Aluminum and Indium Co‐Modified Ni‐Rich Layered Cathode with Enhanced Microstructure and Surface Stability for Lithium‐Ion Batteries

Ni‐rich layered oxides are considered as promising cathodes for next‐generation lithium‐ion batteries. However, they still suffer microstructure and surface instability particularly under high operating voltage, leading to rapid capacity fading and battery failure. In this context, Ni‐rich layered cathodes (LiNixCoyMn1−x−yO2; LNCM) with aluminum and indium co‐modified crystal and surface structures are developed by a simple one‐pot calcination approach. Battery tests show that the Al and In co‐modified LNCM electrodes demonstrate remarkably enhanced rate capability and cycling stability compared with the pristine LNCM, Al‐doped LNCM, and In‐modified LNCM counterparts. Further characterizations reveal a simultaneous suppression of cracking and resistive film growth. The improved microstructural and surface stability originate from the synergistic functions of Al and In co‐modification. The incorporation of Al3+ into transition metal slab significantly reduces the Li+/Ni2+ antisite, which noticeably mitigates the undesired layer to rock‐salt phase transformation. The In3+ dopant dispersed in Li interslab can dissipate the anisotropic lattice strain, enabling greatly improved reversibility of H2↔H3 phase transition occurred in delithiation‐lithiation processes. Meanwhile, the synchronously formed LiInO2 adherent coatings deplete lithium residues, facilitate lithium‐ion transfer, and resist electrolyte corrosion. Microstructure and surface engineering through Al and In co‐modification offer a promising design strategy for Ni‐rich layered cathodes.


Introduction
Ni-rich layered oxides have been considered as one of the most promising cathodes for next-generation lithium-ion batteries to meet the rising demand from electric vehicles and grid energy storage fields, attributed by their large theoretical DOI: 10.1002/admi.2023005789] Cation antisite and anisotropic lattice strain are the two culprits for the microstructure instability of Ni-rich LNCM.Because of the comparable ionic radius between Li + (0.076 nm) and Ni 2+ (0.069 nm), Ni 2+ tends to diffuse into Li interslab and occupy Li sites.The Li + /Ni 2+ antisite defect forms during synthesis, further aggravated in the repeated charging processes.[12][13] To overcome these obstacles, heteroatom lattice doping has been proposed as an effective microstructure engineering methodology to boost the bulk structure stability. [14,15]Among various doping elements, Al is of continuing interest.[22] Near the charge-end in particular under high operating voltage, H2→H3 phase transition accompanied by an abrupt lattice collapse along crystallographic c-axis direction takes place, which triggers internal microcracks and allows the microcracks to propagate to the surface. [23]Substantial efforts have been made to constrain the microcrack formation, like doping, [24,25] constructing textured morphology, [26] building protective coatings in primary particle level. [27,28]Among them, microstructure modulation through incorporating alien cations into Li interslab is a good strategy to dissipate the anisotropic lattice strain.However, the alien cation's elemental species, radius, ionic valence, and doping content should be carefully examined, otherwise, its disadvantageous impacts such as blocking the lithium diffusion path, increasing the antisite defects and capacity loss will dominate.
Surface degradation of Ni-rich LNCM cathodes also has a great influence on the electrochemical performance.[31][32] To relieve the adverse surface deterioration caused by the aforementioned side reactions, surface coating on cathodes is believed to be an effective strategy.35][36][37][38][39][40] Preparation of coatings through post-treatment on an available LiNi x Co y Mn 1−x-y O 2 has been widely adopted, but such method is complicated, and usually gives rise to weak bond between the coatings and host interface, resulting in the detachment of decorated layers.Therefore, constructing tightly adherent coatings via simple method is highly demanded for surface engineering on Ni-rich cathode materials.
