 0. 8 2 precursors were prepared through a solvothermal method [155]. Precursors were composed of flowerlike microspheres with a dimension around 2-6 µm. Tian et al prepared Ni 0.7 Co 0.15 Mn 0.15 (OH) 2 precursors via a hydrothermal treatment after co-precipitation reaction [156]. As shown in figure 1(I), the LiNi 0.7 Co 0.15 Mn 0.15 O 2 particles were nanoplate-like, which have an average diameter around 300 nm. To investigate the effect of concentration, temperature, and reaction time, Ni 0.5 Co 0.2 Mn 0.3 CO 3 precursors were prepared by controlling reaction conditions in the hydrothermal method [157]. SEM images showed that spherical particles were composed of primary grains (200-500 nm). When reaction time was 24 h, LiNi 0.5 Co 0.2 Mn 0.3 O 2 showed the best cyclability and rate capability. Moreover, Zuo et al prepared Ni 0.5 Co 0.2 Mn 0.3 CO 3 precursors by a hydrothermal method [158]. As shown in figure 1(J), the rugby ball-like morphology was the most regular when the molar ratio of urea to the total transition metal cations was set as 5:1.  [159], and then calcined in oxygen and air, respectively. Samples in oxygen were constituted by irregular polyhedral primary particles (200-500 nm) (figure 1(K)) and showed better electrochemical performance. In order to tailor the Li 2 CO 3 surface phase, oxidization extent of Ni 2+ /Ni 3+ was controlled and oxygen vacancies on LiNi 0.9 Mn 0.1 O 2 surface were modulated. Synthesis was accomplished via a sol-gel method and then gel was calcined in air [160]. LiNi 0.9 Mn 0.1 O 2 secondary spheres (10 µm) weere constituted by primary particles with diameters ranging from 200 to 500 nm (figure 1(L)). However, because of the coating effect, the Li 2 CO 3 surface phase can not only suppress the H 2 /H 3 phase transition but also alleviate the side ration of LiNi 0.9 Mn 0.1 O 2 with electrolyte, and eventually lead to a good cycling stability of LiNi 0. 9 [162]. Particle size increased with raised calcination temperature (750 • C-950 • C). As shown in figures 1(N) and (O), NCM811-900 • C with D50 for 7.7 µm displays the best discharge specific capacity and cyclability.

Other methods.
A simple combustion has also been used to synthesize LiNi 0.6 Co 0.2 Mn 0.2 O 2 with a nitrate/urea mixture as fuel [77]. The fabricated samples exhibit a spherical morphology. While, Li et al prepared Ni 1−2x Co x Mn x O y (x = 0.075, 0.05, 0.025) precursors by ultrasonic spray pyrolysis of metal chlorides solution [163]. The spherical precursors were composed of numerous closely packed primary particles and the particle sizes increase with the increase of Ni content. A hydrometallurgical method has also been proposted to treat high-grade nickel matte and produce LiNi 0.80 Co 0.15 Al 0.05 O 2 cathode materials. This strategy includes oxidation roasting, two-stage leaching, solvent extraction, spray pyrolysis and solid sintering [164], abundant micro pores are observed in the products. Moreover, Zhu et al prepared single-crystal LiNi 0.8 Co 0.1 Mn 0.1 O 2 by a spray pyrolysis method, field emission scanning electron microscope (FESEM) images show that their precursors are composed of regular microspheres with diameter of 1-5 µm [165].
In summary, numerous efforts have been devoted on these synthetic methods and great improvements have been obtained for these nickel rich cathode materials for LIBs. Their electrochemistry performance has been summarized in table 1.

Characterizations
Many characterization techniques are used to confirm the structure and morphology of nickel-rich cathode materials and investigate their structural evolution during charge-discharge process [166,167]. Firstly, ex-situ and in-situ characterization techniques are used to show phase transition of nickel-rich cathodes during charge/discharge process. Typical ex-situ measurements include XRD [168], focused ion beam (FIB) cutting-SEM and transmission electron microscopy (TEM)fast Fourier transformed (FFT) [169], time-of-flight secondary ion mass spectrometry (ToF-SIMS), inductive coupled plasma emission spectrometer-atomic emission spectroscopy (ICP-AES), energy-dispersive x-ray (EDX) analysis, x-ray photoelectron spectroscopy (XPS), and Fourier transform (FT) infrared spectroscopy. To confirm the crystal structure of the nickel-rich cathode materials, ex-situ XRD analysis was used to investigate the peak phase [170]. Bulk compositions and surface morphologies of nickel-rich cathode materials can be initially analyzed by SEM and energy dispersive spectrometer (EDS) elemental mapping/ICP-AES/ToF-SIMS [171,172]. To further explore the surface compositions and oxidation states of the nickel-rich cathode materials, XPS examination is an ideal method to analyze the pristine elements and generated byproducts [90,170]. In addition, formation of intergranular cracks and void spaces within surface/interface structure can be evident through cross-sectional SEM images of nickel-rich cathode materials before and after cycling [51]. To analyze the surface phase transformation and oxygen reversibility of the cathode materials, high-resolution TEM with FFT, scanning TEM (STEM) and electron energy-loss spectroscopy (EELS) were used to analyze the structural evolution mechanisms of the cycled cathode materials by showing the atomic surface structure and EELS line scans of the O K edge [173,174]. The phase transition of a cycled LiNi 0.90 Co 0.09 Ta 0.01 O 2 (NCTa90) cathode was analyzed via high-resolution TEM image with FT [175]. The results indicate that Li + -deintercalation NCTa90 cathodes exhibit an exceptional cycling stability, since cation ordering is preserved even after 2000 cycles. However, disassembling the batteries to obtain the cycled nickel rich cathode materials at different states may cause some errors due to the phase transition of the samples during the disassembling process [176]. Recently, in-situ XRD has been considered as a powerful technic in studying the real-time microstructural evolution of nickel-rich cathode materials in LIBs. It is beneficial to observe the tiny phase changes that are usually invisible from ex-situ XRD. Liu  Because the discharge processes (Li + -intercalation) of cathode materials are concomitant with reduction processes of TM ions involving Li + re-intercalation into the nickelrich cathode materials lattice, these dopants from group I and II A (Na, K, Mg, Ca) may impede to Li + reinsertion, due to their possible incorporation in the Li sites and formation a Capacity retention (%)/cycle number/current density (mA g −1 )/cut-off voltage (V)/cycle temperature ( • C), room temperature at 25 • C. b Capacity (mAh g −1 )/current density (mA g −1 ).
of surface layers. However, cycling stability of the modified materials can be improved due to stability of bulk and surface structures of the surface modified nickel rich cathode materials. For example, Na-doped LiNi 0.5 Co 0.2 Mn 0.3 O 2 (NCM523) was synthesized by a hydroxide co-precipitation route [178].
The substitution of Na for Li resulted in a more ordered α-NaFeO 2 structure, accompanied by the slight enlargement of the layered distance and lower cation mixing, which was attributed to the pillaring effect of Na ions with large radius. As a result, Li 0. 97  and 10 000 mA g −1 , which were superior to those of the bare NCM523 due to the rapid diffusion of Li-ion in the bulk lattice. Improved reversibility and electrochemical activity, low potential polarization after the substitution of Na for Li were further verified by stability of both potential and capacity as well as the integrated spherical morphology upon long-term cycling. In order to stabilize both capacity and potential, gradient Na-doped LiNi 0.8 Co 0.15 Al 0.05 O 2 (NCA) was prepared by Wang et al [179]. Na-doped and un-doped NCA delivered specific capacities of 150 and 89 mAh g −1 after 300 cycles at a current density of 200 mA g −1 , corresponding to capacity retention of 81.6% and 48.1%, respectively. Even at a high current rate of 1000 mA g −1 , Na-doped NCA achieved a capacity of 135.5 mAh g −1 . After 300 cycles, Na-doped NCA exhibited only 0.24% decay of discharge midpoint potential in contrast to 4.32% for raw NCA. Only a few of the Nadoped NCA secondary particles broke into cracked particles (figures 2(A) and (B)), which was attributed to the pillaring effect of Na ions with large radius. Meanwhile, many reports provide more insight into the pillaring effect at lower cation mixing. For example, micronsized single crystals (Li 1−x K x Ni 0.8 Co 0.1 Mn 0.1 O 2 , LKNCM) were prepared by a molten salt-assisted growth method, which was beneficial for the formation of Ni 3+ and Co 3+ in single crystals [180]. LKNCM exhibited a capacity of 163 mAh g -1 after 200 cycles at 1.0 • C (capacity retention of 91%), which was much higher than that of Li 1−x NCM (LNCM) for 124 mAh g −1 (capacity retention of 72%). Even at a high current rate of 10.0 • C, LKNCM presented discharge capacity about 116 mAh g −1 and the median voltage of 3.4 V. Pristine LNCM presented a loose particleaggregation morphology after cycling, while LKNCM particles were still ordered and distinguishable.
It is worth noting that the inherent structural and thermal instability can be suppressed by using an optimized LiNi 0.90 Co 0.07 Mg 0.03 O 2 with greater conductivity [45]. In the voltage range of 2.8-4.5 V, LiNi 0.90 Co 0.07 Mg 0.03 O 2 demonstrated a capacity of 148.7 mAh g -1 at 360 mA g −1 after 300 cycles (84.3% capacity retention). Meanwhile, the LiNi 0.90 Co 0.07 Mg 0.03 O 2 could deliver a considerably high capacity of 142.8 mA h g −1 at a high current rate of 1800 mA g −1 . When tested at −5 • C and 60 • C, capacity retention of 81.2% and 76.0% was achieved after 200 cycles at 180 mA g −1 , respectively. As shown in in-situ XRD data, LiNi 0.90 Co 0.07 Mg 0.03 O 2 experienced a H1-H2-H3 phase transition at high voltage range, because Mg doping stabilized the layered structure by suppressing cation mixing. It should also be stated that Mg-doped NCA was synthesized through a solid state reaction [181]. Mg-NCA displayed a typical layered hexagonal structure without impurity phase, with an increased percentage of Ni 2+ on the surface. After 100 cycles at 92.5 mA g −1 , Mg-NCA delivered 84.9% of its initial capacity in contrast to 63.2% for the pristine NCA (PNCA). Electrical impedance spectroscopy (EIS) and galvanostatic intermittent titration technique tests confirmed that Mg-doped NCA showed a lower total resistance value after 100 cycles, and a higher Li + -ion diffusion coefficient compared to PNCA during charge/discharge process. Decoupling ultrathin MgO, Al 2 O 3 and ZrO 2 coating layer atomic layer deposition (ALD) have been reported by Laskar et al [182]. After 180 cycles at 1.0 • C, MgO-coated NCM cathode demonstrated a high capacity of 130 mAh g −1 with capacity retention of 82%, while the capacity retention of pristine NMC was 70%. Even at a high current rate of 10.0 • C, MgO-coated NCM cathode exhibited a capacity of 115 mAh g −1 , in contrast the Al 2 O 3 -coated and uncoated NMC exhibited a capacity of 45 and 30 mAh g −1 , respectively. As shown in figures 2(C)-(E), MgO-coated NCM cathode had a higher Li + -diffusion coefficient, which resulted in lower overpotential on the surface and improved rate performance.
