Suppressing failure mechanisms in thick Ni-rich cathodes using angstrom-level alumina coatings

Surat Prempluem , Thitiphum Sangsanit , Worapol Tejangkura and Montree Sawangphruk *
Centre of Excellence for Energy Storage Technology, Department of Chemical and Biomolecular Engineering, School of Energy Science and Engineering, Vidyasirimedhi Institute of Science and Technology, Rayong 21210, Thailand. E-mail: montree.s@vistec.ac.th

Received 22nd April 2025 , Accepted 10th July 2025

First published on 10th July 2025


Abstract

Enhancing the long-term stability of Ni-rich layered oxide cathodes remains a central barrier to the commercialization of high-energy lithium-ion batteries. Despite progress in surface coating strategies, challenges such as poor uniformity, imprecise thickness control, and limited scalability continue to hinder practical deployment. Herein, we report a scalable, low-temperature atomic layer deposition (ALD) method for conformally coating industrial-scale, thick Ni-rich NMC811 electrodes with ultrathin alumina (Al2O3) layers. By precisely controlling the coating thickness via ALD cycles, we identify a 10-cycle coating (∼2 nm) as the optimal condition—achieving a significant improvement in capacity retention without compromising lithium-ion diffusion or rate capability. Comprehensive electrochemical and structural analyses, including in situ XRD, GITT, and DEMS, reveal that the alumina layer effectively suppresses microcracking, inhibits oxygen release, and minimizes transition metal dissolution—the three major degradation pathways in Ni-rich cathodes. Notably, the protective layer also mitigates electrolyte decomposition under high-voltage abuse conditions in commercial 18650-format cells. This work highlights the transformative potential of nanoscale interface engineering through ALD, offering a practical and industrially viable route to durable, safe, and high-performance lithium-ion batteries for next-generation energy storage systems.


1. Introduction

The rapid adoption of lithium-ion batteries (LIBs) in electric vehicles (EVs) has significantly escalated the demand for high-performance and cost-effective cathode materials. Currently, the cathode accounts for more than one-third of the total cost in LIBs, highlighting its critical role in overall battery economics. Conventional cobalt-rich cathodes, although widely utilized, encounter substantial limitations such as high raw material costs, limited resource availability, ethical concerns related to cobalt mining, and inherent instability at elevated operating voltages.1,2 To overcome these constraints, nickel-rich cathode materials have emerged as attractive alternatives, offering higher specific capacity and substantially reducing cobalt dependency.

Despite these compelling advantages, Ni-rich cathodes face significant challenges that severely affect their long-term stability and impede commercial scalability. Notably, degradation mechanisms including microcracking, lattice oxygen release, transition metal (TM) dissolution, and undesired phase transformations severely undermine their performance, particularly under the rigorous high-voltage and high-stress cycling conditions typical of EV operations.3–6 These degradation pathways accelerate battery failure, safety issues, and rapid capacity fading, necessitating robust mitigation strategies.

Surface modification via protective coatings has emerged as a highly effective strategy to enhance the durability and stability of Ni-rich cathodes. Among various coating materials, alumina (Al2O3) stands out due to its proven capability to effectively suppress microcracking, stabilize lattice structures, mitigate TM dissolution, and reduce unwanted side reactions at electrode–electrolyte interfaces.7 Moreover, alumina's abundance, environmental friendliness, and cost-effectiveness reinforce its suitability for industrial-scale applications.

Several approaches have been explored to deposit alumina coatings on cathode surfaces at a commercial scale; however, conventional methods such as mechanofusion and solution-based processes suffer from poor control over coating thickness and uniformity—critical parameters influencing cathode performance. For instance, Chenxi et al. found that mechanofusion of alumina on LiNiO2 was more effective when applied before lithiation rather than directly onto the active cathode material.8 In contrast, Krisara et al. demonstrated optimal performance with much thicker (150–200 nm) alumina coatings on NMC811 cathodes.9 Meanwhile, Sven et al. reported improved electrochemical stability using solution-based alumina coatings approximately 2 nm thick.10 Although these findings appear contradictory, they collectively underscore the vital importance of precise thickness control and coating uniformity—qualities not adequately addressed by conventional coating techniques.11

Atomic layer deposition (ALD) represents an advanced solution, distinguished by its unparalleled capability to deposit ultrathin, uniform coatings with precise nanoscale control, even on complex and porous cathode architectures.11–14 ALD-coated cathodes have consistently demonstrated improved structural integrity and enhanced electrochemical performance, as illustrated by prior studies on LiCoO2 electrodes that exhibited improved electron transport pathways and reduced overpotential.15 Nevertheless, most prior ALD investigations have predominantly focused on small-scale, thin-film cathodes, leaving a substantial knowledge gap concerning its effectiveness for thick electrodes crucial for achieving commercial-scale battery energy densities.16 Besides, previous studies on the cost breakdown and processing efficiency of the ALD process using TMA as a precursor have demonstrated its scalability for full-scale lithium-ion battery manufacturing.17

In the present study, we systematically investigate ALD-derived Al2O3 coatings on large-scale, thick Ni-rich NMC811 electrodes. We demonstrate the precise control of coating thickness via incremental ALD cycles, achieving uniform, conformal coverage over complex three-dimensional porous structures typical of commercial electrodes. Optimal electrochemical stability is obtained with an ultrathin alumina coating approximately 2 nm thick (10 ALD cycles), as validated by coin-cell experiments. Furthermore, we extend this methodology to commercial large-format 18[thin space (1/6-em)]650 cylindrical cells, confirming the scalability and industrial viability of the approach. In electrochemical abuse tests, ALD-coated electrodes significantly outperform their uncoated counterparts by showing reduced electrolyte decomposition, minimized TM dissolution (confirmed by ICP-OES), suppressed gas evolution (validated by differential electrochemical mass spectroscopy (DEMS) and proton nuclear magnetic resonance (1H NMR)), and a stabilized lattice structure evidenced through in situ X-ray diffraction (XRD) analyses.18–21 Moreover, post-mortem analyses after long-term stability testing—using HR-TEM and STEM—confirm that the coating layer remains uniformly adhered to the NMC surface, indicating its effectiveness in providing long-term electrochemical protection. Overall, this research provides compelling evidence that a precisely engineered, ultrathin alumina coating via ALD can effectively mitigate the critical degradation mechanisms of Ni-rich cathodes, enhancing their durability, safety, and performance in high-energy LIB applications.

2. Experimental

2.1 Production of large-scale electrodes at the pilot plant

The cathode slurry was prepared by thoroughly mixing the active material LiNi0.8Mn0.1Co0.1O2 (NMC811), polyvinylidene fluoride (PVDF), and carbon black (all procured from Gelon LIB Group, China) in a weight ratio of 95.2[thin space (1/6-em)]:[thin space (1/6-em)]2.4[thin space (1/6-em)]:[thin space (1/6-em)]2.4, respectively, with n-methyl-pyrrolidone (NMP) as the solvent. This slurry was then cast onto an aluminum foil current collector using a roll-to-roll semi-automatic coating machine and dried at 120 °C to remove the NMP solvent. After drying, the electrode was pressed at 9 tons, achieving a mass loading of approximately 20 mg cm−2.22 The anode slurry was prepared by mixing natural graphite, sodium carboxymethyl cellulose (CMC), styrene-butadiene rubber (SBR), and carbon black in a ratio of 95.4[thin space (1/6-em)]:[thin space (1/6-em)]1.2[thin space (1/6-em)]:[thin space (1/6-em)]2.4[thin space (1/6-em)]:[thin space (1/6-em)]1.0 respectively (all procured from Gelon LIB Group, China). The slurry was coated on Cu foil in the same way as that for the cathode, followed by pressing at 6.5 tons. All electrode fabrication was done under dry room conditions (dew point −40 °C).

2.2 Atomic layer deposition (ALD) coating on the large-scale electrode

The deposition of alumina (Al2O3) onto the as-prepared large-scale electrode was performed using an ALD system (MTI, Gelon LIB, China). 1 M trimethyl aluminum (TMA) in heptane from Thermo Scientific Chemicals and distilled water were used as the precursors. The ALD process followed a precise sequence: (1) TMA was pulsed for 0.5 s, (2) evacuation, (3) Ar purging for 90 s, (3) evacuation, (4) H2O was pulsed for 0.5 s, (5) evacuation, and (6) Ar purging for 150 s, followed by a final evacuation step.23 The reaction chamber was maintained at 135 °C, while the TMA and H2O precursors were kept at 80 °C and 70 °C, respectively.