Considering one single approach could hardly address both the microstructure and surface challenges of Ni-rich cathodes, a proper combination of different modification routes that can work synergistically is preferable. [41,42]In this work, we employed a simple one-pot calcination method for preparing the Al and In co-modified LiNi 0.8 Co 0.1 Mn 0.1 O 2 .The Al 3+ dopant in transition metal slab and In 3+ dopant in Li interslab collaboratively alleviate the cation antisite and anisotropic lattice distortion noticeably.Meanwhile, the synchronously formed LiInO 2 adherent coatings deplete lithium residues, facilitate lithium-ion transport, and isolate the cathode from electrolyte and moisture air.Benefiting from such crystallographic and surface modulations, the undesired microstructure instability, mechanical cracks, and surface deterioration have been effectively mitigated during cycling.Compared with pristine LNCM, Al-doped LNCM and Inmodified LNCM, the Al and In co-modified LNCM cathode exhibits improved cycling stability and rate capability even under high operating voltage conditions.

Materials Preparation
The Al and In co-modified LNCM cathode materials were synthesized by a simple one-pot calcination method, as depicted in Scheme 1.During the wet grinding, an ethanol solution dissolved with Al(NO 3 ) 3 and In(NO 3 ) 3 infiltrate throughout the Ni 0.8 Co 0.1 Mn 0.1 (OH) 2 precursor.With ethanol evaporation, nitrate salts recrystallize and most of them deposit on the secondary particle surface of hydroxide precursor due to the coffee ring ef-fect.Under subsequent high temperature calcination, the LiOH molten fluid together with Al and In sources penetrates deeply into the bulk, and reacts with Ni 0.8 Co 0.1 Mn 0.1 (OH) 2 precursor to produce the Al and In co-modified LNCM materials.

Materials Characterization
The XRD patterns of as-synthesized Li 1−x In x Ni 0.8-y Al y Co 0.1 -Mn 0.1 O 2 samples, including pristine LNCM, Al0.5%-LNCM, In1%-LNCM, Al0.5%In0.5%-LNCM,Al0.5%In1%-LNCM, Al1%In0.5%-LNCM and Al1%In1%-LNCM are shown in Figure 1 and Figure S1, Supporting Information.Diffraction peaks of all LNCM samples can be indexed to the hexagonal -NaFeO 2 -type layered structure with a space group R-3m.The distinct splitting of (006)/(102) and (108)/(110) diffraction peaks indicates their well-ordered layer structure with high crystallinity. [26]rom Figure 1c-e, it can be observed that typical (003) and (110) Bragg reflections of Al0.5%In1%-LNCM shift slightly to higher 2 values compared with pristine LNCM, implying decreased c and a lattice parameter, which coincided with the Rietveld refinement as shown in Table 1.The slight contraction in a-axis lattice parameter is indicative of the successful incorporation of Al 3+ into the transition metal slab of LNCM crystal lattice because the ionic radii of Al 3+ (0.0535 nm) are slightly smaller than that of Ni 3+ (0.056 nm).Whereas the reduction in c-axis lattice parameter is believed to be ascribed to In-doping.Since the ionic radii of In 3+ (0.08 nm) are comparable to Li + (0.076 nm) but much larger than all the transition metals in LNCM (Ni 3+ 0.056 nm, Co 3+ 0.0545 nm, Mn 4+ 0.053 nm), In 3+ is hardly to enter the transition metal slab, thus Li interslab of LNCM layered structure is proposed as the possible In 3+ doping site.It is worth noting that the length of O-(Li 1−x In x ) depends not only on the ionic radius, but also on the bonding manner.Although the ionic radii of In 3+ are slightly larger than Li + (0.08 versus 0.076 nm), the electronegativity of In (1.78) is much higher than Li (0.98).With In-doping, the charge density of oxygen layers would decrease, and the O─(Li 1−x In x ) bonds become more covalent, leading to weakened electrostatic repulsion between oxygen layers, and thus reduced lattice parameter along c-axis direction.It is known that the reducing of Li interslab would impede Li + diffusion to some extent.However, the rate performance of LNCM cathode is not determined just by Li layer spacing.Other factors, such as the Ni 2+ antisite concentration in Li interslab, microcracks, and surface reconstruction involving the formation of various resistive films can also affect the cathode rate capability.In fact, benefiting from the synergetic effect of Al and In comodification, the cathode exhibits improved rate performance.More details will be discussed later.