, has also been investigated by Chen et al, because Ga has a lower valence state and larger radius compared to Ni 2+ [183]. In the voltage range of 2. B, as a typical element in IIIA group, can reduce the surface energy of the (003) planes of layered oxide cathode materials to produce a highly textured microstructure, because its ionic radius (B 3+ , 0.27 Å) is smaller than that of Li + (0.76 Å), Ni 2+ (0.69 Å), Co 3+ (0.545 Å) and Mn 4+ (0.53 Å). Some work has been done by Park et al to study the effects surface modification by using B 2 O 3 with an in-situ high-temperature calcination method [57]. XRD and TEM results indicated that prepared samples had a rhombohedral crystal structure with a R-3m space group and average lengths of 0.55-1.1 µm. B 1.0 -NCM90 with a highly textured microstructure delivered a capacity retention of 91% after 100 cycles at 90 mA g −1 and 55 • C, which was 15% higher than that of pristine-NCM90. Based on density functional theory calculation results, the higher capacity retention was attributed to the alleviation of intrinsic internal strain and the reduction of a small amount of Ni 3+ to Ni 2+ . In contrast, obvious microcracks were observed in bare NCM (figure 3(A)). Ryu [184]. TEM results indicated that elongated primary particles were ∼500 nm long and ∼100 nm wide. After 100 cycles at 90 mA g −1 and 45 • C, B-NCA88 exhibited a high-capacity retention of 88.7%, and that of pristine P-NCA89 is 74.4%. Even at a pouch-type full-cell with a graphite anode was cycled between 3.0 and 4.2 V at 180 mA g −1 , B-NCA88 retained a capacity retention of 83% after 1000 cycles, while the capacity retention of P-NCA89 was only 49.0%. As shown in figures 3(B) and (C), the cycled B-NCA88 cathode maintained its initial (003) contour plot well, even after repeated lattice contraction and expansion; particles were also free of microcracks. While, Li et al raised an interesting question regarding the state of B 2 O 3 coating in polycrystalline cathode materials [185]. B 2 O 3 -coated LiNi 0.83 Co 0.12 Mn 0.05 O 2 was synthesized through a mechano-chemical bonding technology, and SEM results indicated that B 2 O 3 with a molten state filled in the boundaries among the primary particles. The 0.5% B 2 O 3coated LiNi 0.83 Co 0.12 Mn 0.05 O 2 delivered a discharge capacity of 154.2 mAh g −1 after 200 cycles at 200 mA g −1 with a capacity retention of 87.7%, whereas the pristine sample delivered a capacity retention of 69.4%. The improved cycling stability was associated with a B 2 O 3 coating layer, which could greatly inhibit the irreversible phase transitions and the extension of microcracks. However, the B 2 O 3 coating layer acted as a resistance layer because of its electrically and electrochemically inactive properties. Therefore, lithium-ion conductors have been developed as alternative coating materials. Hashigami et al prepared lithium boron oxide (LBO)-coated LiNi 0.5 Co 0.2 Mn 0.3 O 2 by an antisolvent precipitation method [186]. XRD results indicated that the diffraction peak intensity ratio of the (003) to the (104) decreased by the coating of LBO. After 50 cycles at 200 mA g −1 within 2.5-4.5 V, the 5 wt.% LBO-NCM maintained 85.6% of its initial capacity, in contrast to 53.8% for the pristine NCM. Cross-sectional SEM images of cycled samples showed that crack formation was excluded from spray pyrolyzed particles (<1 µm), which was likely to be the reason why the LBO coating could improve the cycle ability and rate capability. Interestingly, in a more rigorous study on anion substitution, Li[Ni 0.6 Co 0.2 Mn 0.2 ]-(BO 3 ) x (BO 4 ) y O 2−3x−4y (x + y = 0, 0.01 and 0.03) were prepared by adding boric acid (H 3 BO 3 ) before the lithiation of the hydroxide precursors [187]. In the voltage range of 2.8-4.5 V, B 3 -NCM delivered 76.07% of the initial discharge capacity after 100 cycles at 180 mA g −1 , whereas the corresponding capacity retention of the pristine counterpart was only 59.15%. The improved performance was attributed to boracic polyanion doping, which could greatly enhance Li + diffusion coefficient in NCM.
Al 3+ doped nickel-rich compounds have attracted tremendous attention due to their high cycling stability derived from enhanced Li + kinetics and structural stability, although their initial discharge specific capacities were descended due to incorporation of inactive Al 3+ . As expected, studies on heterogeneous Al 3+ doped LiNi 0.7 Co 0.15 Mn 0. 15 O 2 show that it delivered a capacity retention of 90.6% after 100 cycles at 20 mA g −1 and 50 • C, while pristine NCM delivered a capacity retention of only 68.5% [188]. XRD results indicated that the lattice parameter a value decreased Reprinted from [191], Copyright (2019), with permission from Elsevier. (G) Cross-section SEM and high-angle annular dark-field (HAADF)-STEM pictures of NMC76 and 1% Al-NMC76 samples. Reprinted with permission from [192].  [56]. Magnified STEM images indicated that 9-10 µm spherical particles were composed of the rod-shaped primary particles (∼300 nm) with a strong crystallographic texture, which expedited Li intercalation and nano-sized core primary particles suppressed the propagation of interparticle microcracks (figure 3(E)).  [192]. The results indicated that the surface of doped NMC76 particles became rougher as Al doping increased, and Al species displayed a concentration gradient from the surface to the subsurface in the layered structure. In the voltage range of 2.7-4.5 V, 1% Al-doped NMC76 delivered a capacity retention of 79.2% after 500 cycles, whereas pristine NMC76 only retained a capacity retention of 3.4%. This bulk structural stability (figure 3(G)) was ascribed to Al doping into the NMC76 lattice, which suppressed chemical reactions with the acidic electrolyte. Kim et al prepared Al-doped LiNi 0.76 Co 0.09 Mn 0.15 O 2 with concentration gradients to suppress the microcracking (figure 3(H)) and preserve the mechanical integrity of the cathode particles [193]. Benefiting from a slight decrease in cation mixing, the 2 mol% Al doped LiNi 0.76 Co 0.09 Mn 0.15 O 2 retained 95% of its original capacity after 1000 cycles at 1.0 • C in pouch-type full cells with graphite anodes, whereas pristine and gradient NCM (FCG76) retained 80% and 88% of the initial capacity. As shown in in-situ time-resolved (TR)-XRD experiment (30 • C-600 • C) results, partially deintercalated of 2 mol% Al doped NCM transformed to a disordered spinel structure (Fd-3m) at higher temperatures (figure 3(I)). The protective layer formed on the surface of nickel-rich cathode materials can effectively inhibit the dissolution of Ni, Co, and Mn ions and improved thermal stability by suppressing direct contact between electrolyte and Li + -deintercalation nickel-rich cathode materials. Liao et al substantiated this by prepared Al 2 O 3 coated gradient LiNi 0.62 Co 0.14 Mn 0.24 O 2 , and it can deliver 89% of its initial discharge capacity after 50 cycles at 36 mA g −1 and 55 • C in the voltage range of 2.7-4.5 V, while the corresponding capacity retention of the pristine and gradient NCM were only 51% and 81%, respectively [194]. Li[Ni 0.73 Co 0.12 Mn 0.15 ]O 2 coated by Al 2 O 3 have positive effects for suppressing the microcracking and preserving its mechanical integrity [195]. This modification ultimately results in lower cation mixing and leads to a capacity retention of 95.0% after 1000 cycles at 1.0 • C in pouch-type full cells with a graphite anode, whereas pristine Li[Ni 0. 73 [197]. Rietveld refinement results illustrated that there is no obvious change for key lattice parameters a and c, suggesting that the Al 3+ existed as oxides on the LiNi 0.5 Co 0.2 Mn 0.3 O 2 surface rather than inserted into the lattice sites of the bulk materials. Benefiting from the prominent physical-chemical protection function of nanoscale Al 2 O 3 layer, pristine and the 2% Al 2 O 3encapsulated LiNi 0.5 Co 0.2 Mn 0.3 O 2 (A-O-2) delivered capacity retentions of 54% and 83% after 100 cycles at 1.0 • C and 55 • C within 3.0-4.6 V, respectively. In the voltage range of 3.0-4.35 V, the A-O-2 delivered a capacity retention of 80% after 800 cycles at 1.0 • C of pouch-type full-cells. Fourier filtered transforms and cross-sectional SEM images showed that no phase evolutions and cracks were found in the cycled A-O-2 sample, while the Al 2 O 3 still existed in the edge of particles, which was highly consistent with the long-term cycle test results. Nevertheless, the strategy of using Al 2 O 3 as a physical barrier coating is an effective way to overcome these instabilities in layered nickel-rich oxides [198]. As shown in Ni Kedge x-ray absorption near-edge spectroscopy data, the oxidation state of Ni in Al 2 O 3 coated Ni-600 • C electrode is much lower than that of a pristine sample at 4.3 V. Also, the Al 2 O 3 coated Ni-900 • C cathode can be continually oxidized more than at 4.3 V.