2.3 Coin-cell fabrication

For the assembly of 2032-type coin cells, the electrode was punched into 12 mm diameter disks and vacuum-dried at 105 °C overnight to ensure complete moisture removal. The cell configuration included a single-layer polyethylene (PE) separator (Celgard) and lithium metal as the counter and reference electrode, with diameters of 18 mm and 14 mm, respectively. The electrolyte solution consisted of 1.2 M LiPF6 in a 1[thin space (1/6-em)]:[thin space (1/6-em)]4 volume ratio of fluoroethylene carbonate (FEC) to dimethyl carbonate (DMC), with 100 μL added to each cell. Following assembly, the cells were held at open-circuit voltage (OCV) for 12 h prior to electrochemical testing to ensure stabilization. All fabrication processes were done in a glove box (MBRAUN, Germany) with ultra-high purity Ar.

2.4 Production of the 18650 cylindrical cells and jelly rolls

The 18[thin space (1/6-em)]650 cylindrical cells were assembled under dry room conditions (dew point −40 °C). The cathode, separator, and anode were assembled by using a winding machine, followed by transfer to the glovebox for electrolyte injection (1.2 M LiPF6 in a 1[thin space (1/6-em)]:[thin space (1/6-em)]4 (FEC[thin space (1/6-em)]:[thin space (1/6-em)]DMC) ratio). The N to P ratio of the cells was 1.20, calculated based on the manufacturing specifications provided by GELON (China). Electrolyte was injected at 4.5 g for the 18[thin space (1/6-em)]650 cylindrical cells and 14 ml for jelly roll cells.

2.5 Electrochemical performance analysis

For coin cells, the cells underwent an initial formation process consisting of three charge/discharge cycles at a constant current (CC) rate of C/20 (where 1C = 180 mA h g−1) over a voltage range of 3.0–4.3 V vs. Li/Li+. Following formation, long-term cycling was conducted at a rate of C/2 for 499 cycles, with periodic performance assessments at C/20 in the 3rd, 49th, 96th, 199th, and 499th cycles. Rate performance testing was subsequently carried out across multiple C-rates (0.05C, 0.1C, 0.2C, 0.5C, 1C, 2C, and returning to 0.1C) under CC conditions within the same voltage range of 3.0–4.3 V. For the 18[thin space (1/6-em)]650 cylindrical cells, the cells were formed by using CC-charge at C/40 to 4.3 V and constant current constant voltage (CCCV) discharge at C/12.5 to 3.0 V, followed by a capacity determination step at C/10 and C/5 in a voltage window of 3.0–4.2 V. The electrochemical stability was tested under CCCV charge at C/2 to 4.2 V and CC discharge at 1C to 3.0 V.

2.6 Capacitance measurement to probe microcracking

The capacitance measurement was performed in a half-cell configuration using Li-metal as both the counter and reference electrodes.24 The experimental protocols are illustrated in Fig. 5a. The cells were initially tested using an Autolab (PGSTAT204), where they were charged at C/20 for 1 h, followed by discharge to 2.5 V at C/20, and then held in constant voltage (CV) mode for 6 h to achieve blocking conditions. This procedure, referred to as the conditioning step, ensured that all capacitance measurements were conducted at the same state of charge. During this state, potentiostatic electrochemical impedance spectroscopy (PEIS) was applied to gather data over a frequency range of 100 kHz to 1 Hz with an amplitude of 15 mV. Subsequently, the cells underwent a series of charge/discharge cycles, after which they were held in CV mode again for capacitance measurement. This cycle was repeated for a total of 10 charge/discharge cycles. The double-layer capacitance was determined by fitting the capacitive branch with an RQ element within the frequency range of 1 Hz to 100 mHz.24,25

2.7 Differential electrochemical mass spectroscopy (DEMS)

To investigate lattice oxygen release and parasitic reactions between the electrolyte and electrodes, differential electrochemical mass spectroscopy (DEMS) was conducted using a Hiden HPR-40 system (UK). The experiment utilized a jelly-roll configuration consisting of pristine or modified NMC cathodes versus graphite with an N/P ratio of 1.20. The jelly roll was prepared using an automatic winding machine in a dry room with a dew point of −40 °C. The jelly roll was placed in a glass vial filled with electrolyte, sealed with a septum, and connected to the DEMS system. The system was pre-conditioned for 12 h at 5 × 10−8 torr before analysis. The cell was first charged/discharged at C/20 to 4.30 V, followed by charging again to 4.3, where constant-voltage charge (CV) at 4.30 V for 6 h was applied. This step was repeated at 4.40, 4.50, 4.60, and 4.70 V while collecting gaseous products, which were ionized at 70 eV (500 μA).

2.8 X-ray photoelectron spectroscopy (XPS)

XPS (JPS-9010 MC, JEOL) with depth profiling was performed on the cathode electrode to characterize the Al2O3 coating thickness. The analysis used a Mg-Kα radiation source ( = 1253.6 eV) at 12 kV and 25 mA under a vacuum pressure of 10−7 Pa. Ar+ etching was performed in a series of time steps, with varying durations depending on the thickness of the Al2O3 layer. Atomic concentrations were determined after background subtraction using a Shirley-type background model.26

2.9 Inductively coupled plasma atomic emission spectrometry (ICP-OES)

For transition metal deposition on the graphite anode, the jelly roll cells were dissembled in an Ar glove box environment. The graphite material was collected at around 0.5 g under each condition, followed by digestion in 5 ml of nitric acid and 1 ml of hydrogen peroxide overnight. The digested solution was diluted with a matrix solution of 0.5% wt nitric acid, followed by a syringe filter before performing ICP-OES. For transition metal dissolution from the cathode in the electrolyte, the electrolyte was collected from the jelly roll for 3 ml, followed by dilution in matrix solution before performing ICP-OES.

2.10 Differential scanning calorimetry (DSC)

To assess the thermal stability enhancement provided by the ALD coating, post-mortem differential scanning calorimetry (DSC) measurements were performed on charged electrodes after long-term cycling. The electrodes were charged to 4.2 V using a constant-current (CC) protocol, followed by a constant-voltage (CV) hold at 4.2 V for 20 h.27 Approximately 3–5 mg of active material, including residual electrolyte, was sealed in a gold-plated high-pressure crucible with a gold-plated membrane lid. All sample preparation was conducted entirely within an Ar-filled glove box to prevent exposure to air and moisture.3,28 The thermal analysis was carried out using a DSC 8000 (PerkinElmer) with a heating rate of 10 °C min−1.

3. Results and discussion

3.1 Physicochemical properties

The electrode was fabricated via a semi-automated pilot-scale production process, as illustrated in Fig. 1a–h. This process encompassed slurry mixing, electrode coating, calendaring, slitting, tab welding, electrode winding, and electrolyte injection. The resulting 18[thin space (1/6-em)]650-format lithium-ion cells, depicted in Fig. 1i and j, reflect industrial-level manufacturing quality. The electrode (Fig. 1g) served as the substrate for subsequent ALD coating. The economic viability of the ALD process—particularly with TMA precursors—has been comprehensively evaluated in prior studies.17 These analyses provided detailed cost breakdowns for industrial-scale applications, demonstrating that while ALD may appear cost-prohibitive at the laboratory scale, significant cost reductions are achievable through economies of scale and industrial-grade chemical sourcing. Furthermore, the processing efficiency of ALD has been reported to meet throughput requirements compatible with current battery production lines.17 Collectively, these findings support the practical feasibility of integrating ALD into commercial lithium-ion battery manufacturing from both cost and operational efficiency standpoints.
image file: d5ta03170b-f1.tif
Fig. 1 Electrode preparation and the 18[thin space (1/6-em)]650 cylindrical cell production process: (a) slurry mixing, (b) electrode coating, (c) calendaring, (d) slitting, (e) tab welding, (f) electrode winding, (g) prepared electrode, (h) electrolyte injection, (i) assembled 18[thin space (1/6-em)]650 cylindrical cells and (j) cell testing setup.