By comparing Figure 1a,b, it is obvious that the peak intensity ratio I (003) /I (104) of Al0.5%In1%-LNCM (1.84) is much higher than that of pristine-LNCM (1.38), disclosing fewer cation antisite defects in the Al and In co-modified LNCM materials. [35]This result agrees well with the Rietveld refinement analysis (Table 1), which quantitatively shows that Al0.5%In1%-LNCM possesses a favored much lower degree of Li + /Ni 2+ mixing than that of pristine-LNCM.Al-doping is very effective in reducing Li + /Ni 2+ mixing owing to the less electronegativity of Al (1.61) than Ni Scheme 1. Schematic illustration of the preparation procedure.
(1.91). [43]With Al-doping into TM slab, the charge density of oxygen layers is increased, enabling TM─O bonds become more ionic, which would suppress the migration of Ni 2+ from TM slab to Li interslab.
Moreover, from Figure 1f and Figure S1c, Supporting Information, it is found that when the In content is increased from 0.5 to 1 mol%, diffraction peaks corresponding to LiInO 2 secondary phase appear.According to the Rietveld refinement for Al0.5%In1%-LNCM, the weight percentage of LiInO 2 secondary phase is 0.498% (Table 1).Based on this data, a maximum doping amount of In is calculated to be 0.7 mol%.The limited In 3+ doping may stem from its high charge state.For Al0.5%In0.5%-LNCMand Al1%In0.5%-LNCMsamples, all of the In elements are incorporated into Li interslab of the LNCM lattice, and no LiInO 2 phase was detected.In contrast, for Al0.5%In1%-LNCM and Al1%In1%-LNCM samples, 70% portion of In 3+ substitute Li + -site, while the rest of In element remaining on the Ni 0.8 Co 0.1 Mn 0.1 (OH) 2 precursor surface participate in the in situ formation of LiInO 2 phase.
The FE-SEM images of all the as-synthesized Li Each secondary particle consists of fine-grained polyhedral primary particles, but their surface morphologies are dramat-ically different.Under both FE-SEM and TEM observations (Figure 2a-c), a continuous amorphous out layer with nanoscale fuzzy morphology and a thickness of about 4 nm corresponding to the impure lithium residues (LiOH/Li 2 CO 3 ) is identified on the surface of crystalline pristine LNCM. [29]On the contrary, Al0.5%In1%-LNCM presents a relatively clean and smooth surface (Figure 2d).EDS mapping from FE-SEM (Figure 2e) displays the uniform distribution of Ni, Co, Mn, Al, In, and O elements on the surface of Al0.5%In1%-LNCM.To confirm the presence of LiInO 2 in Al0.5%In1%-LNCM as indicated by XRD, TEM, and HRTEM characterizations were undertaken.As shown in Figure 2f, a uniform coating layer composed of nanoscale particles with diameter of ≈10 nm is tightly attached on the bulk surface.Detailed HRTEM observations indicate two sets of lattice fringes in Al0.5%In1%-LNCM.In surface region I as shown in Figure 2g, lattice fringes with interplanar spacing of 0.215 nm can be clearly distinguished, which is indexed to LiInO 2 growing along (200) direction.In region II as shown in Figure 2h, lattice fringes with interplanar spacing of 0.235 nm belong to the characteristic (012) plane of Ni-rich LNCM materials.Taken from the edge of region II (Figure 2i), the selected-area electron diffraction (SAED) pattern along [4 41 ] zone of Ni-rich layered structure is obtained.The uniform LiInO 2 coatings with layered crystalline structure develop synchronously with the in situ generation of LNCM, which not only eliminates harmful lithium residues, but also is expected to protect the cathode from electrolyte corrosion but without sacrificing Li + transfer mobility.