Ionic conductor LiAlO 2 deposited on the surface of nickelrich cathode materials can effectively improve the diffusion coefficient of Li + ions and suppress the undesirable side reactions. Besides that, the partial migration of Al 3+ to the host of nickel-rich materials also improved the structure stability. As  [199]. EDS mapping indicated that most of the Al 3+ existed in the coating layer on the surface of LiNi 0.8 Mn 0.1 Co 0.1 O 2 rather than incorporating into bulk LiNi 0.8 Mn 0.1 Co 0.1 O 2 . High resolution transmission electron microscope (HRTEM) images showed that the thickness of the coating layer was 5-7 nm for 1 wt.% LiAlO 2 coated LiNi 0.8 Mn 0.1 Co 0.1 O 2 . The structural stability of LiNi 0.8 Mn 0.1 Co 0.1 O 2 was enhanced by using LiAlO 2 coating, and it led to a capacity retention of 89.1% after 50 cycles at 280 mA g −1 within 2.5-4.5 V, which was higher than pristine NCM (82.5% capacity retention) [200]. XRD results indicated that the intensity ratio of I (003) /I (104) increased after LiAlO 2 surface modification. Electrical conductivities of pristine NCM and 5% LiAlO 2 coated NCM were 1.12 and 8.21 S m −1 , respectively. AlCl 3 solution has also been used as surface coating solvent, since the oriented deposition of Al(OH) 3 layer was formed on surface of hydroxide precursors [201]. HRTEM images showed that the thickness of the coating layer was about 5 nm for 2.2 wt.% LiAlO 2 coated NCM. Benefiting from lower Li + /Ni 2+ disorder, the 2.2 wt.% LiAlO 2 -coated NCM obtained a discharge capacity of 159.9 mAh g −1 with a capacity retention of 85.8% after 200 cycles at 0.5 • C, while pristine NCM delivered a discharge capacity of only 112.8 mAh g −1 . Even at a high current rate of 10.0 • C, a capacity of 135.2 mAh g −1 was also achieved for the 2.2 wt.% LiAlO 2 -coated NCM. The cycling capability of the materials was also demonstrated by a LiAlO 2 coated LiNi 0.90 Co 0.07 Al 0.03 O 2 cathode, which exhibited a capacity retention of 93.4% at 1.0 • C after 100 cycles, while that of pristine LiNi 0.90 Co 0.07 Al 0.03 O 2 80.8% [202].
XRD and HRTEM results indicated that a γ-NaFeO 2 type structured LiAlO 2 on the surface could be observed, with a thickness of about 10 nm.

The group IVA, metal oxides (MOx, M = Si, Sn)
Carbon is low-cost with high electrical conductivity, and carbon coating on nickel rich cathode materials can resist acute side reactions between Ni 4+ and carbonate electrolyte. LiNi 0.8 Co 0.15 Al 0.05 O 2 (NCA) as the core and carbon composite as shell cathode material were synthesized by a solvent-free mechano-fusion method [203]. Discharge capacity of PNCA was 110 mAh g −1 at 100 mA g −1 after 250 cycles and the capacity retention was 71%, while discharge capacity of modified NCA was 150 mAh g −1 (capacity retention of 84%). In particular, surface modified NCA exhibited a discharge capacity of 139 mAh g −1 even at 800 mA g −1 , while discharge capacity of bare NCA was only 106 mAh g −1 . The core-shell structured materials experienced a much lower Li + diffusion barrier as compared to PNCA materials owing to slight expansion of the inter-slab distance along the c-axis (figure 4(A)). Volumetric change along the grain boundaries is also inevitable in cathode materials. Accommodated 3D graphene aerogel network aerogel was first introduced in 2019 for LiNi 0.6 Co 0.2 Mn 0.2 O 2 by a modified Hummers' method [204]. When tested at 200 mA g −1 , a capacity retention of 92.69% was obtained after 250 cycles. Particularly, the composites exhibited an outstanding rate performance of 130.9 and 106.8 mAh g −1 at 1000 and 2000 mA g −1 , respectively. It was beneficial from the rapid Li + and electrons transportation mobility during the Li + -intercalation/deintercalation process. Son et al discovered a facile chemical vapor deposition process with CO 2 and CH 4 for LiNi 0.6 Co 0.1 Mn 0.3 O 2 , and yielded a conductive and protective artificial SEI layer consisting of alkyl lithium carbonate (LiCO 3 R) and Li 2 CO 3 [205]. With cut-off voltage of 4.4, 4.5, and 4.6 V, CO 2 + CH 4 − LiNi 0.6 Co 0.1 Mn 0.3 O 2 maintained capacity retentions of 98.55%, 96.5%, and 95.1% at 90-100 mA g −1 after 100 cycles, respectively. With the same voltage ranges, lower retentions of 93.8%, 92.7%, and 91.1% were showed for pristine LiNi 0.6 Co 0.1 Mn 0.3 O 2 , respectively. The high electrochemical performance of CO 2 + CH 4 − LiNi 0.6 Co 0.1 Mn 0.3 O 2 was attributed to the surface coating effect toward mitigated side reactions, following suppression of oxygen evolution and consequent structural instability ( figure 4(B)). Similar to CO 2 + CH 4 , N 2 + CH 4 is also helpful for the modification of LiNi 0.6 Co 0.1 Mn 0.3 O 2 , which yielded a conductive and protective artificial SEI layer consisting of amorphous carbon, alkyl lithium carbonate, and lithium carbonate (Li 2 CO 3 ) [206]. As shown in cross-sectional TEM and EDS analysis results (figure 4(C)), both the surface and internal voids were throughout coated during the coating process when using CH 4  structure was also modified by reduced graphene oxide (rGO) with electrostatic interactions and forming a ∼2 nm thickness coating layer [207]. Because of its interfacial stability, the obtained rGO-LiNi 0.6 Co 0.2 Mn 0.2 O 2 maintained a reversibility of 179.9 mAh g −1 at 1.0 • C over 100 cycles in a potential range of 3.0-4.5 V without inner phase transformation and structural degradation (figures 4(D) and (E)). Following this concept of rGO wrapping, high-quality rGO-encapsulated LiNi 0.815 Co 0.15 Al 0.035 O 2 was prepared to improve its Listorage performances and thermal stability [208]. In addition, forming a Si-based coating layer, on nickel rich cathode materials surface has also been considered as a promising strategy to improve its electrochemical performance, e.g. SiO 2 [209], ultrathin Si-O film [210], Li-Si-O [211]. The effects on the crystal structure stability and electrochemical property vary for silicon compositions. For example, SiO 2 has been used to suppress the growth of interfacial impedance due to a scavenging effect for the HF. A SiO 2 coating layer (thickness <10 nm) was designed on LiNi 0.6 Co 0.2 Mn 0.2 O 2 [209]; the 3.0 wt.% SiO 2 -coated LiNi 0.6 Co 0.2 Mn 0.2 O 2 exhibited a discharge capacity of 164.7 mAh g −1 even at 0.5 • C and 60 • C after 50 cycles, while that of pristine NCM was 144.7 mAh g −1 . Meanwhile, a molecular coating strategy easily forms a ultrathin Si-O film on LiNi 0.8 Co 0.1 Mn 0.1 O 2 [210]. The lithium silicate acted as a sintering agent on calcination, and resulting in a high-density grain boundary.
XRD results indicated that Si was usually coordinated to oxygen in a tetrahedral configuration, and TEM images indicated that the Si-O film could exist in a monolayer or a very few layers. After 300 cycles at 200 mA g −1 , the electrode with Si-O film surface-modified LiNi 0.8 Co 0.1 Mn 0.1 O 2 delivered 87.5% of its initial capacity, while that of pristine LiNi 0.8 Co 0.1 Mn 0.1 O 2 was 62.57%. A reversible capacity of 155.8 mAh g −1 was also achieved even at a high current rate of 400 mA g −1 . The characteristic peaks of electrodes obtained after 300 cycles at 1.0 • C remained a typical layered α-NaFeO 2 structure. The Li + diffusion coefficients increased from 1.22 × 10 −11 cm 2 s −1 to 2.80 × 10 −11 cm 2 s −1 . It was reported that the formation of lithium silicate coating layer is effective for preventing crack formation and decomposition of electrolyte [212]. Cross-sectional TEM analysis indicated that Si-rich regions were concentrated not only on the particles surface, but also at the grain boundaries even at a depth of 400 nm. In the voltage range of 2.5-4.6 V, capacity retentions of 71.4% and 81.6% were obtained after 100 cycles for pristine and lithium silicate-coated LiNi 0.5 Co 0.2 Mn 0.3 O 2 , respectively. It revealed that in-situ Li-Si-O coating can enhance structural reversibility of LiNi 0.8 Co 0.1 Mn 0.1 O 2 , where density of secondary particles was increased and pores between primary particles was greatly reduced [211]. TEM-EDX elemental maps showed that Si was localized at the grain boundaries and near the center of the particle. In the voltage range of 3.0-4.5 V, the 1 wt.% Li-Si-O-NCM811 delivered a discharge capacity of 183 mAh g −1 after 100 cycles at 20 mA g −1 (capacity retention ratio of 85.9%), whereas the corresponding capacity retention of the pristine NCM811 was only 79.0%. Cross-sectional SEM images showed that visible microintergranular cracks were observed for pristine NCM811, while the 1 wt.% Li-Si-O-NCM811 particle were almost free from visible microcracks after 100 cycles (figure 4(F)).
Recently, it was found that Sn-modification can obviously enhance the cycle behavior of nickel-rich cathode materials.For example, the cyclic stability of LiNi 0.5 Co 0.2 Mn 0.3 O 2 was greatly improved at high voltage with SnO 2 surfacemodification [213]. XRD results indicated that the intensity of the (104) diffraction peak reduced, and the value of I (003) /I (104) ratio was significantly enhanced. In the voltage range of 2.8-4.4 V, the 3 wt.% SnO 2 modified LiNi 0.5 Co 0.2 Mn 0.3 O 2 exhibited a capacity retention of 92.3% after 150 cycles at 0.5 • C, whereas the corresponding capacity retention of the other milled control materials and a commercial LiNi 0.5 Co 0.2 Mn 0.3 O 2 electrode were only 77.4% and 87.1%, respectively. The improved performance was attributed to the synergistic effect of high energy plasma and milling-induced oxygen vacancies in the SnO 2−x surface protection layers, which enabled greatly increased conductivity of the active materials and stable interfaces. Furthermore, the complex relationship between the Li-excess Li 1 The amounts of residual lithium compounds (e.g. Li 2 O, LiOH, Li 2 CO 3 ) increase with increasing Ni content in nickel-rich cathode materials, those residual lithium compounds hinder Li + transport, which in turn causes resistance increased. Li 3 PO 4 coating can partly inhibit the reaction of active material with the air because it is stable up to 6.0 V versus Li/Li + , and has excellent ionic conductivity. Also, PO 4 3− anions react with those residual lithium compounds. Thus, strong covalent bonding of the PO 4 3− contributes to improved high-voltage performance and thermal stability of nickel-rich cathode materials. Jo [221], and it was determined that the use of the coated NCA in aqueous electrode processing reduced the alkalinity of the slurry and thus the lithium leaching in water could be at least partially sup-

The group VI A, metal sulfides/selenides/tellurides
Due to the sublimated gas-phase, S can react with detrimental residual Li compounds on the surface of the particles. Sulfur-coated LiNi 0.8 Co 0.1 Mn 0.1 O 2 (SNCM) was prepared by a simple and novel sublimation-induced gas-reacting process [222]. As a result, the reacted layer of Li x S y O z phases forms on the outside surface of the secondary particles and boundaries within primary particles inside the secondary particles. TEM images indicated that a thin reacted layer containing S uniformly formed on the entire surface of the secondary particles ( figure 5(B)). The 200 nm length boundary within the primary particles inside a secondary particle were completely filled with the S-containing phase, and the thickness of the reacted layer was about 15-20 nm. Because of its anisotropic volume stability, the obtained SNCM maintained a capacity retention of 83 To improve the reversibility of H2-H3 phase transition for LiNi 0.9 Co 0.1 O 2 , Ti 4+ , doping was designed using fluidized bed chemical vapor deposition [236]. XPS and TEM images indicated that more Ni 2+ ions were observed at the surface after Ti 4+ doping, forming a Li/Ni mixed region with a thickness of ∼5 nm in the 5% Ti doped sample. As a result, 5% Ti doped sample obtained a capacity retention ratio of 97.94% after 100 cycles at 38 mA g −1 , which was higher than the pristine sample with a capacity retention ratio of 89.08%. Even at a high current rate of 950 mA g −1 and 1900 mA g −1 , the 5% Ti doped sample delivered capacity of 159.5 mAh g −1 and 143.4 mAh g −1 , respectively. After 30 cycles, the H3 diffraction peaks of 5% Ti doped sample showed unobservable changes (figure 6(A)), implying the contraction in crystallographic c-direction was reversible, thus the generation of microcracks could be suppressed ( figure 6(B)). The interfacial chemistry of LiNi 0.8 Mn 0.1 Co 0.1 O 2 was improved by Ti 4+ doping using a modified co-precipitation method [237].