To elucidate the uniformity of the electrode and coating layer on the thick, porous electrode, cross-sectional SEM images (Fig. 2a) show that the composite electrode has a thickness of approximately 88 μm produced at the pilot plant. Fig. 2b shows the top image of the electrode, showing that the electrode uniformly consists of the active material (NMC811), conductive carbon, and the binder (PVDF). To confirm the quality of coating on the 3D porous, thick electrode, energy dispersive spectroscopy (EDS) was also conducted on the cross-section of the electrode that was coated with 10 cycles of ALD, as shown in Fig. 2c–f. The consistent overlap of Ni, Al, and C signals suggests that TMA can diffuse effectively into the porous structure, indicating that ALD is a viable technique for coating complex, thick electrodes. Top surface SEM-EDS analyses were also conducted to investigate the whole particle properties on the top surface. As shown in Fig. S1–S3, for NMC-10ALD, NMC-25ALD, and NMC-50ALD, respectively, the Al intensity is also uniformly distributed throughout the overall spherical NMC particles as the Al signal overlaps well with the Ni signal. Moreover, Al2O3 was not only coated on the active material but also coated on the carbon compound (conductive carbon and the binder), as shown well in the overlaps between Al and C. This result also indicates that Al2O3 is not only on the surface of the active material but also on the surface of conductive carbon, where parasitic reactions can also occur.29,30 The elemental concentrations obtained from SEM-EDS (Tables S1–S3) reveal that the Al concentration increases proportionately to the number of ALD cycles, indicating the relationship between the cycle number and Al content.


image file: d5ta03170b-f2.tif
Fig. 2 SEM images of the NMC811 electrode: (a) cross-sectional view and (b) top view. EDS mapping of the cross-section of the NMC811 electrode coated with a 10-cycle alumina layer, highlighting the distribution of (c) nickel, (d) aluminum, and (e) carbon, with (f) the corresponding composite mapped image.

A high-resolution transmission electron microscope (HRTEM) was employed to investigate the atomic structure and thickness of the alumina coating. As shown in Fig. 3a, the pristine NMC811 particles display a pure layered structure where the magnified area, Fig. 3b, shows the feature of the layered structure with a d-spacing of 0.175 nm corresponding to the 105 planes which can be further confirmed using the corresponding fast Fourier transform (FFT) pattern (Fig. 3c). After coating with Al2O3 for 10 cycles, a thin alumina coating layer is visible as a brighter contrast at the surface as shown in Fig. 3d.31,32 The core material retains its layered structure, as indicated by d-spacings of 0.198 and 0.449 nm, which correspond to 104 and 003 planes, as shown in the enlarged views in Fig. 3e. In addition, the FFT pattern shown in Fig. 3f also demonstrates the characteristic pattern of a layered structure. Fig. 3g shows a magnified area of the shell material showing the amorphous feature of the alumina coating layer which was further confirmed using the FFT pattern in Fig. 3h. Increasing the number of ALD cycles results in a thicker alumina layer, as seen in the NMC-25ALD and NMC-50ALD samples (Fig. 3i and n respectively), while the core retains its layered structure (Fig. 3j, k, o and p) and the shells were in the amorphous phase (Fig. 3l, m, q and r). The amorphous structure of alumina synthesized in this work aligns with the previous literature that Al2O3 synthesized by the ALD technique is in an amorphous structure.32,33 The reason behind the amorphous structure of alumina is that the temperature is not high enough to create the crystalline phase.32,34 The coating thickness, determined from TEM images (Table S4 and Fig. S4), exhibits a linear growth rate of approximately 0.15 nm per ALD cycle, consistent with prior studies summarized in Table S5.


image file: d5ta03170b-f3.tif
Fig. 3 High-resolution TEM images of the active materials extracted from the electrodes: (a) NMC, (d) NMC-10ALD, (i) NMC-25ALD, and (n) NMC-50ALD. Enlarged views of the core material (NMC), highlighted with yellow squares, are shown in (b), (e), (j) and (o) for NMC, NMC-10ALD, NMC-25ALD, and NMC-50ALD, respectively, while the shell material (Al2O3), marked with green squares, is presented in (g), (l) and (q). Fast Fourier transform (FFT) patterns of the selected core areas are displayed in (c), (f), (k) and (p), and those of the shell areas are shown in (h), (m) and (r) for NMC, NMC-10ALD, NMC-25ALD, and NMC-50ALD, respectively.

The XRD patterns of NMC electrodes are illustrated in Fig. S5a; the peak pattern shows the characteristic of the α-NaFeO2 structure with the R[3 with combining macron]m space group. Fig. S5b–d shows the diffraction patterns of NMC-10ALD, NMC-25ALD, and NMC-50ALD, respectively, confirming that the ALD process does not affect the crystal structure of the NMC material, as no new peaks were observed after coating suggesting that alumina is in the amorphous phase, which is consistent with the HRTEM images. Moreover, the Rietveld refinement results are shown in Table S6; there are no significant changes in the lattice parameters and cation mixing, indicating that the Al2O3 coating layer did not significantly affect the lattice parameters.

XPS depth profiling was performed to assess the distribution of elements across the electrode surface. Fig. S6a–c shows the atomic concentrations of Al, O, C, and Ni in NMC-10ALD, NMC-25ALD, and NMC-50ALD tracked as a function of Ar+ etching time. The presence of residual carbon from the TMA precursor was minimal, as it disappeared after only 1 second of Ar+ bombardment, originating from incomplete oxidation of TMA that occurred at the low deposition temperature.35,36 Afterward, Ar bombardment was used to investigate the composition in the deeper layer of the material. All samples show reducing Al and O concentrations, which can be attributed to the thinner layer after being bombarded, which coincides with the increase in C and Ni concentrations originating from the conductive carbon, binder, and NMC, respectively. The thickness of the layers was determined from the bombarding time where the concentrations of Al and O started to become not parallel, implying that the oxygen concentration is not coming from alumina anymore but rather from the oxygen lattice. Moreover, at this point, C becomes almost constant, indicating that the bombarded layer reached the conductive carbon layer, which is also consistent with the initial appearance of Ni content coming from NMC. The bombarding time for thickness determination was indicated to be 22, 50, and 100 s for NMC-10ALD, NMC-25ALD, and NMC-50ALD respectively. The etching time shows a linear relation with the number of cycles, as shown in Table S7 and Fig. S7, coinciding with the thickness determined by TEM.

3.2 Electrochemical properties

The cycling performance of the NMC electrodes is shown in Fig. 4a. Among the samples, NMC-10ALD exhibited the highest cycling stability within the voltage range of 3.0–4.3 V. As the number of ALD cycles increased beyond 10 cycles, the cycling performance deteriorated, indicating that a thicker alumina layer hinders electrochemical performance. This suggests that 10 ALD cycles represent the optimal thickness for protecting the cathode while maintaining performance. Although the discharge capacity at C/2 at around 500 cycles seems to be very low as thickness increases, the discharge capacity in the check-up cycle (C/20) seems not different. This result suggests that the bulk structure was preserved as it can store lithium at a very low C-rate, further suggesting that the problems can be attributed to kinetic sluggishness after long cycling. The performance rate was evaluated through CC-charge discharge mode at 0.05C, 0.1C, 0.2C, 0.5C, 1C, 2C, and 0.1C, as shown in Fig. 4b. The result demonstrated comparable capacities for all samples up to a 2C rate, beyond which NMC-50ALD exhibited a noticeable decline in capacity. This result suggests that thicker alumina coatings act as kinetic barriers, impeding lithium-ion transport.12,33Fig. 4c shows the galvanostatic charge–discharge (GCD) profile of the first cycle, showing larger polarization as the thickness of the alumina increases, obviously at 25ALD and 50ALD. However, this larger polarization diminishes in subsequent cycles. Fig. 4e suggests that there is some transformation that occurs during the initial cycle, and the main charge storage mechanism of the NMC material does not alter. Moreover, Fig. 4d shows larger overpotential as the thickness of alumina increases. The reason is that alumina is considered an insulator, which would require more potential for lithium-ion to diffuse through the layer.15,37 However, Fig. 4f shows disappearance of this overpotential in the following cycle, indicating that this insulator has transformed into a conductive layer.12,38–40 This finding is consistent with previous reports, which have shown that only ultrathin coatings are necessary to protect the cathode while maintaining transport properties.41
image file: d5ta03170b-f4.tif
Fig. 4 Electrochemical performance in coin cells of NMC samples: (a) capacity retention of NMC (black), NMC-10ALD (green), NMC-25ALD (blue), and NMC-50ALD (red), (b) rate performance of NMC, NMC-10ALD, NMC-25ALD, and NMC-50ALD at various rates (0.05C, 0.1C, 0.2C, 0.5C, 1C, 2C, and 0.1C), (c) first charge–discharge profiles, with enlarged sections highlighting (d) overpotential, and (e) second charge–discharge profiles, with enlarged sections highlighting (f) overpotential. (g) The electrochemical performance of the 18[thin space (1/6-em)]650 cylindrical full cells with a graphite anode is demonstrated through capacity versus cycles.