To confirm the co-modification of Al and In, and to investigate the surface evolution from pristine to Al0.5%In1%-LNCM, the chemical states of In 3d, Al 2p, Ni 2p, Co 2p, Mn 2p, and O 1s are evaluated by XPS tests.For Al0.5%In1%-LNCM, the In 3d spectrum(Figure 3a) shows two peaks at 444.1 and 451.7 eV, corresponding to In 3d 5/2 and In 3d 3/2 spin-orbit doublet, respectively.The Al 2p peak (Figure 3b) is found to be located at 73.8 eV, but its signal is not obvious.This is because Al 2p overlaps with Ni 3p spectrum in the binding energy region between 65 and 81 eV, moreover, since the doping amount of Al is much lower than the content of Ni in the modified LNCM sample, the Al 2p peak becomes not prominent.In pristine LNCM, both Al and In signals are absent.Figure 3c reveals the Ni chemical environment and the percentage of Ni at different valence states on the surface of pristine and Al0.5%In1%-LNCM.Obviously, the relative content of Ni 3+ to Ni 2+ on the surface of Al0.5%In1%-LNCM is higher than that of pristine LNCM.Considering the substitution of Li + with high-valence In 3+ , to maintain the charge balance, the relative content of Co 2+ (Figure 3d) as well as Mn 3+ (Figure 3e) on the surface of Al0.5%In1%-LNCM is increased correspondingly.Therefore, the XPS results further verify that our Al and In co-modification strategy effectively prohibits the formation of Ni 2+ on material surface and lowers the Li + /Ni 2+ disorder, which is consistent with the XRD results.There is no noticeable difference between the O 1s spectra of pristine and Al0.5%In1%-LNCM (Figure 3f).

Electrochemical Performance
The CV profiles of pristine LNCM, Al0.5%-LNCM, In1%-LNCM, and Al0.5%In1%-LNCM for the first three consecutive cycles between voltage of 2.7 and 4.3 V at a scan rate of 0.1 mV s −1 are presented in Figure S3, Supporting Information.Except the initial cycle, the subsequent cycles of all electrodes exhibit typical three pairs of redox peaks, corresponding to a series of phase transitions between hexagonal (H1) and monoclinic (M), monoclinic (M) and hexagonal (H2), hexagonal (H2) and hexagonal (H3). [7]Owing to the solid-electrolyte interphase formation and the lack of Li + during the first charge process, only one dominant peak followed with H2→H3 oxidation peak is identified in the first anodic sweep.The potential interval between the initial anodic peak and its corresponding cathodic peak of pristine LNCM is 274 mV, which is remarkably larger than that of Al0.5%-LNCM (235 mV), In1%-LNCM (199 mV) and Al0   we compared its rate capacities with pristine LNCM, Al0.5%-LNCM (Figure S5, Supporting Information), and In1%-LNCM as shown in Figure 4b.Because both Al and In components are believed not involved in the redox reactions, the initial discharge capacities at 0.1 C of Al0.5%-LNCM and In1%-LNCM are slightly lower than that of pristine LNCM.But to our surprise, the optimized Al0.5%In1%-LNCM manifests a much higher initial discharge capacity of 193.7 mAh g −1 than that of pristine (182.5 mAh g −1 ), which definitely declares the superior synergistic effect afforded by Al and In comodification rather than single Al or In doping.In addition, the first-cycle coulombic efficiency also shows an improvement from 77.7% of the pristine LNCM to 85.0% of Al0.5%In1%-LNCM.The detailed initial charge-discharge voltage profiles at 0.1 C are shown in Figure 4c.When the current rate reaches 5 C, the discharge capacity of Al0.5%In1%-LNCM is still higher than that of the pristine LNCM, Al0.5%-LNCM, and In1%-LNCM.The Al0.5%In1%-LNCM cathode delivers the highest discharge capacities of 163.9 and 149.6 mAh g −1 at 2 and 5 C, while the pristine LNCM can only provide 126.7 and 82.3 mAh g −1 , respectively.As shown by their discharge profiles at various rates from 0.1 to 5 C in Figure S4b, Supporting Information, a much more severe polarization is developed in pristine LNCM than in Al0.5%In1%-LNCM.