The EELS scanning showed that Ti 4+ was present throughout the primary particles and enriched at the top 1-2 nm surface. When Ti 4+ was doped into NCM, capacity retention at 1.0 • C within 2.5-4.5 V was increased from 69% to 80% over 300 cycles and from 58% to 70% over 500 cycles compared to the pristine sample. STEM-EDS mapping illustrated that Ti 4+ chemical environment was highly reversible due to intact Ti 4+ distribution after 300 cycles (figure 6(C)). To inhibit formation of NiO inactive phase for LiNi 0.8 Co 0.15 Al 0.05 O 2 during cycling, Ti-doping was used [238]. SEM images indicated that prepared samples were spherical particles with a size of 3-5 µm and 10-13 µm in diameter. TEM-EDX elemental mapping indicated that Ti was found to be mostly concentrated at the boundaries and surface edges of the grains. In the voltage range of 2.5-4.5 V, Ti-doped NCA obtained 74% of its initial capacity in contrast to 67% for PNCA after 50 cycles at 1.0 • C. For perfectly balancing the coating quality and scale-up production, fluidized bed chemical vapor deposition was used to prepare TiO 2 -coated LiNi 0.8 Mn 0.1 Co 0.1 O 2 [239]. There was a good linear relationship between TiO 2 coating content and coating time, and a smooth particle surface caused by increasing flow rates of carrier gas Ar (TiCl 4  78.1% for pristine NCM622 after 50 cycles at 140 mAh g −1 between 3.0 and 4.5 V. Such improved electrochemical properties could be ascribed to the improvement of the structural stability of the materials and the suppression of the interface reaction between the cathode and the electrolyte. Performance degradation of NCM622 at high cut-off voltage was mitigated by designing a functional interfacial layer that consists of a Ti 4+ surface doping and a TiO 2 surface coating [241]. TEM and SEM images indicated that the surface of the primary particles became rougher, the thickness of the layer/particle agglomerations ranged from 5 to 20 nm. As a result, TiO 2coated NCM622 obtained nearly 70% of its original capacity under a 1025 mA g −1 between 2.8 and 4.4 V at 30 • C. As shown in figure 6(D), the coated sample showed a slightly lower leakage current and smaller static leakage current values compared to the bare materials. It could be ascribed to the TiO 2 acting as a protection layer to suppress the irreversible phase transformation, whilst also reducing the rate of the electron-transfer reaction between the Li + -deintercalation cathode and solvent. NCM622 was prepared with a highquality ultrathin TiO 2 -coating layer by a commercial ALD system [242]. TEM images and SEM mapping indicated that Ti elements were evenly distributed across the micrograph and coating layer of thickness was about 5 nm. Benefiting from the high-quality ultrathin oxide TiO 2 layer coated on the surface of NCM622, HF attack and the dissolution of metal ions were suppressed and Li + diffusion wwas improved, thus, TiO 2 -coated NCM622 achieved a capacity retention of 85.9% at 1.0 • C after 100 cycles. For pristine NCM622, a capacity retention of 67.5% was obtained under the same test condition.
Even at a high current rate of 5.0 • C, TiO 2 -coated NCM622 delivered a capacity of 130.2 mAh g −1 . To demonstrate the criticalness of the interfacial parasitic reactions, ALD system was used to prepare TiO 2 -coated NCM622 [243]. After 100 cycles at 4.4 and 4.5 V vs Li/Li + under 0.1 • C, capacity retention ratio of 95.6% and 93.0% were obtained, respectively. This enhancement of capacity retention was also obtained for NMC622/graphite full cells. Both results illustrated those interfacial parasitic reactions were suppressed including the oxidation of the electrolyte at high potentials, dissolution of transition metals, cation mixing and particle fragmentation (figure 6(E)). Li et al prepared pouch cells with graphite anodes and bare and TiO 2 -coated NCM523 cathodes were used to test their thermostability and cycle life [244]. TiO 2coated NCM523 had a smaller volume contraction (1.1%) than pristine NCM523 (1.53%). When cycled at 200 mAh g −1 at room temperature between 2.7 and 4.5 V, a capacity retention of 30% was obtained after 120 cycles for pristine NCM523 and 80% was obtained after 200 cycles for TiO 2 -coated NCM523. This improved cycling stability was ascribed to the signal of the TM species being weak, as shown in ToF-SIMS chemical mapping (figure 6(F)). TiO 2 coating was helpful for enhancing NCM523 thermostability because of the higher phase transformation temperatures (figure 6(G)). TiO 2 nanofibers (TNFs)-coated NCA were prepared via an electrospinning method [245]. XRD results indicated that the characteristics of TNF on the secondary particle surface included higher amounts of anatase phase TiO 2 than the rutile phase TiO 2 , and FESEM images showed that TNF was acted like a connecting bridge between particles. The 1 wt.% TiO 2 -coated NCA could deliver 81.4% of the initial discharge capacity after 50 cycles at 0.1 • C and 60 • C, whereas the corresponding capacity retention of the pristine counterpart was only 64.7%. NCA was prepared with the high quality, dense, ultrathin and amorphous TiO 2 coating layer via ALD system [246]. XRD results and TEM images indicated that lattice parameters a, c and c/a slightly increased, and the thickness of dense thin coating layer was 5 and 8 nm for 5% and 8% TiO 2 -coated NCM, respectively. Because of its structure stability imparted by the TiO 2 coating layer, which protected the cathode particles from hydrofluoric acid attack, the obtained 5% and 8% TiO 2 -coated NCA obtained capacity retention of 90.21% and 85.54% after 100 cycles at 1.0 • C in a potential range of 2.5-4.5 V, respectively, whereas the corresponding capacity retention of the PNCA was only 50.13%. In particular, the 8% TiO 2 -coated NCA exhibited a capacity retention of 64.72% even at 1.0 • C rate and 55 • C after 100 cycles, while the PNCA was only 43.97%. To restore structural stability and electrochemical performance, a targeted solvothermal followed by re-oxidation approach was used to prepare Li 3 TiO 3coated NCA [247]. XPS results indicated that intensity of Ni 2+ decreased to the lower level compared with the PNCA, SEM images indicated that the thickness of coating layer was 20 nm. Benefiting from the minimal polarization and Theoretical analysis showed that V would be a promising doping source among transition metals configured with an oxidation state of 3 + . The two electrons occupied in the t 2g band of V 3+ would be donated for Ni redox reactions, resulting in V 5+ because of its lower electronegativity in comparison with Ni's electronegativity [248]. Sim [255]. The NCM111-Mo and NCM622-Mo showed capacity retentions of 80% and 81% at 170 mAh g −1 over 20 cycles, respectively. The porous structure with the expanded Li gap and reduced cation mixing should be responsible for the improved performance, which can shorten Li + ions diffusion distance in the electrode and enhance the structural stability. MoO 3 coated NCA was prepared via an ALD system injecting the gasphase Mo(Co) 6 source into the grain boundaries of primary particles and obtained a high capacity of 156.3 mAh g −1 after 100 cycles at 180 mA g −1 [256]. Even at a high current rate of 1800 mA g −1 , a capacity of 113.4 mAh g −1 was also achieved. After cycling, the modified NCA particles suggested a substantially thinner CEI layer than bare NCA, and a well-maintained layered structure of the modified NCA particles ( figure 8(A)). Thus, MoO 3 coating layer can not only suppress the interfacial side reactions, but also provide additional Li + insertion sites. Li x Ni y W z O-like heterostructures were designed on surface of LiNi 0.91 Co 0.045 Mn 0.045 O 2 by a solid-state reaction method [172]. Benefiting from the formation of radially aligned grains, doping W to the primary particle surface and the effect of armor-type tungsten-rich heterostructure, LiNi 0.91 Co 0.045 Mn 0.045 O 2 @2 W retained a capacity of 179.6 mAh g −1 at 200 mA g −1 after 500 cycles with a capacity retention ratio of 85%. After modification, the inner cracks and outer Fm-3m rock salt phase were reduced greatly during cycles ( figure 8(B)). Becker et al constructed WO 3 coated LiNi 0.8 Co 0.1 Mn 0.1 O 2 via a sol-gel method, followed by a heat treatment process [257]. The (NH 4 ) 10 H 2 (W 2 O 7 ) 6 acted as a reactant, which could react with residual lithium species during the annealing process. The Li 2 WO 4 possessed good ionic conductivity and WO 3 enhanced the thermal stability of the composite. Therefore, when utilized as coating layer of NCM, the WO 3 coated NCM/graphite full cells delivered state-of health of 80% after 865 cycles at 180 mA g −1 , exhibiting superior high-rate cycling stability. As shown in figure 8(C), the coating layer was able to suppress particle cracking caused by the reaction between the active material surface and the electrolyte during cycling. WO 3 modified NCM811 was synthesized by mixing Ni 0.8 Co 0.1 Mn 0.1 (OH) 2 precursors, LiOH · H 2 O and WO 3 , followed by heat treatment process [258]. In the voltage range of 2.8-4.5 V, WO 3 modified NCM811 (WNCM-0.5) delivered a capacity retention ratio of 92.1% after 100 cycles at 200 mA g −1 . Even at a high current rate of 2000 mA g −1 , a capacity of more than 160 mAh g −1 was also achieved. As shown in figure 8(D), the WO 3 coating layer contributed to the restraint of the diffusion of the rock salt phase from the surface to the interior, and enlarge the interplanar distances, thus enhancing its better cycling performance and retarding the degradation rate of capacity. WO 3 coated NCM622 was prepared via a liquid phase method using H 2 WO 4 as a raw material [259].    [263]. As shown in figures 9(B) and (C), Mn concentration was increased at the surface of the 1 wt.