The differential capacity (dQ/dV) plots for NMC, NMC-10ALD, NMC-25ALD, and NMC-50ALD are presented in Fig. S8a–d. These plots reveal the characteristic redox peaks corresponding to the H1–M, M–H2, and H2–H3 phase transitions, which are intrinsic to the charge storage mechanism of layered NMC cathodes. The presence of these peaks across all samples confirms that the Al2O3 coating does not disrupt the fundamental electrochemical behavior of the NMC material. As cycling progresses, a gradual shift of the redox peaks toward higher potentials during delithiation is observed—particularly in the H1–M region, as shown in the enlarged views in Fig. S9a–d. Notably, the peak shift for NMC-10ALD after prolonged cycling is only 0.0009 V, in contrast to a shift of 0.0208 V for the uncoated NMC, indicating superior structural and electrochemical stability of the coated sample.42

Moreover, the H2–H3 phase transition—often associated with microcrack formation due to pronounced lattice distortion—is more reversible in the NMC-10ALD sample, as evidenced by the reduced peak separation.43,44 This suggests that the optimized Al2O3 coating helps maintain lattice integrity and mitigates stress-induced degradation.

To further evaluate the thermal robustness of the coated cathode, cycling tests were conducted at an elevated temperature of 60 °C, where accelerated degradation of Ni-rich cathodes is typically observed.3,45 As shown in Fig. S10, the NMC-10ALD sample exhibited significantly improved capacity retention, even after only a few cycles, underscoring the effectiveness of the alumina coating in enhancing cathode stability under thermally stressful conditions.

To further strengthen the understanding, ALD was applied to 18[thin space (1/6-em)]650 full-cell configurations with a graphite anode to evaluate its feasibility for large-scale production. The properties of the electrode used in cell fabrication are presented in Table S8. The coulombic efficiency is shown in Fig. S11a, while the charge–discharge profile during the first formation cycle is depicted in Fig. S11b. Capacity determination cycles at C/10 and C/5 are illustrated in Fig. S11c and d, respectively. The electrochemical stability, shown in Fig. 4g, was assessed using CCCV charging at C/2 and CC discharging at 1C within a voltage range of 3.0–4.2 V. The results indicate that discharge capacity remains comparable for both samples until approximately 575 cycles, after which the capacity of the uncoated NMC rapidly declines, whereas NMC-10ALD maintains its performance. After 1000 cycles, NMC-10ALD can retain capacity at 71.20% while non-coated NMC remains at only 63.33%. For comparative analysis, the electrochemical performance metrics from previously reported coating strategies are summarized in Table S9. These findings further highlight the ability of the ultrathin alumina layer to protect the cathode from degradation at the commercial cell level.

To understand the effect of the alumina coating layer on the diffusion properties of lithium ions, the galvanostatic intermittent titration technique (GITT) was employed to determine the diffusion coefficient. The GITT was performed to observe the kinetics of lithium ions after formation cycles where the majority of surface area change occurred.24,25,46–48 The procedure is shown in Fig. S12a, consisting of (1) the conditioning step (black), (2) the formation steps (green), and (3) the GITT step (purple), where the capacitance was obtained after finishing the formation steps. The diffusion coefficient from the GITT was calculated from eqn (1):49

 
image file: d5ta03170b-t1.tif(1)
where τ, mB, MB, Vm and S are the time interval during the current pulse, mass, molecular weight, molar volume, and active surface area. The derived equation shown is based on semi-infinite, planar diffusion, steady state voltage change, and small increments of cell voltage during the pulse current. These parameters are shown in Fig. S12b. The diffusion coefficient strongly depends on the active surface area, meaning that changes in surface area must be accounted for to prevent overestimation of the diffusion coefficient.50 To correct this, we merged a capacitance measurement technique that can provide relative electrochemical active surface area compared with the pristine state, allowing corrections to be made for the effects of microcracking.

The diffusion coefficient determined from the GITT is presented in Fig. S13a and b. Overall, the trend of the increasing diffusion coefficient is consistent with the nature of the layered oxide material, which exhibits the highest diffusion rate at around a voltage of 3.85 to 4.15 V due to lattice expansion.48,51–54 The diffusion coefficient for NMC-10ALD is higher than that of NMC, indicating that the 10-cycle alumina coating does not hinder lithium-ion kinetics. This can be attributed to the transformation of the alumina layer into a passivation layer that does not impede lithium-ion diffusion. Although microcracking was suppressed (discussed in the last part), leading to a reduced surface area for lithium to intercalate from the electrolyte into the active material, the higher diffusion coefficient was still observed, even during the early formation cycles.

3.3 Cylindrical jelly-roll configuration for in situ gas detection, transition metal dissolution, and electrolyte decomposition

The jelly roll configuration cell was employed to study the effect of the coating layer at a large-scale level. The experimental setup is shown in Fig. 5a, where the corresponding testing protocols are shown in Fig. 5b and c. High voltage abuse was used because it can severely reflect the degradation of the cells.55 To track the real-time change during the electrochemical abuse, Video S1 shows the cell's behavior during abuse at various voltages, including 4.3, 4.6, and 4.9 V.56 At the abuse voltage of 4.3 V, no obvious change was observed, which might be because 4.3 is not far from normal operating voltage; thus, it is still safe. However, at 4.6 V, obvious gas bubbling was observed at the beginning of the constant voltage step. This can be attributed to the limited operating voltage of NMC811. The operating voltage is too high beyond the redox potential of transition metals, resulting in the participation of the oxygen lattice, finally resulting in oxygen release.6 Under the most abusive condition, 4.9 V, a considerable amount of gas bubbling was observed, suggesting that a high degree of oxygen release was occurring.55 Moreover, after the cell underwent constant voltage discharge, the color of the electrolyte became brownish, which presumably is the color from transition metal dissolution as shown in Fig. S15a–c.18 The color of NMC is browner than that of NMC-10ALD, suggesting a higher degree of dissolution. This hypothesis was confirmed using the ICP-OES of electrolytes in Fig. S15d, which shows that the percentages of TM dissolution were suppressed after ALD coating. However, because of the high content of organic compounds, high dilution is required. As a result, the transition metal dissolution at 4.3 V and 4.6 V was not high enough to be precisely detected, which will be evaluated through transition metal deposition on the anode instead.57 After the color changed shortly, the cells exploded because of the pressure build-up in the cells. In the case of NMC, the cell was in CV mode for 9.27 h before the explosion. However, NMC-10ALD can tolerate the CV mode longer at 12.56 h, which roughly suggests a lower degree of gas evolution. The transition metal dissolution was further evaluated through the amount of TM deposition on the anode by ICP-OES, as shown in Fig. 5d, showing that transition metal content on the anode is lower when the ALD layer is applied on the cathode, suggesting a lower degree of TM dissolution. The TM dissolution severely occurs at a high state of charge. At this state, TMs are in a high oxidation state, which is further reduced by the oxygen lattice that participates in the redox during high voltage, resulting in a lower oxidation state according to Scheme 1 instantaneously with oxygen lattice release.58
image file: d5ta03170b-f5.tif
Fig. 5 Evaluation of transition metal dissolution and degradation behavior in 18[thin space (1/6-em)]650-format jelly-roll cells under high-voltage abuse conditions. (a) Schematic of the jelly-roll configuration used for in situ gas evolution and degradation studies. (b and c) Electrochemical abuse protocols for uncoated NMC and NMC-10ALD samples, respectively: cells were charged and then held at constant voltages of 4.3 V, 4.6 V, and 4.9 V for 20 h. (d) Quantitative analysis of transition metal (Ni, Mn, Co) dissolution measured by ICP-OES from the anode side, showing significantly reduced TM deposition in cells with the ALD-coated cathode.

image file: d5ta03170b-s1.tif
Scheme 1 Transition metal dissolution mechanism with oxygen lattice liberation.

This lower oxidation state of TMs has been reported to have labile stability and higher solubility, thus resulting in dissolution. The TM dissolution process can deteriorate performance in various aspects. Loss of the active material is the first thing to happen; as TMs continue to dissolve, there is a less host structure for lithium. Another problem is that the TMs can crosstalk to the anode and then be deposited on the anode; this process can cause impedance build-up because the TM layer on the anode hinders lithium-ion transport.59,60 When coating with alumina, it is clearly shown that TM dissolution was suppressed, which can be attributed to the protective effect of alumina. The previous study also points out that a few layers of ALD are selectively coated on TMs, which might result in effective TM protection, preventing direct contact between the active material and electrolyte.61 Even though the previous work shows that alumina coating by various techniques can suppress TM dissolution, our work demonstrates that only 2 nm sufficiently protects the cathode.