Figure 4d displays the cycling performance of pristine LNCM, Al0.5%-LNCM, In1%-LNCM, and Al0.5%In1%-LNCM electrodes at a low current density of 0.2 C. As observed, Al0.5%In1%-LNCM delivers the highest initial discharge capacity of 187.1 mAh g −1 , and after 100 cycles, its capacity retention is as high as 96.8% with a remaining capacity of 181.2 mAh g −1 , whereas pristine LNCM cathode retains 81.5% of its initial discharge capacity with 143.7 mAh g −1 being left.Figure 4e compares the cycling performance at a medium current density of 1 C.After 200 cycles, the capacity retention rate of pristine LNCM, Al0.5%-LNCM, In1%-LNCM, and Al0.5%In1%-LNCM cathodes are 72.6%,81.8%, 82.9%, and 86.7%, respectively, demonstrating that Al0.5%In1%-LNCM with optimum co-modification content possesses the best cycling behavior.In order to further validate the advantages of Al and In co-modification strategy, the cycling performances of pristine LNCM, Al0.5%-LNCM, In1%-LNCM, and Al0.5%In1%-LNCM at a high current density of 5 C, and under high operating voltage of 2.7-4.5 V are investigated.After 200 cycles at 5 C, the discharge capacity of pristine LNCM only remains at 69.4 mAh g −1 , in contrast, a much higher discharge capacity of 120.5 mAh g −1 is still preserved in Al0.5%In1%-LNCM cathode (Figure 4f).It is noted that the cyclic capacity curve of Al0.5%In1%-LNCM at 5 C displays relatively large fluctuation, and the possible main reason is supposed to be temperature influence.Detailed analysis and discussion can be found in Figure S6, Supporting Information.The initial charge-discharge voltage profiles of pristine LNCM, Al0.5%-LNCM, In1%-LNCM, and Al0.5%In1%-LNCM cathodes between 2.7 and 4.5 V at 0.1 C are shown in Figure 4g.While all these electrodes exhibit similar curve shapes, Al0.5%In1%-LNCM shows the highest initial discharge capacity (202.5 mAh g −1 ), and its coulombic efficiency is also higher than that of other samples.Figure 4h   can be delivered by pristine LNCM.Their discharge profiles at the 200th cycle under 2.7-4.5 V and 1 C are shown in Figure 4i, which further confirms the stabilization effect afforded by Al and In co-modification.

Mechanism of Enhanced Cycling Stability and Rate Capability
In order to reveal the different cycling behavior between pristine and Al0.5%In1%-LNCM cathodes, the differential capacity (dQ dV −1 ) profiles were obtained by differentiating the 5th, 20th, 50th, and 100th charge-discharge curves (Figure 5a,b).Both electrodes show similar multiple phase transitions during delithiation-lithiation processes, which is consistent with the CV results.However, for the pristine LNCM electrode as shown in Figure 5a, the intensity of H2↔H3 peak decreased rapidly, and the polarization voltage significantly increased from 42 mV at the 5th cycle to 118 mV at the 100th cycle, implying the irreversible structural damage during cycling.Upon deep Li + extraction at the end of charging, the final H2→H3 phase transition is reported to be responsible for the abrupt anisotropic lattice contraction in caxis direction, which induces intrinsic mechanical strain on the internal particles, leading to microcracks nucleation then propagation, and consequently the capacity fading. [22,23]In contrast, the intensity and position of H2↔H3 peak are well maintained in Al0.5%In1%-LNCM electrode with cycles (Figure 5b), demonstrating that the irreversibility of structural evolution is effectively mitigated through Al and In co-modification.