% Mn-NCM811 particle, and its bulk structure was not affected by the formation of a reliable Mn-rich surface layer. FESEM images revealed that the Mn-rich surface could enhance interconnection between the primary particles in NCM811 due to relatively strong Mn-O bonds, thus Mn-NCM811 achieved a capacity retention of 87.3% in contrast to 57.0% for the raw NCM811 after 50 cycles at 190 mA g −1 and 60 • C. Following Mn 4+ sites existed in the TM layer, a certain amount of Ni 3+ has to be reduced to Ni 2+ to maintain the electric neutrality. In order to investigate this effect, MnO 2 -modified NCM811 was prepared via wet chemical method using KMnO 4 as precursors [264]. The obtained MnO 2 -modified NCM811 showed superior electrochemical performance, including long-term cycling stability (96.3% capacity retention after 50 cycles at 200 mA g −1 ), thermal stability (92.5% capacity retention after 50 cycles at 200 mA g −1 , 50 • C), and high rate-capability (204.7 and 200 mAh g −1 at 600 and 1000 mA g −1 , respectively). Some portion of the Ni 2+ ions that migrate from the transition metal layer to the lithium layer should be positive for preventing inter-slab collapse and improving structural stability as well as rate capability. XRD and HRTEM analysis indicated that some Mn 4+ ions were incorporated into the bulk showing increased lattice parameters. Thus, partial Ni 3+ would be reduced to Ni 2+ , and there was a coating layer with thickness of 6-12 nm for MnO 2 -modified NCM811. Electrolyte decomposition was suppressed by forming a thin SEI film. In addition, MnO 2coated NCM811 (LNMC and M) was prepared via treating the precursor with strong oxidant KMnO 4 [265]. The increase of c/a value indicated the priority of lattice grew along the c axis and provides large channels for Li + transportation process. LNMC and M delivered a capacity of 200.6 mAh g −1 after 50 cycles at 200 mA g −1 and 50 • C with the capacity retention of 99.8%, and exhibited superior rate performance (190.3 and 185.2 mAh g −1 at 400 and 1000 mA g −1 , respectively). The improved performance was ascribed to the ∼12 nm coating layer that suppressed electrolyte decomposition and side reactions, thus a thin SEI film was formed on the surface of the LNMC and M. The electrochemical performance of NCM811 was improved by coating a thin MnO 2 layer and combined with the formation of single crystal primary particles (S and T) [266]. A 4-7 nm MnO 2 coating layer on the single crystal morphology was formed by a reaction between MnO 4 − and Ni 2+ under calcined process. When applied as a cathode for LIBs, a high capacity of 193.5 mAh g −1 could be obtained at 200 mA g −1 after 50 cycles, which was better than the performance of pristine NCM (S and U). As shown in figures 9(D) and (E), S and T showed a smaller shift of (003) peak and cracks than S and U.
The decreased migration barrier of Ni ion due to the oxygen vacancy and the lowering of the system energy by the Li + compensation causes extra cation mixing and lattice distortion [267]. Hence, Ni ion migration introduces strain perpendicular to the Ni layers with oxygen defects and Li vacancies, which leads to dislocations and defects with lattice distortion. Yang et al proposed to adjust the valence state of Ni and avoid the oxygen defects by designing a manganese oxide coating method and using 2 wt.% (CH 3 COO) 2 Mn as precursors [268]. EDS mapping results indicated that the lattice of the interstitial coating layer matched with the internal Ni-rich layered structure and released stress concentration to avoid the damage by lithium migration. Rietveld refinement results indicated that the unit cell volume was extended at a degree of 0.4%. As shown in figures 9(F)-(H), the effect of Mn coating was that the ratio of Ni 3+ 2p 3/2 to Ni 2+ 2p 3/2 decreased from 9.49 to 2.46, and oxygen content increased after cycles. In the voltage range of 2.8-4.25 V, the M-NCM88 cathode delivered 80.6% of its initial capacity in contrast to 47.2% for the B-NCM88 at 190 mA g −1 after 200 cycles. This report effectively solves the issues resulting from extra cation mixing and lattice distortion. Additionally, Wu et al prepared LiNi 0.71 Co 0.09 Mn 0.2 O 2 with a uniform layered R-3m phase following a self-assembled concentration-gradient shell via a wet chemical method [269]. EDX scanning analysis showed that the thickness of the concentration-gradient shell was over 1 µm on the obtained  2 with Mn(CH 3 COO) 2 · 4H 2 O and sintering with LiOH · H 2 O at high temperature [270]. The LOD@NCM showed a stable capacity of 166 mAh g −1 with a capacity retention of 90.4%, while pristine NCM exhibited a discharge capacity of 136 mAh g −1 with only 77.2% of capacity retention after 200 cycles at 200 mA g −1 . The improved performance could be ascribed to the improved surface oxygen-keeping capacity, which was benefited for decreasing particle crack and Li + /Ni 2+ disorder.
Among the group VIII metals, Fe and Co have been investigated as modified elements for nickel-rich materials. The lithium-deficient Li[Ni 0.78 Co 0.11 Mn 0.11 ]Fe 0.0023 O 2 (NCMF) was prepared by co-precipitation method [271]. The related characterization indicated that secondary particles exhibited a bimodal size distribution with a size of ∼10 µm. The parts of Fe 3+ and Ni 2+ ions that resided in the 3a sites could reduce the ratio of Ni 2+ and cation mixing, and thus restrain the local collapse of the LiO 2 inter-slab space. Moreover, the Fe 3+ doping could reduce the residual lithium compounds on the surface of NCMF. Therefore, NCMF delivered a capacity retention of 70.9% after 100 cycles at 360 mAh g −1 in the voltage range of 3.0-4.5 V. High-angle annular dark-field (HAADF)-STEM analysis displayed that the cation-disordered region of cycled NCMF was increased from ∼1 nm to ∼5 nm, whereas, the raw sample was significantly increased from ∼2 to ∼35 nm ( figure 10(A)). Du et al constructed partial substitution of Co with Fe in LiNi 0.8 Co 0.15 Al 0.05 O 2 via ball-milling followed by a two-step solid reaction [272]. The significantly improved cyclability was mainly attributed to the effect of the FeO 6 octahedron on the edge-shared NiO 6 octahedra via enhanced electron localization. The partial substitution of Co with Fe could suppress the formation of the single monoclinic phase and SEI film. Therefore, when cycled at 200 mA g −1 at room temperature in the voltage range 2.8-4.3 V, LiNi 0.8 Co 0.125 Fe 0.025 Al 0.05 O 2 (NCF1) and LiNi 0.8 Co 0.075 Fe 0.075 Al 0.05 O 2 (NCF3) delivered capacities of 114.2 and 122.8 mAh g −1 after 350 cycles, respectively. Even at 200 mA g −1 , 55 • C, NCF3 delivered a capacity of 127.6 mAh g −1 after 350 cycles, corresponding to a capacity retention ratio of 77.5%. As shown in XPS spectra of NCA and NCF3 after cycles ( figure 10(B)), the relative intensities of C-C, C-O, ROLi, ROCO 2 Li, LiCO 3 and LiF bands for NCF3 are weaker than those for NCA, indicating thinner SEI film formed and side reactions greatly suppressed on NCF3.
In addition, Co 3 O 4 , LiCoO 2 and Li 0.73 CoO 2 have been investigated as coating materials for nickel-rich cathode materials and exhibited superior electrochemical performance. To suppress the release of divalent nickel ions from the LiNi 0.8 Co 0.1 Mn 0.1 O 2 , a nanostructured stabilizer (Co 3 O 4 , Li x CoO 2 and LiCoO 2 ) with an epitaxial structure was created on the NS-NCM surface [273]. In the crystal structure, nanostructured stabilizer on the NS-NCM could prevent the nickel ion crossover from the cathode to the anode side by mitigating the nickel defect formation at the cathode surface during battery cycling. The HAADF-STEM images with the line EDX results showed that the TM concentration gradient throughout the NS-NCM particles in combination with the physically blocked surface led to the gradual oxidation of the Ni 2+ towards the Ni 3+ , thus decreasing the thickness of the cation mixing layer. In particular, the suppression of the TM defect formation could restrain the catalytic reaction of the solvent, maintaining a uniform and stable anode SEI layer. As a result, NS-NCM delivered a ∼33% higher discharge capacity retention and lower overpotential than raw NCM811 after 150 cycles at 60 • C with charge and discharge rates of 0.5 • C and 1.0 • C. The designed NS-NCM/graphite full-cell delivered a ∼51% higher discharge areal capacity retention and lower overpotential than the NCM/graphite full-cell after 500 cycles at 45 • C within a voltage window of 2.8-4.2 V. Tao et al prepared Co 3 O 4 coated NCM622 by a PVP-assisted method [274]. The 1 wt.% Co 3 O 4 modified NCM622 delivered a discharge capacity of 114.1 mAh g −1 with a capacity retention of 60.3% after 100 cycles at 180 mA g −1 within a voltage window of 2.8-4.6 V. The improved capacity and cyclability were mainly attributed to inactive lithium residues consumed and the interface reactions being suppressed by a coating layer Co 3 O 4 with a thickness of ∼7 nm. Otherwise, LiCoO 2 -coated NCA was synthesized by a molten salt method [275]. Benefiting from the decrease in NiO-type surface layer on the coated sample, the 3.0 wt.% LiCoO 2 -coated NCA delivered 95.8% of its initial capacity in contrast to 87.9% for the raw NCA after 50 cycles at 2.0 • C. LiCoO 2 -coated NCA was synthesized by a molten salt method, the thickness of coating layer was 40-100 nm [276]. Among the four composites, 3.0 wt.% LiCoO 2 -coated NCA sintered at 800 • C for 4 h at weight ratio of KCl/NCA = 2 displayed the highest specific capacity of 193.6 mAh g −1 with capacity retention of 98.7% at 36 mA g −1 after 50 cycles. Even at a high current rates such as 90, 180, 360 and 540 mA g −1 , as well as at a high temperature of 55 • C, the discharge capacities of coated NCA were obviously more than that of the raw NCA. Liu et al synthesized 3 wt.% LiCoO 2 -coated NCA by a co-oxidation-controlled crystallization method [277]. This fresh LiCoO 2 -coated NCA delivered a discharge capacity of 193.6 mAh g −1 with a capacity retention of 98.7% after 50 cycles at 0.2 • C. LiCoO 2 -coated NCA also maintained a fiftieth discharge capacity of 183.4, 175.1 and 167.9 mAh g −1 after storage at a relative humidity of 40%, 60% and 80% for 3 months, respectively. After FTIR spectra and XPS analysis, the improved stability was ascribed to the LiCoO 2 coating layer that suppressed effectively the reactions between NCA and atmosphere and was resistant to H 2 O and CO 2 in air. Lu to the synergistic effects of the Co-O bond and the coexistence of Co 4+ and Al 3+ , which could help to form a stable cathodeelectrolyte interface in the Li + intercalation/deintercalation process, as well as boost the ion diffusion and electrical conductivity of the coated NCA cathode (figures 10(C) and (D)). This surface chemical state modulation strategy provides a new possibility of surface engineering with nickel-rich cathodes for high-performance LIBs.