According to Scheme 1, when the alumina suppresses TM dissolution, the gas evolution named oxygen release from the cathode should also be inhibited. To evaluate this, differential electrochemical mass spectrometry (DEMS) was employed to examine the effect of the coating on gas evolution. The jelly-roll configuration with the graphite anode was used, as shown in Fig. 6a. The experimental setup shown in Fig. 6b consisted of a jellyroll placed in a glass tube filled with electrolyte, connected to a mass spectrometer via a probe. Fig. 6c shows that the experimental protocol used under gas detection involved an initial charge–discharge cycle at C/20 to 4.3 V, followed by consecutive charges to 4.3 V with a constant voltage (CV) held for 6 h at 4.3, 4.4, 4.5, 4.6, and 4.7 V, before discharging to 3.0 V.


image file: d5ta03170b-f6.tif
Fig. 6 Differential electrochemical mass spectroscopy (DEMS): (a) cell configuration, (b) experimental setup, (c) electrochemical performance of NMC and NMC-10ALD under abuse conditions, and gas evolution profiles for (d) CO2, (e) CO, and (f) H2 during the test.

Oxygen evolution was only weakly detected, as shown in Fig. S16. This limited signal can be attributed to the high reactivity of lattice oxygen released from the cathode, which readily reacts with electrolyte solvents before it can escape into the gas phase. Similar observations have been reported in previous studies, where molecular O2 was detectable only when released rapidly and in substantial quantities—conditions under which it could escape prior to undergoing secondary reactions.55 This is consistent with our experimental setup, which employs a large-scale jelly-roll configuration with an excess volume of electrolyte. Such conditions promote rapid consumption of the reactive oxygen species, thereby suppressing detectable O2 evolution and instead leading to predominant formation of CO2 and CO. Consequently, in this study, the extent of oxygen release under high-voltage operation is evaluated indirectly by analyzing the gaseous and liquid-phase decomposition products generated through reactions between singlet oxygen and electrolyte components.

Fig. 6d and e depict the same on-set potential for carbon dioxide (CO2) and carbon monoxide (CO) generation at 4.5 V for NMC and NMC-10ALD. However, the result illustrates that CO2 and CO were significantly inhibited when alumina was coated. These gases are primarily attributed to chemical oxidation processes, as at operating voltages below 5.0 V, chemical oxidation dominates over electrochemical oxidation, rendering the latter negligible.62 The chemical oxidation of electrolyte solvents, including DMC and FEC, with the released oxygen lattice finally yields CO and CO2, as shown in Schemes 2–4 for hydrolysis below.19–21


image file: d5ta03170b-s2.tif
Scheme 2 Decomposition pathways of DMC starting from chemical oxidation.

image file: d5ta03170b-s3.tif
Scheme 3 Decomposition pathways of FEC starting from chemical oxidation.

To further confirm the following mechanisms, the electrolyte was collected, and 1H NMR was used to evaluate the decomposition products in the liquid phase. Fig. S17 shows NMR spectra of fresh electrolyte, NMC, and NMC-10ALD after facing abuse, according to Fig. 5b and c. In the beginning, fresh electrolytes show a trace amount of water at 3.35; however, after facing abuse at 4.3 V, the water disappears, suggesting the hydrolysis of water according to Scheme 4, which results in LiOCH3 formation at 3.33 ppm, which can further participate in the alkalosis of formaldehyde as shown in Scheme 2. At an abuse voltage of 4.6 V, acetals were detected at 5.70, 5.71, and 5.79 ppm. According to Scheme 2, acetals are the decomposition products of formic acid that originate from the chemical oxidation of DMC. Therefore, the suppression of CO2 and CO formation in the alumina-coated samples indicates that the protective alumina layer mitigates oxygen release from the lattice. The oxygen release process is a critical degradation mechanism in lithium-ion batteries that can cause several problems. First, the phase transformation from a layered structure to spinel or rock salt structures can cause impedance to build up in the cells that finally results in a fatigue phase resulting in capacity loss.63,64 Moreover, the surface degradation caused by oxygen release also results in permanent cracking in the active material that also results in capacity fading.65 In addition, according to Scheme 1, transition metal dissolution is also closely related to oxygen loss, meaning that transition metal dissolution is also suppressed.18,47 Lastly, the safety aspect is important since the oxygen release can cause pressure to build up in the cell, which could potentially explode at a critical amount of gas.


image file: d5ta03170b-s4.tif
Scheme 4 Hydrolysis of DMC.

Fig. 6f illustrates the hydrogen (H2) evolution, which occurs at the beginning of the charging process and at 4.4 V. The H2 generation can be attributed to these two main reasons: (1) the reduction of a trace amount of water in the electrolyte and (2) the reduction of protic species that are formed via oxidation at the cathode during high-voltage operation and crosstalk to the anode.26,66 During the early stages of charging, H2 generation is likely due to water reduction at the graphite anode, as there is no obvious increase in the following cycle and absence of water in NMR according to Fig. S17, suggesting that the trace amount of water already transforms into hydrogen and other products. Afterward, at the high operating voltage of 4.4 V, the H2 generation is more likely attributed to the latter pathway. The decomposition products of electrolyte solvents can result in the formation of H2O and other protic species that are formed due to the oxidative decomposition of electrolytes. According to Schemes 2 and 3, DMC and FEC can be oxidized with the liberated singlet oxygen, resulting in protic species and water formation.19–21 These generated products then crosstalk to the anode where reduction occurs, finally resulting in hydrogen evolution. The results clearly show that hydrogen evolution at high voltage is significantly suppressed when NMC is coated with alumina. This suppression suggests that fewer protic species are formed during high-voltage operation, resulting in a lower degree of protic species reduction and, consequently, less hydrogen evolution.

3.4 In situ XRD for lattice parameter change evaluation and microcracking

In situ XRD was employed during the lithiation and delithiation process to evaluate lattice parameter changes. Fig. S18 shows the experimental set-up, including an electrochemical in situ cell in which one site is sealed with a Be window exposed to an X-ray source. The contour plots at the selected sites of (003), (101), and (104) planes are shown in Fig. 7a–c for NMC and Fig. 7e–g for NMC-10ALD, respectively, with response to the electrochemical profile in Fig. 7d for NMC and Fig. 7h for NMC-10ALD. The overall peak shift shows consistent trends with the previous report on layered oxide materials.67 The (003) region first starts to shift toward a lower scattering angle while the (101) and (104) planes shift to a higher angle. Typically, (003) is directly related to the c lattice parameter, while those 2 planes are related to the a lattice parameter.68 Near the end of discharge, (003) rapidly moves to a lower angle, indicating rapid change in c lattice parameters. During discharge, the change was opposite to that in the charging process showing the reversibility of lithiation/delithiation. To elucidate more, Rietveld refinement was performed to demonstrate lattice parameter changes as shown in Fig. 7i–k. First, during charging, the c lattice becomes larger, which can be attributed to the repulsion of the transition metal as the oxidation state becomes higher due to Li deintercalation. Afterwards, when a critical amount of Li was extracted, the c lattice started to shrink. Meanwhile, the a lattice shrinks due to the smaller size of TMs at higher oxidation states. During discharge, the process was the opposite, indicating a reversible process. The percentage changes of c, the a lattice parameter, and the volume of the unit cell can be calculated for NMC which are 3.04, 2.04, and 5.21%, while for NMC-10ALD they were 2.73, 2.00, and 4.73%, respectively. The result shows that lattice parameter changes were milder when the ALD layer was applied, which can be attributed to the effect of the coating layer that can stabilize the structure. A smaller change in lattice parameters is indicative of reduced microcracking, thereby mitigating surface degradation.
image file: d5ta03170b-f7.tif
Fig. 7 In situ XRD during electrochemical testing: the contour plots of XRD patterns in selected regions of (a–c) NMC and (e–g) NMC-10ALD with the corresponding charge–discharge curves of (d) NMC and (h) NMC-10ALD. The refinement results are shown as (i) c lattice parameter, (j) a lattice parameter, and (k) cell volume.

To further investigate the effect of the coating layer on the electrochemical surface area, the capacitance measurement was conducted, which can quantitatively reflect the properties of the whole electrode instead of the representative particles. Fig. S19a shows the protocols used in this work consisting of charge/discharge for 1 h at C/20 followed by a CV step at 2.5 V for 6 h to measure the capacitance of the pristine electrode called the conditioning step (shown as a black line). After completing the subsequent 1 charge/discharge cycle, in both formation (shown as a green line) and cycling (shown as an orange line), the capacitance was obtained by applying PEIS after the CV step at 2.5 V for 6 h to achieve the blocking condition. An equivalent circuit used to fit the EIS spectra shown in Fig. S19b was fitted in the frequency range of 1 Hz – 100 mHz. The example of the fitted Nyquist plot is shown in Fig. S19c showing that the fitting line is consistent with the experimental data in the tail region where the constant phase element of the double layer appears. Under this blocking condition, the charge transfer resistance is in a quasi-infinite state because the cathode is in a fully lithiated state. As a result, the low-frequency response shows purely capacitive behavior, which can be described through a constant phase element of a double layer.69,70

Fig. S19d shows the percentage of increase in capacitance measured after cycling which represents the increasing electrochemical active surface area change. The change in capacitance would purely reflect alterations in the surface area of the active material (NMC) since surface area changes of carbon black and PVDF are expected to be constant.25 The relative percentage increase in capacitance was calculated compared to the pre-formation capacitance (after the conditioning step). The increasing surface area can be attributed to microcracking.24,25 The result shows that the increase in capacitance was lower when the ALD layer was coated, indicating a lower degree of microcracking. The cause of this microcracking reduction would be the alumina layer on the surface of the electrode that acts as an elastic shell holding the primary particles together, preventing the severe separation of the primary particles that cause microcracking. This result first quantitatively demonstrates that the microcracking was suppressed when coated with alumina, consistent with the qualitative aspect.