The electrode-electrolyte interface is the key area for Li + /e - transfer and parasitic reactions.Therefore, the electrochemical impedance variations were monitored during cycling to further understand the mechanism behind the improved electrochemical performances.As shown in Figure 5c,d, all the measured Nyquist plots consist of two depressed semicircles followed by an inclined line.The semicircle in the high frequency region is generally interpreted as the surface film resistance (R film ), and the one at medium frequency represents the charge-transfer resistance (R ct ), while the inclined line in the low frequency region is related to the Warburg impedance (W) describing the solid-state lithium-ion diffusion.The EIS spectra are simulated using equivalent circuit depicted in the inset of Figure 5c, and the fitting results are listed in Table 2.As shown in Figure 5c, for pristine LNCM after 1 cycle, the surface film resistance R film (61.33 Ω) contributes most to the total electrode impedance.On the contrary, the R film value of Al0.5%In1%-LNCM after 1 cycle (12.87Ω) is nearly four times smaller than that of pristine LNCM, proving that the depletion of insulating lithium residues significantly reduces the Li + migration resistance across the electrodeelectrolyte interface.In addition, the Al0.5%In1%-LNCM after 1 cycle displays nearly identical R ct value (28.96Ω) to that of pristine (30.91 Ω), indicating that the influence of LiInO 2 coatings on the interfacial charge-transfer resistance is negligible.
By comparing Figure 5d with Figure 5c, after 100 cycling, a huge increment of R ct from 30.91 to 138.9 Ω is observed on pristine LNCM, while there is only a minimum R ct growth for cycled Al0.5%In1%-LNCM (from 28.96 to 41.06 Ω), demonstrating that the electron transport in cycled pristine LNCM is dramatically degraded at the electrode level.In order to explore the variations of Li + mobility during cycling, the diffusion coefficients of lithiumion (D Li + ) are estimated from the EIS profiles in low frequency re-gions (Figure 5e). [36]As listed in Table 2, although Al0.5%In1%-LNCM always shows higher D Li + values than that of pristine either after 1 or 100 cycles, for each electrode the D Li + value change after cycling is negligible.Therefore, the minimum impedance growth during cycling afforded by Al and In co-modification is mostly in the form of notably reduced ohmic loss, which means that the improved electrochemical performance of Al0.5%In1%-LNCM mainly originates from the persistent rapid electron transfer throughout the whole electrode.
To gain deeper insight into the origin behind the collaborative effect of Al and In co-modification on boosting the long-term cycling stability, especially under high-rate and high operating voltage for Ni-rich cathodes, the morphologies of cycled electrodes are also examined.After 100 cycles at 1 C, pristine LNCM and Al0.5%In1%-LNCM exhibit distinct microstructures.From the FE-SEM image of cycled pristine LNCM shown in Figure 6a, severe microcracks can be observed within the primary particles, and some of them even propagate to the particle surface.Meanwhile, intensive cracks throughout the secondary particles also appear.The intragranular microcracks within the primary Table 2. Fitting results of EIS and the calculated D Li + values of pristine and Al0.5%In1%-LNCM electrodes.[46][47] But in cycled Al0.5%In1%-LNCM (Figure 6c), the mechanical integrity is well preserved with no visible structural damage on its primary particles.The contrasting cracking behaviors further evidence the greatly improved reversibility of H2↔H3 structural evolution in Al0.5%In1%-LNCM as revealed by our dQ dV −1 analysis.The mitigated mechanical strain is supposed to be highly associated with the incorporation of In 3+ into the Li interslab.During the initial period of Li removal, strong In─O bonds limit the extent of c-axis expansion via reducing the oxygen-oxygen repulsion.Once H3 phase evolves, these In─O bonds serve as reinforcing beams dispersed in the crystal lattice to restrict the abrupt shrinkage along c-direction.Therefore, In 3+ doping effectively dissipates the anisotropic lattice strain, and avoids crack nucleation and propagation during cycling.

