1 g NMC811 particles (Xiamen TOB New Energy Technology) were dispersed in ethanol (2.5 mL) using ultrasonication for 30 min. Then, 0.12 mmol of Zn(CH₃COO)₂ (Sigma-Aldrich) and 0.14 mmol of LiOH·H₂O (Sigma-Aldrich) precursor materials were added; and then, an additional ethanol (2.5 mL) was added, followed by dispersing the solution with ultrasonication for 30 min. Afterwards, an additional ethanol (45 mL) was added, accompanied by homogeneous mixing with magnetic stirring on a hot plate at 25 °C for 3 h. Subsequently, the ethanol was evaporated during stirring at 50°C for 5 h to produce intermediate product particles. Finally, annealing was conducted at 500 °C for 6 h with a heating rate of 5°C/min to yield LZO on NMC811 cathode materials. To evaluate the influence of Li doping in ZnO on LIB performance, pure ZnO was synthesized using the procedure mentioned above, with the exception of excluding the LiOH·H₂O precursor material. Supplementary Figure S1 demonstrates a schematic diagram that illustrates the synthetic procedures of Li-doped ZnO on NMC811 using the sol-gel process.

Phases of the as-synthesized LZO on NMC811 cathode materials were identified using powder X-ray diffraction (XRD) analysis using a X-ray diffractometer (XRD, D2 Phaser, Bruker). The surface morphology of the as-synthesized and cycled LZO on NMC 811 cathode materials were analyzed using a field emission scanning electron microscope (FESEM, SmartSEM Sigma, Carl Zeiss). The surface morphology and thickness of the 2 wt% LZO thin layer coated on NMC811, which exhibited the highest LIB performance, were characterized using a 200 kV field emission transmission electron microscope (FE-TEM, JEM-2100F(HR), JEOL). The chemical composition and element distribution of the LZO on NMC811 were examined using energy dispersive X-ray spectroscopy (EDS) with a Bruker Quantax microanalysis system equipped with an XFlash^® 6–30 detector. To identify elemental composition and chemical state of materials, X-ray photoelectron spectroscopy (XPS) analysis was conducted (Thermo Scientific Sigma). All spectra were calibrated with the C 1s peak assigned to 284.6 eV. Particularly, peak deconvolution procedures of Ni 2p XPS spectrum and peak-preserving smoothing of Zn 2p XPS spectrum were carried out using Casa XPS software.

To evaluate the LIB performance of the LZO on NMC811, a slurry was prepared by mixing active material constituted of the LZO on NMC811 (70 wt%) (the NMC811 serves as the active material, while the LZO thin layer plays a role in improving cyclability), conductive additive (20 wt%, Super-P), and binder (10 wt%, poly (vinylidene fluoride) (PVDF)). The mixture was ground for 30 min in a mortar with a pestle; then, N-methyl-2-pyrrolidone (NMP) solution was added. The slurry was cast onto an carbon coated Al foil current collector (MTI Korea) with a doctor blade. Subsequently, the slurry on the current collector was dried at 80°C for 12 h. The disk-type of working electrode was fabricated by a hole punch with 16 mm diameter. We included digital images of the disk-shaped cathode before and after ZnO coating on NMC811 in the Supplementary Figure S2. A CR2032 coin cell was assembled in an Ar-filled glovebox under the extremely low levels of humidity and oxygen of less than 0.1 ppm. As cell components, a Li metal was used as a counter and reference electrode; a polypropylene-polyethylene-polypropylene (PP-PE-PP) trilayer was served as a separator. A 1M LiPF₆ in a mixture solvent of ethylene carbonate (EC), ethyl methyl carbonate (EMC), and dimethyl carbonate (DMC) in a volume ratio of 1:1:1 (EC:DMC:EMC) was utilized as an electrolyte. The LIB performance of the LZO on NMC811-based cathode was tested in the coin-cell by using a multi-channel battery testing system (Neware) at 25°C under Galvanostatic mode (constant current density). We also evaluated the cycling performance of an LIB cell with the 2 wt% LZO on NMC811 sample at higher temperature (55°C) for practical application in real LIBs. Additionally, we investigated the thermal stability of LZO coating layer under the elevated temperature. The cell was cycled in the voltage window from 2.7 to 4.4V. After completing 300 cycles, electrochemical impedance spectroscopy (EIS) measurement was conducted on the tested cells using an impedance analyzer (ZIVE MP2; Won-A Tech) at 25 °C. During the EIS measurement, a sinusoidal AC voltage was applied in the frequency range from 1 MHz to 10 mHz with an amplitude of 5 mV. The measurement of transition metal concentrations within the electrolyte employed in the coin cell used for LIB testing was conducted using an inductively coupled plasma mass spectrometer (ICP-MS, Agilent 7900 ICP-MS).

Figure 1 presents the surface morphology of bare NMC811 particles and NMC811 particles coated with Li-doped ZnO (LZO) at ratios of 1, 2, and 3 wt% using the sol-gel synthetic method. Figures 1A, E depict images captured at the lower and higher magnifications, respectively. Upon coating, insignificant changes in particle shape and size are observed, indicating that LZO is uniformly and thinly coated on the surface of NMC811 without altering its surface characteristics. The spherical particles have an approximate size of 10 μm, comprising agglomerated nano-sized primary particles with a length of approximately 500 nm (Xu et al., 2021). This meatball-like structure provides a higher tap density, crucial for achieving superior electrochemical performance compared to other dispersed particle structures. Furthermore, we measured a thickness of 3–5 nm for a 2 wt% LZO thin layer coated on NMC811 using TEM analytical technique for both bare NMC811 and the 2 wt% LZO-coated NMC811 particulate samples (see Supplementary Figure S3).