3.5 Post-mortem analysis

To assess the structural stability of the NMC cathode and the integrity of the alumina coating after prolonged cycling, post-mortem TEM analyses were performed. The 18[thin space (1/6-em)]650 cells were disassembled in an Ar-filled glove box, and the cathodes were collected and transferred using an air-tight, air-sensitive TEM holder to prevent exposure to ambient conditions. As shown in Fig. 8a, HRTEM imaging of the uncoated NMC cathode reveals substantial surface degradation, characterized by a phase transformation from the original layered structure to spinel and rock-salt phases. This transformation is further confirmed by the magnified region in Fig. 8b and its corresponding fast Fourier transform (FFT) pattern in Fig. 8c. In contrast, Fig. 8d presents the HRTEM image of the cycled NMC-10ALD sample, where the layered structure is well preserved. The enlarged view in Fig. 8e and the corresponding FFT in Fig. 8f further confirm the retention of the original layered structure. Notably, a uniform, amorphous surface layer—attributed to the Al2O3 coating—is observed in Fig. 8g, with its amorphous nature validated using the FFT pattern in Fig. 8h. These observations indicate that the alumina coating remains structurally intact and uniformly adhered to the NMC surface even after extended electrochemical cycling, thereby effectively protecting the cathode against parasitic reactions with the electrolyte.
image file: d5ta03170b-f8.tif
Fig. 8 High-resolution TEM images of (a) pristine NMC and (d) NMC-10ALD extracted from electrodes after 1000 cycles of electrochemical testing. (b and e) Enlarged views of the outer surface of the active material for NMC and NMC-10ALD, respectively, with the corresponding FFT patterns shown in (c) and (f). (g) Magnified image of the alumina shell layer on NMC-10ALD, with its corresponding FFT pattern displayed in (h), confirming the amorphous nature of the coating.

To further validate the presence and distribution of the alumina layer, high-angle annular dark-field scanning TEM (HAADF-STEM) and bright-field STEM (BF-STEM) imaging were performed. As shown in Fig. S20a and b, the uncoated NMC sample exhibits no discernible surface layer, whereas the NMC-10ALD sample (Fig. S20c and d) displays a distinct outer region with reduced contrast, corresponding to the lower atomic number of aluminum in the Al2O3 coating.71 STEM-EDS elemental mapping (Fig. S21a–d) further confirms the distribution of aluminum across the NMC particle surface. The Al signal consistently overlaps with the Ni signal, indicating that the Al2O3 coating remains uniformly distributed and firmly attached after long-term operation. Collectively, these results demonstrate that the ALD-derived alumina coating provides robust interfacial protection, preserves the cathode's layered structure, and maintains its physical and chemical stability under extended cycling conditions.

Fig. S22 presents differential scanning calorimetry (DSC) profiles of both coated (NMC-10ALD) and uncoated NMC electrodes, scanned over the temperature range of 100 °C to 350 °C. All samples exhibit clear exothermic peaks, which can be attributed to the exothermic reactions between the electrolyte and oxygen released from the charged Ni-rich cathode. This oxygen evolution is typically associated with surface Ni4+ species and has been widely recognized as a critical factor contributing to thermal runaway in Ni-rich layered oxides.72 Notably, the exothermic peak in the NMC-10ALD sample is shifted to a slightly higher temperature relative to the uncoated NMC, suggesting that the ∼2 nm Al2O3 coating effectively delays the onset of thermal decomposition. This enhancement in thermal stability is likely due to the alumina layer serving as a physical and chemical barrier that suppresses oxygen release and, in turn, retards the oxidative degradation of the electrolyte. Alumina's high thermal stability further reinforces its protective function at elevated temperatures.

To elucidate the chemical evolution of the alumina coating after prolonged cycling, post-mortem X-ray photoelectron spectroscopy (XPS) was conducted, as shown in Fig. S23. The Al 2p peak in the pristine NMC-10ALD electrode appears at 74.60 eV, which is characteristic of Al2O3.32,73 Variability in the Al 2p binding energy across studies has been attributed to differences in ALD film thickness, surface charging, and local bonding environments.

Following electrochemical cycling, the Al 2p spectra reveal the emergence of two additional components: a peak at 75.80 eV, attributed to aluminum oxyfluoride (AlOFx),74 and another at 73.40 eV, corresponding to lithiated alumina (LiAlO2).75,76 The formation of AlOFx is proposed to result from a chemical reaction between the Al2O3 layer and PF6 anions from the LiPF6-based electrolyte, as described by the following pathway (Scheme 5).77


image file: d5ta03170b-s5.tif
Scheme 5 Proposed reaction pathway for the formation of aluminum oxyfluoride (AlOFx) from Al2O3via chemical interaction with PF6 anions in the electrolyte.

This reaction pathway also generates lithium difluorophosphate (LiPO2F2), a byproduct recognized as a beneficial electrolyte additive that enhances the electrochemical performance and interfacial stability of NMC-based lithium-ion batteries.78,79 Furthermore, the formation of aluminum oxyfluorides (AlOFx) on the NMC surface has been reported to facilitate lithium-ion transport across the electrode–electrolyte interface by reducing interfacial resistance.80–82

In parallel, LiAlO2 is formed through the electrochemical lithiation of Al2O3 during battery operation. This transformation has been associated with an increased Li+ diffusion coefficient,38,39 thereby improving both ion insertion kinetics and charge-transfer processes. The presence of LiAlO2 also offers additional surface protection, mitigating electrolyte-induced degradation and structural deterioration of the cathode.75,76

Collectively, these post-cycling chemical transformations demonstrate that the ALD-deposited Al2O3 coating not only remains chemically active but also evolves into functional interfacial species. These species contribute synergistically to enhanced lithium-ion transport and long-term surface stabilization of Ni-rich cathodes, thereby supporting the practical implementation of thin ALD coatings in high-energy-density lithium-ion battery systems.

4. Conclusion

This study underscores the feasibility and substantial efficacy of employing low-temperature ALD to significantly enhance the electrochemical performance and long-term durability of Ni-rich NMC811 cathodes, particularly in industrial-scale, thick-electrode applications. Among various coating techniques, ALD uniquely offers atomic-level precision and uniform conformal coverage, crucial for effectively mitigating surface-driven degradation processes in high-energy lithium-ion batteries. Our comprehensive analysis demonstrates that an optimal ultrathin Al2O3 coating of approximately 2 nm thickness (achieved via 10 ALD cycles) effectively balances protective capability and electrochemical functionality.

This precisely engineered nanoscale coating notably suppresses key degradation mechanisms, including transition metal dissolution, oxygen release, microcracking, and electrolyte decomposition, without adversely affecting lithium-ion transport or rate capability. Electrochemical testing, encompassing prolonged cycling in commercial 18[thin space (1/6-em)]650 cell configurations, confirms substantial improvements in capacity retention, cycle stability, and internal resistance reduction. Detailed structural and chemical characterization through SEM-EDS, GITT, ICP-OES, DEMS, in situ XRD, and post-mortem analysis further validates the protective efficiency of the alumina layer, highlighting its critical role in stabilizing cathode integrity under high-voltage and demanding operational conditions.

The successful implementation of precise nanoscale coatings on thick commercial-scale electrodes via ALD addresses one of the key technical barriers currently limiting the widespread commercialization of Ni-rich cathodes. Collectively, these findings significantly advance our understanding of robust cathode design and present a viable pathway toward industrial-scale integration of ALD technology. Ultimately, this research emphasizes the transformative potential of controlled surface engineering, setting a robust foundation for the future development of safer, more reliable, and commercially viable high-performance lithium-ion batteries.