Sample 1st cycle 100th cycle
The HRTEM observations are carried out to assess the extent of surface damage on the pristine LNCM and Al0.5%In1%-LNCM cathodes after 100 cycles at 1 C.As shown in Figure 6b, the cycled pristine LNCM shows large areas of NiO-like rock-salt layer, which rapidly increases the charge-transfer resistance and hinders Li + transfer.On the contrary, the surface damage on cycled Al0.5%In1%-LNCM is substantially suppressed, and the layered structure is well preserved in the bulk region as shown in Figure 6d.]48,49] According to our aforementioned XRD results and Rietveld refinement, Al 3+ /In 3+ co-doping achieves a much lower degree of Li + /Ni 2+ antisite than that of pristine LNCM.Therefore, Al and In co-modification can suppress the undesired surface phase transformation, especially during high voltage charging, which also contributes to the significantly reduced ohmic loss for cycled Al0.5%In1%-LNCM as revealed by EIS results.Moreover, as shown in Figure 6d, the coating remains intact after 100 cycles, confirming the strong bonding built between LiInO 2 and bulk LNCM.
To verify the protective effect of LiInO 2 coatings on the cathode-electrolyte interface, XPS measurements are performed on pristine LNCM and Al0.5%In1%-LNCM after 100 cycles at 1 C as shown in Figure 6e.From the C 1s spectra, the deconvoluted peak at 289.8 eV corresponds to CO 3 2-species of Li 2 CO 3 and ROCO 2 Li, which arise from lithium residues and organic CEI components produced by electrolyte decomposition, while, other peaks come from the conducting carbon materials and binder in the electrode.The proportion of Li 2 CO 3 /ROCO 2 Li peak for the cycled Al0.5%In1%-LNCM is obviously lower than that of pristine LNCM, indicating that the LiInO 2 coatings can effectively restrain lithium residues formation and protect the electrode surface from electrolyte attack.Spectra of F 1s for both samples exhibit three peaks of CF 2 from the binder (688.5 eV), Li x PO y F z (687.4 eV), and NiF 2 /LiF (684.6 eV).The percentage of NiF 2 /LiF peak on the surface of cycled Al0.5%In1%-LNCM is lower than that of pristine LNCM, revealing that the LiInO 2 coatings can also slow down the production of CEI components of insulating metal fluorides.Due to the rapid growth of CEI and lithium residues, the Ni 2p signal can hardly be detected on the surface of cycled pristine LNCM electrode.As a result, on the basis of the above XPS observations, the adherent LiInO 2 coatings deplete harmful lithium residues and stabilize the cathode-electrolyte interface, which also contributes to the electrochemical performance improvement of Al0.5%In1%-LNCM.
Figure 6f presents the XRD patterns of pristine LNCM and Al0.5%In1%-LNCM cathodes before and after 100 cycles at 1 C with the corresponding magnified (003) and ( 006)/(102) peaks.The same peak positions of Al substrates from different electrodes declare that the measuring error is negligible.Although both cycled electrodes display the typical layered structure, the peak intensities of cycled pristine LNCM fade much more severely than Al0.5%In1%-LNCM, which verifies our aforementioned result that the bulk structure of pristine LNCM suffers massive destruction during cycling.Besides, it is observed that the (003) diffraction peak of Al0.5%In1%-LNCM shifts to lower angle after cycling (Δ2 ≈ 0.07°), implying a tiny increase of c lattice parameter owing to lithium deficiency.Meanwhile, the separation between (006) peak and (102) peak of cycled Al0.5%In1%-LNCM becomes smaller, which represents an increase of a lattice parameter due to the accumulation of Ni 2+ during cycling.Conversely, the (003) peak of cycled pristine LNCM moves toward higher angle (Δ2 ≈ 0.04°).The slight c-directional shrinkage upon repeated cycling is supposed to be relevant to the poor reversibility of H3→H2 phase transition during discharging, which would lead to the accumulation of untransformed H3 phase with collapsed interlayer distance.