Multi-element coating and doping
To date, many studies have reported multi-element modification for nickel-rich cathode materials with high-rate capability and cycle stability. Constructing multi-element ion surface modification is an effective approach to improve electrochemical performance compared to their single-element ion modification, originating from the synergetic effect of multiple metal species [279,280] [281]. As a result, the sample for x = 0.01 exhibited a capacity retention rate of 85% in contrast to 67% for bare NCM at 1.0 • C after 300 cycles. Similarly, Mg and Zr doped LiNi 0.6 Co 0.2 Mn 0.2 O 2 were prepared by co-precipitation method, respectively [282]. SEM showed that surface area increased with the addition of dopants, which was partly due to the reduced secondary particle size. The 1% Zr-dopants at 900 • C had a capacity of 8 mAh g −1 lower than the pristine material at 90 mA g −1 . After 25 cycles, the 0.5% Mg-dopants at 900 • C have a capacity that is only 2 mAh g −1 lower than the pristine materials, which is likely an effect of the improved crystallinity at higher calcination temperature. XRD plots showed that Mg-doped NCM showed evidence of the NCM breaking down into two phases ( figure 11(A)). The Mgdoped NCM appeared to be stable and underwent phase transformation from 200 • C to 400 • C, followed by the pristine, and then the Zr-doped materials ( figure 11(B)). Choi [291]. There are segregation phenomena near the particles' surfaces that may increase the electrodes' impedance. All of the dopants improved the electrodes' cycling stability during prolonged cycling at 360 mA g −1 and 720 mA g −1 (45 • C), compared to the un-doped NCM811 electrodes, which was due to lower volumetric changes in the active mass upon lithiation/de-lithiation processes. Mg 2+ doping worsened the initial capability of the material because of its possible incorporation in the lithium sites and the formation of surface layers. Recently, singlecrystal LiNi 0.874 Co 0.09 Mn 0.03 Al 0.003 Zr 0.003 O 2 (AZ0.3-SNCM) was prepared by a combination of co-precipitation and calcination processes [292]. It can achieve a reversible capacity of 163.0 mAh g −1 at 100 mA g −1 after 150 cycles within a voltage range of 2.75-4.6 V. It is confirmed that Al/Zr codoping in a SNCM lattice is stable against internal strain and Li/Ni cation mixing upon cycling at a high cut-off voltage. A pouch cell of single-crystal AZ0.3-SNCM reveals an alternative path towards developing ideal cathode materials beyond the spherical secondary particles. Nanthagopal et al prepared a nitrogen-doped carbon (NC)coated NCM811 cathode via a dry-solid state reaction method followed by calcination in Ar [293]. Benefiting from a coating thickness of 4 nm NC, preventing surface degradation and HF acid attacks, the 0.2 wt.% NC@NCM811 composite exhibited an initial discharge capacity of 183 mAh g −1 and maintained a capacity retention of 65.74% after 50 cycles at 1.0 • C, which was higher than bare NCM811 with a capacity retention of 63.15%. Chen et al utilized NC as a coating material, and synthesized NCM622/NC composites by simple mechanical activation and pyrolysis methods [294]. The incorporation of N makes carbon more electron-rich; thus, the electron clouds will be biased to NCM622 from NC, generating a stronger electronic coupling between NC and NCM622. Benefiting from the enhanced electron-conductivity, cation ordering, and decreased surface impurities, the NCM622/NC-0.1 delivered a capacity retention of 92.0% after 100 cycles at 195 mA g −1 within 3.0-4.5 V, better than that of bare NCM622 with capacity retention of only 76.4%. Even at 975 mA g −1 , NCM622/NC delivered a high discharge capacity of 156.6 mAh g −1 . Nitrogen and sulfur dual-doped carbon (NSC) was prepared and subsequently coated over NCM811 by a simple and one-step process [295]. In the voltage range of 3.0-4.5 V, the 0.3 wt.% NSC@NCM811 delivered a capacity of 177.89 mAh g −1 with the corresponding capacity retention of 90.64% after 100 cycles at 0.1 • C, while raw NCM811 delivered 65.67% of its initial capacity. The 0.3 wt.% NSC@NCM811 delivered a discharge capacity of 111.39 mAh g −1 even at a high rate of 2.0 • C. These results were ascribed to a coating layer that improved ionic and electronic transfer as well as prevented the degradation of the electrodes. NCM811 was wrapped with a pyrazine-linked 2D sheet by the condensation reaction of amine and ketone in the presence of NCM811 particles. The TEM image indicated that the coating layer was uniform with an approximate thickness of 1 nm (figure 11(D)) [296]. After 100 cycles at 600 mA g −1 and 60 • C, this composite delivered 68.2% of its capacity retention in contrast to 17.9% for the raw NCM811. In addition, the Pyr-2D coating turned out to improve the air stability and rate performance of NCM811 as indicated by the robust cycling performance. This improved performance was ascribed to a protective layer mitigated by the growth of a resistive layer and did not inhibit the Li + diffusion much in the electrode. Cyclizedpolyacrylonitrile (cPAN) coated NCM622 was prepared via a wet-coating method to combine the high ion conductivity of the conductive polymer and the good electronic conductivity of the delocalized sp 2 π bonds in graphite group with pyridinic-N [297]. The TEM image of the 2 wt.% cPANcoated NCM622 revealed that the thickness of the amorphous film was about 25-30 nm. In the voltage range of 2.8-4.5 V, the 2 wt.% cPAN-coated NCM622 delivered a capacity of 161.5 mAh g −1 with a capacity retention of 88.6% at 1.0 • C after 200 cycles, while bare NCM622 obtained a capacity retention of 73.3%. The polypyrrole coated NCA (PL-LNCA-2) was prepared by a liquid-phase chemical oxidation polymerization method [298]. Among the three composites, the PL-LNCA-2 displayed the highest capacity retention of 88.8% at 180 mA g −1 after 100 cycles and an outstanding rate performance of 163.5 mAh g −1 at 900 mA g −1 , while the pristine LNCA material exhibited a capacity retention of 71.6% at 180 mA g −1 after 100 cycles. The PL-LNCA-2 cathode material still remained the spherical morphology and as the rhombohedral phase after cycles. This likely indicated that the PPy maintained the structural stability of cathode material during the cycles. Xu et al fabricated polymer poly (3,4-ethylenedioxythiophene) skin decorated on both secondary and primary particles of LiNi 0.85 Co 0.1 Mn 0.05 O 2 by an oxidative chemical vapor deposition technique [299]. The ultra-conformal poly(3,4-ethylenedioxythiophene) skin significantly restrained the undesired phase transformation and the associated oxygen loss, retarded intergranular and intragranular mechanical cracking, boosted the transport of lithium ions and electrons, and effectively stabilized the cathodeelectrolyte interface. As a result, this composite obtained 91% of its initial capacity in contrast to 54% for bare LiNi 0.85 Co 0.1 Mn 0.05 O 2 upon 100 cycles at 200 mA g −1 . A ∼3 µm sized NCM622 with a thin polyphenylene film was prepared via the spontaneous reaction between NCM622 and C 6 H 5 N 2 + BF 4 − [300]. In this unique structure, the aromatic diazonium cations are reduced to phenyl radicals (C 6 H 5 N 2 + → C 6 H 5 N 2 * ) via electron transfer from bare NCM622, forming LNCM ( figure 11(E)). Therefore, polyphenylene/LNCM-3 displayed a thin conductive polymer layer with a thickness of ∼10 nm, which endowed this composite with a high discharge capacity of ∼148 and ∼105 mAh g −1 at 0.1 • C and −20 • C or at 1.0 • C and −20 • C, respectively, and a superior low-temperature cycle stability (capacity retention was up to 90% at 0.5 • C over 1150 cycles). Similarly, NCM811 with a 5-8 nm cathode electrolyte interphase (CEI) layer was modified through chemical reactions between the lithium tetra(trimethylsilyl) borate (LTB) layer and the NCM811 electrode (figure 11(F)) [301], and 0.025 LTB-NCM811 delivered a capacity retention of 75.5% at 200 mA g −1 and high temperature after 100 cycles. This cycling stability was ascribed to the CEI layer preventing boundary between the primary particles covered with decomposed adducts from electrolyte decomposition.