Data availability

The data that support the findings of this study are available from the corresponding author, Dr Montree Sawangphruk, upon reasonable request. The raw and processed data, including ALD coating parameters, electrochemical testing results, and characterization data (TEM, XPS, GITT, and gas evolution analysis), have been included in the ESI file.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

This work was financially supported by the Program Management Unit for National Competitiveness Enhancement (PMU-C) by the Office of National Higher Education Science Research and Innovation Policy Council (NXPO) and IRPC Public Company Limited, Thailand Science Research and Innovation (TSRI) under the Fundamental Fund by TSRI (FRB680014/0457), Vidyasirimedhi Institute of Science and Technology (VISTEC), and the Energy Policy and Planning Office (EPPO), Ministry of Energy, Thailand. In addition, the Frontier Research Centre (FRC) supported this work through VISTEC.

References

  1. T. Liu, L. Yu, J. Liu, J. Lu, X. Bi, A. Dai, M. Li, M. Li, Z. Hu, L. Ma, D. Luo, J. Zheng, T. Wu, Y. Ren, J. Wen, F. Pan and K. Amine, Nat. Energy, 2021, 6, 277–286 CrossRef CAS.
  2. H. Li, M. Cormier, N. Zhang, J. Inglis, J. Li and J. R. Dahn, J. Electrochem. Soc., 2019, 166, A429 CrossRef CAS.
  3. H.-J. Noh, S. Youn, C. S. Yoon and Y.-K. Sun, J. Power Sources, 2013, 233, 121–130 CrossRef CAS.
  4. G. W. Nam, N.-Y. Park, K.-J. Park, J. Yang, J. Liu, C. S. Yoon and Y.-K. Sun, ACS Energy Lett., 2019, 4, 2995–3001 CrossRef CAS.
  5. S. Yin, W. Deng, J. Chen, X. Gao, G. Zou, H. Hou and X. Ji, Nano Energy, 2021, 83, 105854 CrossRef CAS.
  6. A. Manthiram, Nat. Commun., 2020, 11, 1550 CrossRef CAS PubMed.
  7. M. J. Herzog, D. Esken and J. Janek, Batteries Supercaps, 2021, 4, 1003–1017 CrossRef CAS.
  8. C. Geng, A. Liu and J. R. Dahn, Chem. Mater., 2020, 32, 6097–6104 CrossRef CAS.
  9. K. Srimanon, S. Vadivel and M. Sawangphruk, J. Power Sources, 2022, 550, 232150 CrossRef CAS.
  10. S. Neudeck, F. Strauss, G. Garcia, H. Wolf, J. Janek, P. Hartmann and T. Brezesinski, Chem. Commun., 2019, 55, 2174–2177 RSC.
  11. J.-M. Kim, X. Zhang, J.-G. Zhang, A. Manthiram, Y. S. Meng and W. Xu, Mater. Today, 2021, 46, 155–182 CrossRef CAS.
  12. M. J. Herzog, N. Gauquelin, D. Esken, J. Verbeeck and J. Janek, Energy Technol., 2021, 9, 2100028 CrossRef CAS.
  13. J. Duan, X. Tang, H. Dai, Y. Yang, W. Wu, X. Wei and Y. Huang, Electrochem. Energy Rev., 2020, 3, 1–42 CrossRef CAS.
  14. B. Han, B. Key, A. S. Lipton, J. T. Vaughey, B. Hughes, J. Trevey and F. Dogan, J. Electrochem. Soc., 2019, 166, A3679 CrossRef CAS.
  15. Y. S. Jung, A. S. Cavanagh, L. A. Riley, S.-H. Kang, A. C. Dillon, M. D. Groner, S. M. George and S.-H. Lee, Adv. Mater., 2010, 22, 2172–2176 CrossRef CAS PubMed.
  16. D. Mohanty, K. Dahlberg, D. M. King, L. A. David, A. S. Sefat, D. L. Wood, C. Daniel, S. Dhar, V. Mahajan, M. Lee and F. Albano, Sci. Rep., 2016, 6, 26532 CrossRef CAS PubMed.
  17. N. L. Chang, G. K. Poduval, B. Sang, K. Khoo, M. Woodhouse, F. Qi, M. Dehghanimadvar, W. M. Li, R. J. Egan and B. Hoex, Prog. Photovolt., 2023, 31, 414–428 CrossRef CAS.
  18. R. Sahore, D. C. O'Hanlon, A. Tornheim, C.-W. Lee, J. C. Garcia, H. Iddir, M. Balasubramanian and I. Bloom, J. Electrochem. Soc., 2020, 167, 020513 CrossRef CAS.
  19. B. L. D. Rinkel, J. P. Vivek, N. Garcia-Araez and C. P. Grey, Energy Environ. Sci., 2022, 15, 3416–3438 RSC.
  20. B. L. D. Rinkel, D. S. Hall, I. Temprano and C. P. Grey, J. Am. Chem. Soc., 2020, 142, 15058–15074 CrossRef CAS PubMed.
  21. K. Homlamai, T. Sangsanit, R. Songthan, W. Tejangkura and M. Sawangphruk, ChemSusChem, 2025, 18(11), e202500238 CrossRef CAS PubMed.
  22. N. Phattharasupakun, P. Bunyanidhi, P. Chiochan, N. Chanlek and M. Sawangphruk, Electrochem. Commun., 2022, 139, 107309 CrossRef CAS.
  23. Y. Seok Jung, A. S. Cavanagh, Y. Yan, S. M. George and A. Manthiram, J. Electrochem. Soc., 2011, 158, A1298 CrossRef.
  24. S. Oswald, F. Riewald and H. A. Gasteiger, J. Electrochem. Soc., 2022, 169, 040552 CrossRef CAS.
  25. S. Oswald, D. Pritzl, M. Wetjen and H. A. Gasteiger, J. Electrochem. Soc., 2020, 167, 100511 CrossRef CAS.
  26. W. M. Dose, I. Temprano, J. P. Allen, E. Björklund, C. A. O'Keefe, W. Li, B. L. Mehdi, R. S. Weatherup, M. F. L. De Volder and C. P. Grey, ACS Appl. Mater. Interfaces, 2022, 14, 13206–13222 CrossRef CAS PubMed.
  27. S. Albrecht, J. Kümpers, M. Kruft, S. Malcus, C. Vogler, M. Wahl and M. Wohlfahrt-Mehrens, J. Power Sources, 2003, 119–121, 178–183 CrossRef CAS.
  28. I. Belharouak, W. Lu, D. Vissers and K. Amine, Electrochem. Commun., 2006, 8, 329–335 CrossRef CAS.
  29. S. Ko, Y. Yamada, L. Lander and A. Yamada, Carbon, 2020, 158, 766–771 CrossRef CAS.
  30. K. G. Araño, B. L. Armstrong, G. Yang, C. Kumara, T. Z. Ward, H. M. Meyer, III, A. M. Rogers, E. Toups and G. M. Veith, Energy Fuels, 2024, 38, 6446–6458 CrossRef.
  31. J. Alvarado, C. Ma, S. Wang, K. Nguyen, M. Kodur and Y. S. Meng, ACS Appl. Mater. Interfaces, 2017, 9, 26518–26530 CrossRef CAS PubMed.
  32. Y. Shi, M. Zhang, D. Qian and Y. S. Meng, Electrochim. Acta, 2016, 203, 154–161 CrossRef CAS.
  33. L. Wang, Q. Su, W. Shi, C. Wang, H. Li, Y. Wang, G. Du, M. Zhang, W. Zhao, S. Ding and B. Xu, Electrochim. Acta, 2022, 435, 141411 CrossRef CAS.
  34. S. Cava, S. M. Tebcherani, I. A. Souza, S. A. Pianaro, C. A. Paskocimas, E. Longo and J. A. Varela, Mater. Chem. Phys., 2007, 103, 394–399 CrossRef CAS.
  35. S. Kim, S.-H. Lee, I. H. Jo, J. Seo, Y.-E. Yoo and J. H. Kim, Sci. Rep., 2022, 12, 5124 CrossRef CAS.
  36. H. S. Jin, D. H. Kim, S. K. Kim, R. M. Wallace, J. Kim and T. J. Park, Adv. Electron. Mater., 2019, 5, 1800680 CrossRef.
  37. L. A. Riley, S. Van Atta, A. S. Cavanagh, Y. Yan, S. M. George, P. Liu, A. C. Dillon and S.-H. Lee, J. Power Sources, 2011, 196, 3317–3324 CrossRef CAS.