Conclusions
In summary, we have developed an aluminum and indium codoped Ni-rich layered oxide with ionic permeable LiInO 2 coatings via a simple one-pot calcination method.The effect of such co-modification strategy on the microstructure, morphology, and electrochemical performance has been systematically investigated.By utilizing various measurements, we have revealed that the Al 3+ dopant in transition metal slab reduced the Li + /Ni 2+ antisite to a large extent, noticeably mitigating the undesired layer to rock-salt phase transformation which usually evolves from surface to bulk.The In 3+ dopant dispersed in Li interslab can restrict the anisotropic lattice expansion and shrinkage occurred in delithiation-lithiation processes, thus the intrinsic mechanical stress is effectively dissipated during cycling.Besides, the synchronously formed LiInO 2 adherent coatings deplete lithium residues and protect the cathode from side reactions with electrolyte and moisture air without sacrificing lithium-ion mobility.Under the synergistic microstructure and surface engineering afforded by Al and In co-modification, the undesired mechanical cracks and passive films accumulation including CEI, rock-salt phase, and lithium residues have been effectively mitigated during cycling.Compared with the pristine cathode, the Al and In co-modified cathode exhibits significantly suppressed impedance growth, thus demonstrating enhanced rate capability and cycling stability even under high operating voltage.Material Characterizations: The crystallographic structures and morphologies were characterized by XRD (D8 Advance, Bruker, with Cu K radiation), FE-SEM (Apreo HiVac, FEI), high-resolution TEM (JEM-2100F, JEOL), and SAED.The Rietveld refinement was performed using the GSAS program.The surface elemental distribution and chemical state were characterized by EDS and XPS (Thermo Scientific K-Alpha, Thermo Fisher).

Experimental Section
Electrochemical Measurements: The cathodes were fabricated by mixing 80 wt% active materials, 10 wt% Super P, and 10 wt% PVDF in NMP.The resulting slurry was cast onto aluminum foil, dried in vacuum, and cut into disks (ϕ14 mm).CR2032-type coin cells were assembled in an Arfilled glovebox, with Li foil as the anode, and Celgard 2400 PP membrane as the separator.The electrolyte was composed of 1.2 m LiPF 6 in 3:7 v/v EC/EMC with 2 wt% VC additives.CV experiment was carried out on a Potentiostat/Galvanostat (PARSTAT 4000A, Princeton Applied Research) in a voltage range of 2.7-4.3V versus Li/Li + at a scan rate of 0.1 mV s −1 .Galvanostatic charge/discharge tests were conducted on LAND-CT2001A instruments at various C rates (1 C = 200 mA g −1 ).The EIS measurements were taken over in a frequency range of 0.1 MHz to 10 mHz with an AC amplitude of 5 mV using PARSTAT 4000A.Based on Warburg impedance in the low frequency region, the diffusion coefficient of lithium ion (D Li + ) can be calculated according to the following equations. (1) wherein  is the Warburg coefficient, R is the gas constant, T is the absolute temperature, A is the surface area of cathode, n is the number of transferred electrons per molecule, F is the Faraday constant, and C is the concentration of lithium ions.The cycled batteries were disassembled in the glovebox, then the cathodes were washed with EMC, dried in vacuum, and analyzed by XRD, FE-SEM, high-resolution TEM, and XPS.
1−x In x Ni 0.8-y Al y Co 0.1 Mn 0.1 O 2 samples are shown in Figure 2a,d and Figure S2, Supporting Information.All the samples have the typical sphere shape with an average diameter of about 20 μm.
.5%In1%-LNCM (212 mV).Moreover, the polarization voltage in the third cycle of Al0.5%In1%-LNCM (85 mV) becomes the smallest one compared to that of pristine LNCM (127 mV), Al0.5%-LNCM (116 mV) and In1%-LNCM (96 mV), suggesting that appropriate Al and In co-modification can effectively restrain polarization and improve the electrochemical reversibility of Nirich LNCM cathode, thus the enhancement of rate performance and cycling stability is reasonably expected.The electrochemical performances of Al and In co-modified LNCM cathodes are investigated by a series of galvanostatic tests.
displays their cycling performances in the voltage range of 2.7-4.5 V at 1 C.After prolonged 200 cycles operated under such high voltage, Al0.5%In1%-LNCM still retains a satisfactory discharge capacity of 145.1 mAh g −1 , whereas only 83 mAh g −1 discharge capacity