Li et al designed a NCM811 surface with a layer of 4vinylbenzeneboronic acid (4-VBBA) to make the surface less hydrophilic [302]. . Additionally, the practicability of the GB-NCM composite was further evaluated by fabricating full cells using graphene balls as anodes. At both 25 • C and 60 • C, it delivered a capacity retention of 84.1% and 78.6% at 700 mA g −1 after 500 cycles, respectively, which showed a great promise for the practical application of the GB-NCM electrode. The high electrochemical performance of the GB-NCM composite was ascribed to the homogeneously integrated SiO 2 and graphene, which could effectively enhance the electric conductivity of LiNi 0.6 Co 0.1 Mn 0.3 O 2 powders, block electrolyte penetration and phase transition. To hinder the direct interface reaction, rGO and SiO 2 nanoparticles coated NCM523 were prepared [305]. The rGO with hierarchically organized 2D nanosheet subunits exhibited a large surface and a shortened Li + diffusion pathway, facilitating the kinetic properties and ultimately increased power density. As a result, rGO-SiO 2 @NCM523 sample exhibited an excellent cyclic retention of 80% compared to 46.1% for the pristine cathode after 100 cycles at 180 mA g −1 and 55 • C. Even at 360 and 900 mA g −1 , the rGO-SiO 2 @NCM sample delivered high capacities of 146.9 and 126.7 mAh g −1 , respectively. A facile wet chemical method was used to prepare rGO (outer)/SiO 2 (inner) double layer-coated NCM523 composite with a homogeneous graphene sheet surface [306]. Benefiting from the rGO-SiO 2 coating layer with an improved Li + and electron conductivity, this material could maintain a capacity retention of 88.5% with an intact morphology compared to 57.8% for pristine NCM523 with damaged particles after 100 cycles at 18 mA g −1 . Even at a high current rate of 900 mA g −1 , a capacity of 103.2 mAh g −1 was also achieved, which was higher than bare NCM523 (48.6 mAh g −1 ) and 0.5 wt.% SiO 2 coated NCM 523 (42.2 mAh g −1 ). In recent studies, a simultaneous bulk doping and surfacecoating approach of multi-transition metals was proposed to exploit their advantages while maintaining cathode stability, high capacity, safety, and good charge transport. Among them, LiNi 0.8−x−y Co 0.15 Al 0.05 Cu x Fe y O 2 (NCACF) was prepared to decrease the relative ratio of Ni 2+ /Ni 3+ on the surface and increase the performance of the Li + diffusion coefficient, structure stability, and charge transfer [307]. When applied as a cathode for LIBs, NCACF displayed discharge capacity of 164.2 mAh g −1 after 100 cycles at 180 mA g −1 , corresponding to a capacity retention of 92.50%, whereas the capacity retention of NCA, LiNi 0.8−x Co 0. 15 [308]. Due to LNFO and NFO being used as the coating layer, bonding with the surface oxygen vacancies of the bulk material meant it was confirmed to be an effective approach to improve the electrical and ionic conductivity at the interface. As a result, the LNFO and NFO delivered 74.32% of its initial capacity in contrast to 51.59% for raw NCM811 after 100 cycles at 200 mA g −1 and 55 • C within a voltage range of 2.75-4.4 V. Even at a high current of 2000 mA g −1 , LNFO and NFO also obtained a capacity of 140.56 mAh g −1 .
The in-situ XRD results showed that the maximum position shifts of the (003), (101) and (104) peaks for the LNFO and NFO sample are less than those of NCM during the charge process. Chen et al prepared single crystalline NCM811 coated on cobalt (0.6 wt.%) and manganese (0.08 wt.%) in comparison with NCM811 coated by cobalt (0.6 wt.%) and bare NCM811 [309]. Benefiting from the formation of the spinel Co 3 O 4 and LiMn 2 O 4 layer on the surface, Co 3 O 4 and LiMn 2 O 4 coated NCM811 (Co and Mn-NCM811), Co 3 O 4 coated NCM811 (Co-NCM811) and NCM811 delivered capacity retentions of 91.4%, 89.9% and 86.5% at 185 mA g −1 after 100 cycles, respectively. Even at a high current rate of 925 mA g −1 , Co and Mn-NCM also achieved a capacity of 161.7 mAh g −1 .
To improve cycle stability and rate performance of NCA, Sb-doped tin oxide (ATO) and Li 2 TiO 3 (LTO) were used as coating materials [310]. The composite delivered a capacity retention of 88.56% in contrast to 76.15% for bare NCA at 200 mA g −1 and 60 • C after 200 cycles. Even cycled at 1000 mA g −1 , a discharge capacity of 153 mAh g −1 also was obtained, which was 24 mAh g −1 higher than that of the bare NCA. Such an enhanced performance could be ascribed to the ATO/LTO coating layer, which suppressed increase of lithium residues during storage and facilitated the Li + diffusion. To improve cycling stability and increase the lithium-ion diffusion coefficient of LiNi 0. 92 figure 12(A)). LiNi 0.79 Mn 0.1 Co 0.1 Al 0.01 O 2 was encapsulated in the glassy LiBO 2 network (NMCA-LBO) within the primary particles [313], and NMCA-LBO retained 92.2% of its initial capacities after 200 cycles at 100 mA g −1 , while that of pristine NMCA-was ∼89.0%. Even at a high current rate of 2000 mA g −1 , NMCA-LBO achieved a capacity of 96.4 mAh g −1 . In a full cell with Si/C-400 as an anode, capacity retention of 85% was maintained after 1000 cycles at 100 mA g −1 even under −20 • C. The suppression of the microcrack evolution and phase transition (figure 12(B)) and enhanced ion conduction upon the LBO encapsulation accounted for the satisfactory rate capability of the modified cathode. The cycling stability of LiNi 0.91 Co 0.06 Al 0.03 O 2 was improved via the synergistic effect of excellent fluidity of H 3 BO 3 and the outstanding stability of Y 2 O 3 (NCA-BY0.1) [314]. Benefiting from the interface reaction being mitigated, NCA-BY0.1 showed a capacity retention of 93.7% at 1 • C over 100 deep charge-discharge cycles, which was much better than bare NCA (91.6%). The soft-packing battery of NCA-BY0.1 also exhibited a capacity retention of 81.7% after 800 cycles at 1 • C. Chae et al prepared LiNi 0.6 Co 0.2 Mn 0.2 O 2 electrode using metal-organic framework (MOF) as a multifunctional additive for selectively trapping transition-metal components [315]. The MOF can trap the volume of nickel ions through its porous structure to suppress surface degradation triggered by the nickel-ion dissolution. As a result, the obtained compositions retained more than 95.0% of their capacity at 1 • C, 25 • C and over 50 cycles, and 5 wt.% MOF-NCM622 exhibited 82.2% of its initial capacity in contrast to 65.5% for raw NCM622 after 100 cycles at 1 • C,  [316]. Among the three composites, the 0.5% BT-NCM622 displayed the highest specific capacity of 128 mAh g −1 at 825 mA g −1 within 3.0-4.5 V after 200 cycles with capacity retention of 86.6%, which was better than bare NCM622 with a capacity retention of 84.2%. The 0.5% BT-NCM622 also delivered a outstanding rate performance of 140 mAh g −1 at 1650 mA g −1 . The NiTiO 3 (NTO) nanocoated NCA was fabricated by directly bonding NTO to surface framework of NCA and the surface lithium residues evolved into a Li-Ni-Ti-O mixed phase coating [317]. The improved performance was ascribed to the minimization of the side-reaction, and the undesirable SEI layer, surface and charge-transfer resistances were obviously suppressed after cycles. The PO 4 3− polyanion and Mn 4+ cation co-doped NCA (NCA-PM) was prepared to enlarge the channel for Li + -intercalation/deintercalation, lower the cationic mixing, and suppress the structural degradation during cycling [325]. As a result, 3 wt.% NCA-PM (NCA-PM3) could deliver 80.9% of the initial discharge capacity after 100 cycles at 180 mA g −1 and 55 • C, whereas the corresponding capacity retention of the 1 wt.% NCA-PM (NCA-PM1) and PNCA were only 60.4% and 33.6%, respectively. In particular, the NCA-PM3 and NCA-PM1 exhibited capacity retention of 85.4% and 85% even at a 900 mA g −1 rate, while the counterpart of NCA was 81.2%. In the voltage range of 2.7-4.5 V, NCA-PM3, NCA-PM1 and NCA retained capacity retentions of 78.8%, 71.0% and 61.3% after 100 cycles at 180 mA g −1 , respectively. Song et al prepared Ca 3 (PO 4 ) 2 (CP) modified NCM811 via a solid phase method [326]. Benefiting from cathode-electrolyte interphases layer with thickness of 1.0 nm, 0.1 CP-modified NCM811 showed less electrolyte decomposition, and exhibited 72.5% of its capacity retention in contrast to 39.9% for raw NCM811 at 45 • C after 150 cycles. Cross-sectional SEM images illustrated that the formation of microcracks were suppressed during electrochemical cycling. Benefiting from the development of an anionic PO 4 functional group on the CEI layer and strong binding of calcium ions with oxygen elements, 0.1 CP-NCM811/graphite full-cell delivered a capacity retention of 86.7% after 150 cycles at 0.3 • C. To achieve a long-term cycling stability at elevated temperatures and voltages, Mn 3 (PO 4 ) 2 (MP) coated NCM622 was fabricated [327]. When cycled at 10.0 • C and 60 • C within voltage range of 3.0-4.3 V, bare NCM and MP-NCM delivered capacity retentions of 68.8% and 83.1% after 100 cycles, respectively. MP-NCM also delivered 86.9% of its capacity retention in contrast to 67.4% for raw NCM within voltage range of 3.0-4.6 V after 50 cycles at 0.1 • C. Even at a high current rate of 10.0 • C, MP-NCM electrode obtained a discharge capacities of 114.5 mAh g −1 . Such enhanced improvement was attributed to the MnPO 4 layer, which could dramatically reduce transition metal dissolution and the occurring side reactions with the electrolyte, especially at higher temperatures and cut-off voltages. Because of the strong PO 4 (P) covalent bonds, nano-sized crystalline MP coating layer was designed on NCM622 to improve its thermal properties [328]. After 50 cycles at 0.5 • C and 60 • C, discharge capacities of 142.2 and 160 mAh g −1 were obtained for P-NCM and MP-NCM, corresponding to a capacity retention ratio of 81.3% and 90.9%, respectively. This improved performance also was associated with decreased interfacial impedance investigated by AC impedance spectroscopy. In view of this, Lee et al manufactured Ni 3 (PO 4 ) 2 -coated NCA by a ball milling [329], and its capacity retention after 100 cycles at 100 mA g −1 , 55 • C was 75%, which was notably greater than that of PNCA (53%). For instance, the discharge capacity of Ni 3 (PO 4 ) 2 -coated NCA at a 2000 mA g −1 rate was 149 mAh g −1 , while the PNCA delivered only 127 mAh g −1 . This result encouraged us to believe that the Ni 3 (PO 4 ) 2 coating layer on NCA did protect the cathode from chemical attack by HF and thus suppress an increase in charge transfer resistance. A thin LaPO 4 coating layer was designed on an NCM811 surface through a facile wet chemical route to improve its cycle stability and rate performance [330]. A LaPO 4 -coated sample showed smaller charge transfer resistance and higher Li + diffusion coefficient in compare with pristine sample, and therefore it delivered 91.2% of its capacity retention in contrast to 76.4% for raw NCM811 after 100 cycles at 1 • C. Even at a high current rate of 10.0 • C, a capacity of 124 mAh g −1 was also achieved, which was higher than those of the pristine sample (108.6 mAh g −1 ). An in-situ ZrP 2 O 7 coating layer was introduced to improve cycle and rate performance of NCM811 [331]. After 100 cycles at 1.0 • C, pristine NCM811 and ZrP 2 O 7 coated NCM811 (1.0ZPO-NCM811) delivered capacity retention ratios of 66.35% and 86.92%, respectively. The discharge capacity of 1.0ZPO-NCM811 at a 10 • C rate was 148.0 mAh g −1 , while the PNCA delivered only 108.2 mAh g −1 . The high