  38. A. M. Nolan, D. Wickramaratne, N. Bernstein, Y. Mo and M. D. Johannes, Chem. Mater., 2021, 33, 7795–7804 CrossRef CAS.
  39. S. C. Jung and Y.-K. Han, J. Phys. Chem. Lett., 2013, 4, 2681–2685 CrossRef CAS.
  40. S.-Y. Kim and Y. Qi, J. Electrochem. Soc., 2014, 161, F3137 CrossRef CAS.
  41. A. M. Wise, C. Ban, J. N. Weker, S. Misra, A. S. Cavanagh, Z. Wu, Z. Li, M. S. Whittingham, K. Xu, S. M. George and M. F. Toney, Chem. Mater., 2015, 27, 6146–6154 CrossRef CAS.
  42. F. Xin, A. Goel, X. Chen, H. Zhou, J. Bai, S. Liu, F. Wang, G. Zhou and M. S. Whittingham, Chem. Mater., 2022, 34, 7858–7866 CrossRef CAS.
  43. F. Wang, S. Tang, X. Han, Y. Wu, L. Lu, C. Yu, X. Sun and M. Ouyang, Chem. Eng. J., 2024, 500, 157026 CrossRef CAS.
  44. J. B. Adamo and A. Manthiram, ACS Appl. Energy Mater., 2025, 8, 2200–2208 CrossRef CAS.
  45. K. Homlamai, N. Anansuksawat, T. Sangsanit, S. Prempluem, K. Santisuk, W. Tejangkura and M. Sawangphruk, J. Power Sources, 2024, 617, 235150 CrossRef CAS.
  46. F. Riewald, P. Kurzhals, M. Bianchini, H. Sommer, J. Janek and H. A. Gasteiger, J. Electrochem. Soc., 2022, 169, 020529 CrossRef CAS.
  47. S. Oswald, D. Pritzl, M. Wetjen and H. A. Gasteiger, J. Electrochem. Soc., 2021, 168, 120501 CrossRef CAS.
  48. E. Trevisanello, R. Ruess, G. Conforto, F. H. Richter and J. Janek, Adv. Energy Mater., 2021, 11, 2003400 CrossRef CAS.
  49. W. Weppner and R. A. Huggins, J. Electrochem. Soc., 1977, 124, 1569 CrossRef CAS.
  50. S. D. Kang and W. C. Chueh, J. Electrochem. Soc., 2021, 168, 120504 CrossRef.
  51. N. Phattharasupakun, M. M. E. Cormier, E. Lyle, E. Zsoldos, A. Liu, C. Geng, Y. Liu, H. Li, M. Sawangphruk and J. R. Dahn, J. Electrochem. Soc., 2021, 168, 090535 CrossRef CAS.
  52. K. Kang and G. Ceder, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 74, 094105 CrossRef.
  53. A. Van der Ven, J. C. Thomas, Q. Xu, B. Swoboda and D. Morgan, Phys. Rev. B: Condens. Matter Mater. Phys., 2008, 78, 104306 CrossRef.
  54. A. Van der Ven and G. Ceder, J. Power Sources, 2001, 97–98, 529–531 CrossRef CAS.
  55. R. Jung, M. Metzger, F. Maglia, C. Stinner and H. A. Gasteiger, J. Electrochem. Soc., 2017, 164, A1361 CrossRef CAS.
  56. S. Prempluem, T. Sangsanit, K. Santiyuk, K. Homlamai, W. Tejangkura, R. Songthan, N. Anansuksawat and M. Sawangphruk, J. Power Sources, 2024, 606, 234538 CrossRef CAS.
  57. R. Jung, F. Linsenmann, R. Thomas, J. Wandt, S. Solchenbach, F. Maglia, C. Stinner, M. Tromp and H. A. Gasteiger, J. Electrochem. Soc., 2019, 166, A378 CrossRef CAS.
  58. Z. Ruff, C. Xu and C. P. Grey, J. Electrochem. Soc., 2021, 168, 060518 CrossRef CAS.
  59. M.-T. F. Rodrigues, K. Kalaga, S. E. Trask, I. A. Shkrob and D. P. Abraham, J. Electrochem. Soc., 2018, 165, A1697 CrossRef CAS.
  60. J. A. Gilbert, I. A. Shkrob and D. P. Abraham, J. Electrochem. Soc., 2017, 164, A389 CrossRef CAS.
  61. A. L. Hoskins, W. W. McNeary, S. L. Millican, T. A. Gossett, A. Lai, Y. Gao, X. Liang, C. B. Musgrave and A. W. Weimer, ACS Appl. Nano Mater., 2019, 2, 6989–6997 CrossRef CAS.
  62. R. Jung, M. Metzger, F. Maglia, C. Stinner and H. A. Gasteiger, J. Phys. Chem. Lett., 2017, 8, 4820–4825 CrossRef CAS PubMed.
  63. C. Xu, K. Märker, J. Lee, A. Mahadevegowda, P. J. Reeves, S. J. Day, M. F. Groh, S. P. Emge, C. Ducati, B. Layla Mehdi, C. C. Tang and C. P. Grey, Nat. Mater., 2021, 20, 84–92 CrossRef CAS PubMed.
  64. S. S. Zhang, Energy Storage Mater., 2020, 24, 247–254 CrossRef.
  65. S. Lee, L. Su, A. Mesnier, Z. Cui and A. Manthiram, Joule, 2023, 7, 2430–2444 CrossRef CAS.
  66. M. Metzger, B. Strehle, S. Solchenbach and H. A. Gasteiger, J. Electrochem. Soc., 2016, 163, A798 CrossRef CAS.
  67. P. Bunyanidhi, N. Phattharasupakun, C. Tomon, S. Duangdangchote, P. Kidkhunthod and M. Sawangphruk, J. Power Sources, 2022, 549, 232043 CrossRef CAS.
  68. W.-S. Yoon, K. Y. Chung, J. McBreen and X.-Q. Yang, Electrochem. Commun., 2006, 8, 1257–1262 CrossRef CAS.
  69. N. Ogihara, Y. Itou, T. Sasaki and Y. Takeuchi, J. Phys. Chem. C, 2015, 119, 4612–4619 CrossRef CAS.
  70. N. Ogihara, S. Kawauchi, C. Okuda, Y. Itou, Y. Takeuchi and Y. Ukyo, J. Electrochem. Soc., 2012, 159, A1034 CrossRef CAS.
  71. H. Huang, L. Qiao, H. Zhou, Y. Tang, M. J. Wahila, H. Liu, P. Liu, G. Zhou, M. Smeu and H. Liu, Sci. Rep., 2024, 14, 18180 CrossRef CAS PubMed.
  72. Y.-K. Sun, S.-T. Myung, B.-C. Park, J. Prakash, I. Belharouak and K. Amine, Nat. Mater., 2009, 8, 320–324 CrossRef CAS PubMed.
  73. H.-M. Cheng, F.-M. Wang, J. P. Chu, R. Santhanam, J. Rick and S.-C. Lo, J. Phys. Chem. C, 2012, 116, 7629–7637 CrossRef CAS.
  74. R. Ramos, G. Cunge, B. Pelissier and O. Joubert, Plasma Sources Sci. Technol., 2007, 16, 711 CrossRef CAS.
  75. F. Wu, Q. Shi, L. Chen, J. Dong, J. Zhao, H. Wang, F. Gao, J. Liu, H. Zhang, N. Li, Y. Lu and Y. Su, Chem. Eng. J., 2023, 470, 144045 CrossRef CAS.
  76. X. Xiao, P. Lu and D. Ahn, Adv. Mater., 2011, 23, 3911–3915 CrossRef CAS PubMed.
  77. D. S. Hall, R. Gauthier, A. Eldesoky, V. S. Murray and J. R. Dahn, ACS Appl. Mater. Interfaces, 2019, 11, 14095–14100 CrossRef CAS PubMed.
  78. Q. Q. Liu, L. Ma, C. Y. Du and J. R. Dahn, Electrochim. Acta, 2018, 263, 237–248 CrossRef CAS.
  79. W. Zhao, G. Zheng, M. Lin, W. Zhao, D. Li, X. Guan, Y. Ji, G. F. Ortiz and Y. Yang, J. Power Sources, 2018, 380, 149–157 CrossRef CAS.
  80. S. U. Woo, C. S. Yoon, K. Amine, I. Belharouak and Y. K. Sun, J. Electrochem. Soc., 2007, 154, A1005 CrossRef CAS.
  81. Y. K. Sun, S. W. Cho, S. W. Lee, C. S. Yoon and K. Amine, J. Electrochem. Soc., 2007, 154, A168 CrossRef CAS.
  82. K. J. Rosina, M. Jiang, D. Zeng, E. Salager, A. S. Best and C. P. Grey, J. Mater. Chem., 2012, 22, 20602–20610 RSC.

Footnote

Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta03170b

This journal is © The Royal Society of Chemistry 2025
Click here to see how this site uses Cookies. View our privacy policy here.