electrochemical performance of the 1.0ZPO-NCM811 cathode was ascribed to the dense inorganic pyrophosphate layer, which not only had strong Zr-O covalent bonds against the electrolyte but also eliminated surface residual lithium. Zhu et al prepared LiFePO 4 (LFP)-coated LiNi 0.5 Co 0.2 Mn 0.3 O 2 by the high-speed dispersion and mechanical fusion method [332]. In the voltage range of 2.75-4.5 V, the capacity retentions of NCM/MesoCarbon MicroBead graphite anode (MCMB) and NCM-LFP/MCMB pouch cells were 81.4% and 90% after 100 cycles at 180 mA g −1 , respectively. In-situ XRD patterns showed that the phase transition temperatures were increased from 405 • C to 445 • C after coating. Thus, postponed structure evolution led to the increased thermal decomposition temperature, which was quite significant toward a broad application prospect of NCM-LFP composites as cathode materials in LIBs. Thermal stability of NCA was improved by constructing a LiFePO 4 coating layer, which helps to reduce the formation of a cathode-electrolyte interface layer and the extent of cation mixing in NCA [333]. Even in a full pouch cell configuration versus graphite-based anodes, NCA-LFP obtained a capacity retention of 95% after 150 cycles at 90 mAh g −1 and 55 • C, whereas the NCA cathodes faded 2.5 times faster. The SEM images showed that NCA-LFP surface seemed relative smooth and clean after 100 cycles, indicative of less electrolyte decomposition, whereas the NCA surfaces appeared to be covered by a thick layer of CEI. A sol-gel method was used to fabricatea fast ionic conductor Li 1.5 Al 0.5 Zr 1.5 (PO 4 ) 3 (LAZP) coated NCM811 [334], and the discharge capacity of NCM@LAZP-1 was 179.3 mAh g −1 after 200 cycles at 200 mA g −1 in the potential range of 2.8-4.5 V, which was notably greater than that of pristine NCM (112.1 mAh g −1 ). Accordingly, the capacity retention of NCM@LAZP-1 after 50 cycles was 84.8%, which was a significant improvement over pristine NCM (52.3%). The good cycle stability of cathode material at high cutoff voltage resulted from the decreased Li + /Ni 2+ cation mixing, enhanced Li + diffusion rate and structural stability. Because the LiTa 2 PO 8 coating layer had an ordered channel with threedimensional Li + ion transport, the side reaction was inhibited and the surface structure stability of the NCM materials was enchanced. As such a 2-4 nm LiTa 2 PO 8 coating layer was designed on the surface of NCM811 to improve electrochemical performance of NCM811 at high voltage [335]. In the voltage range of 2.7-4.6 V, LiTa 2 PO 8 coated NCM811 (NCMT3) delivered a high capacity of 175.9 mAh g −1 at 180 mA g −1 after cycling 100 times, corresponding to a capacity retention ratio of 84.85%, while the pristine NCM811 cathode delivered only 140.8 mAh g −1 , matching with a capacity retention rate of 71.58%. The discharge capacity of NCMT3 at 540 mA g −1 rate was 164.1 mAh g −1 , while the pristine NCM811 delivered only 125.6 mAh g −1 . Highvoltage performance of LiNi 0.87 Co 0.1 Al 0.03 O 2 was improved by Li 2 O-BPO 4 (LBP) coating [170]. Benefiting from the decreased cation mixing and increased Li + conductivity, NCA@LBP nanotubes delivered 76.2% of its initial capacity in contrast to 68.7% for raw NCA after 100 cycles at 200 mA g −1 within 2.7-4.7 V. TEM-FFT images ensured that NCA@LBP cathode still maintained the R-3m space group from surface to bulk during electrochemical cycling ( figure 12(E)).
An anion slab is forced closer to the TM slab when nitrogen is used as anionic dopants in nickel-rich cathode materials as nitrogen has higher charge relative to oxygen, resulting in an enlarged inter-slab distance for the migrating [336]. ToF-SIMS results indicated that fluoride ions were concentrated near the outer surface of the secondary particles and not detected in the core of the particles. In contrast, nitrogen was primarily found within the core of the particles. Benefiting from the samples with higher interslab distances, the 100 and 200 N-doped samples delivered the capacity of 147 and 123 mAh g −1 at 2 • C, respectively, as compared to 116 mAh g −1 for the pristine sample. However, when tested at a rate of 2 • C, initial discharge capacities of 116, 107, 66, 50, and 41 mAh g −1 were obtained for pristine and 1, 1.3, 2, and 4 wt.% F-doped sample, respectively. Except for a fluoride containing surface layer with a low ionic conductivity, it was also found that the transition metal (TM) reduction led to the contraction of the Li-O interslab distances, and an increase in cation site mixing with increasing fluoride concentration was observed, thereby hindering the performance at such high C-rates. To suppress the increase in charge transfer resistance and NCA material pulverization during cycling, NCA was prepared with a 50 nm thick AlF 3 layer [337]. When cycled at 100 mA g −1 and 55 • C, AlF 3 -coated NCA delivered a capacity retention of 84.7% compared to that of PNCA (79.1%) after 100 cycles. Subsequently, a full cell was constructed using graphite as the anode material and capacity retentions of 66.5% and 86.2% were obtained for bare and AlF 3 -coated NCA after 1000 cycles. The stability of NCM811 at highvoltage was enhanced by ALD AlW x F y /AlF 3 coating [338]. The impedance rise was mitigated by ∼80% compared to the NMC811 baseline after ∼300 h of high-voltage exposure during cycling. The NCM811 electrode was prepared with LiAlF 4 solid thin film interfacial layer to realize the wide electrochemical window [339]. This electrode showed excellent capacity retention within 100 cycles at 200 mAh g −1 , 50 • C with an electrochemical window of 2.75-4.50 V vs Li + /Li, which was ascribed to the LiAlF 4 interfacial layer not only improving cycle stability through reduced parasitic reactions, but also maintaining satisfactory ion conductivity. The structural degradation and intergranular cracks of NCM622 were mitigated by constructing a LiAlO 2 /LiF and AlF 3 hybrid coating layer [340]. Benefiting from the synergistic effect of LiAlO 2 with high ionic conductor and LiF and AlF 3 in suppressing the attack by HF, the NCM622@LiAlO 2 /LiF and AlF 3 displayed a discharge capacity of 124.3 mAh g −1 with a capacity retention of 74.5% after 300 cycles at 5 • C in the voltage window of 2.7-4.5 V. However, NCM622 and NCM622@LiAlO 2 showed discharge capacities of close to 7.5 and 88.1 mAh g −1 after 300 cycles at 5 • C, respectively. A sol venting-out crystallization process was used to form a 10-20 nm thick FeF 3 layer on the surface of NCA, accordingly, the thermal stability of NCA was enhanced [341]. This composite delivered a capacity of 105.4 mAh g −1 at 1 • C, 55 • C over 100 cycles with capacity retention of 57.8%, which was higher than that of bare NCA (capacity retention of 45.2%). This enhanced thermal stability was ascribed to the protection effect of the FeF 3 coating layer, which suppressed the reactivity of Ni 4+ ions and the release of oxygen through prohibiting the oxidized cathode from being directly contacted with electrolyte solution.

Conclusions and perspectives
As a promising alternative cathode for LIBs, Li[Ni x Co y Mn 1−x−y ]O 2 (NCM) and Li[Ni x Co y Al 1−x−y ]O 2 (NCA) (x ⩾ 0.8) nickel-rich cathode materials have received noteworthy achievements in commercial applications, due to their low-cost and high theoretical capacities. However, large amount of cation disorders were formed in the cathode materials during the charge/discharge process, which would inevitably lead to phase transition from layered to disordered spinel/rock-salt structure and microcracks following severe capacity loss. In addition, other issues include oxygen loss, transition metal dissolution, side reactions between electrode and electrolytes, as well as thermal instability under high voltage and deep charging/discharging. In recent years, many efforts have been devoted to alleviate these problems and obtained considerable progresses by optimizing secondary particles, pre-oxidating of precursors, designing electrolyte additives, introducing extrinsic element with strong electrons, or designing 2-5 µm single crystals. In this review, we summarize these works in order to provide a clear understanding about the present achievements. According to the previous reports, the synthesis of core-shell/concentration gradient secondary particles is an effective method to mitigate Li + /Ni 2+ mixing and particle crack with side reactions during synthesizing and cycling process. The typical core-shell structures of concentration gradient composites are high nickel structure coated on the low nickel structure. For modifying the surface structure of precursors, one Na 2 S 2 O 8 is commonly the oxidizing agent for the enhanced oxidation state of Ni. The poor cycle stability of nickel-rich system can be partly ascribed to the anisotropic shrinkage/expansion breaking layered structure. The results indicate that by simply synthesizing the coreshell secondary particles with a concentration gradient, the cycle stability of NCM/NCA towards high capacity can be significantly enhanced. In addition, designing various electrolyte additives, such as lithium bis(oxalato)borate (LiBOB) additive in the LiPF 6 can stabilize the interface, is also an effective way to improve cycling stability. Most of the nickel-rich cathode materials exhibit a combination of phase transition and oxygen release when charging/discharging. Integrating extrinsic element into bulk and surface can control the degree of cation disorder and realize a higher reversible reaction. Rational improvements to the mechanical stability of nickel-rich cathode materials focused mostly on three aspects: increasing the layered distance to improve Li + diffusion path; constructing strong metal-oxygen bonds to improve the structure stability or electric conductivity of materials, such as Zr-O, Ta-O, Ce-O, etc; designing an ionic conductor with suitable void space to use its inner void space improving the lithium ions diffusion, such as LiZrO 2 and LBP coating structure. Moreover, optimization of the synthesis process, such as processing active material into micron-sized single crystals, also exhibits a good promising in applications. Single crystal-based cathode materials possess the performance of low specific surface area and outstanding mechanical stability, which is capable of maintaining particle integrity and eliminating intergranular cracks during charge/discharge processes. As the price of Co is high, combining the effective modified strategies and high voltage performance, low/zero Co materials show the most promising for practical application.
Although some effective measures have been taken to improve the capacity and cycle stability of nickel-rich cathode materials in LIBs, there are still several challenges to be addressed. Research on the degradation mechanism of materials can provide guidance for rational design improvement strategies, but there are still some degeneration mechanisms to be clarified, and there also exists some controversial perspectives. The in-situ and ex-situ characterizations combined with theoretical calculation is a promising method to further explore the degeneration mechanisms. Moreover, many modified nickel-rich cathode materials displayed excellent performance in half cells, but their application in full LIBs should be noticed in the future. There is no doubt that the performance of nickel-rich cathode materials in full LIB is crucial for practical applications.