Successful coating of ZnO on NMC811 is confirmed by EDS color mapping. In Figure 2, the color mapping and spectra obtained through energy dispersive spectroscopy (EDS) display the distribution of various elements (Zn, Ni, Co, O, and Mn) present on the surface of NMC811 particles coated with LZO. The EDS analysis reveals that all the elements in LZO on NMC811 are uniformly distributed across all the particles. Noticeably, Figure 2B confirms the uniform coating of Zn element on the surface of NMC811 particles. These results clearly demonstrate the presence of uniformly coated Li-doped ZnO on the surface of NMC811 particles. According to the EDS spectra (Figure 2F), Zn exhibits a composition of approximately 0.15 at%, while the atomic ratio of Zn was recorded as a value of zero for the bare NMC811. Through EDS data, we can confirm that LZO synthesized via a mass-producible and highly-efficient sol-gel synthetic method could be uniformly and effectively coated on commercially available NMC811 cathode particles.

As shown in Figure 3, the diffraction peaks can be correlated to the hexagonal a-NaFeO₂ structure with the R-3m space group in the absence of any additional diffraction peaks associated with secondary phases or impurities. The separation of peak pairs (006)/(102) and (108)/(110) in all the samples indicates the formation of a layered structure (Yang et al., 2012). Furthermore, the LZO coating material has only a miniscule impact on the NMC811 cathode structure. Considering samples under all conditions before conducting battery cycling test, we can observe peaks corresponding to the crystalline planes of NMC811 including (003) at 18.8°, (101) at 36.7° and (104) at 44.5° (Figure 3B). In contrast, among the samples subjected to the LIB test up to 300 cycles (Figure 3C), we can only identify peaks related to the crystalline planes of NMC811 from the samples coated with 2 and 3 wt% LZO. This suggests that coating materials with a thickness of at least 2 wt% could effectively maintain the structural integrity of the crystalline planes of the cycled NMC811 samples. Nevertheless, we could find no peak regarding Zn_1-xLi_xO (0.00 ≤ x ≤ 0.50) phases with a wurtzite hexagonal structure (JCPDS card (ID-36–1,451)) (Lin et al., 2010; Ardyanian and Sedigh, 2014; Rakkesh and Balakumar, 2014). This finding demonstrates there is a limitation in obtaining XRD patterns for LZO, which is attributed to low-energy X-ray source in a laboratory-scale XRD equipment, particularly for detecting thin coating layers with sub-nanometer grain size.

Figure 4 shows the results of XPS surface analysis for the electrodes fabricated with bare NMC811 and 2 wt% LZO on NMC811 particles, which exhibits the best battery performance. For Ni in both samples (Figures 4A, B), we can acquire peaks corresponding to Ni 2p_3/2 and Ni 2p_1/2 binding energies at 855.8 and 872.7 eV, respectively (Yu et al., 2022). For the 2wt% LZO on NMC811 sample (Figure 4A), the Ni 2p_3/2 binding energy peak can be deconvoluted into two components located at 855.5 and 858 eV, respectively, which are consistent with the binding energies of Ni³⁺ and Ni²⁺, indicating a mixed state of Ni²⁺ and Ni³⁺ oxidation states near the surface of particles. Similarly, the Ni 2p_1/2 binding energy peak can be deconvoluted into two components located at Ni²⁺ (873.2 eV) and Ni³⁺ (875.6 eV) (Kong et al., 2014). In contrast, for Mn with a lower composition (Figure 4C), there is insignificant peak corresponding to Mn 2p_3/2 and Mn 2p_1/2 binding energies at 642.5 and 654 eV, respectively (Pereira et al., 2020). Similarly, for Co with a lower composition (Figure 4D), we can attain a small broad peak corresponding to Co 2p_3/2 binding energy at 781 eV (Fantauzzi et al., 2019). Our main interest lies in determining whether LZO is uniformly coated on the surface of NMC811 particles through XPS analysis. Figure 4E shows the observation of peaks corresponding to Zn (1,022.7 eV for Zn 2p_3/2 and 1,045.9 eV for Zn 2p_1/2) for the 2 wt% LZO on NMC811, thus manifesting the presence of Zn-O chemical bonding in ZnO. Based on the spectrum, Zn shows a composition of 30.81 at%. These findings can justify the formation of LZO coating materials on the surface of NMC811 (Kong et al., 2014). In contrast, we could find insignificant Zn element detected for the bare NMC811 cathode material as illustrated in Figure 4E.

Figure 5A presents the voltage profiles of the 2 wt% LZO on NMC811/Li half-cell during the initial two cycles between 4.4V and 2.7V at 25°C. The half-cell exhibits charge capacities of 239.3 and 208.1 mAh/g and discharge capacities of 205.5 and 204.8 mAh/g in the first and second cycle, respectively. Coulombic efficiencies (CE) of the first and second cycles are recorded as 85.9% and 99.7%, respectively. To compare, CE of the bare NMC811 for the first and second cycles are measured as 81.2% and 97.5%, respectively (see Supplementary Figure S4). The superior CE values of 2 wt% LZO on NMC811 is attributed by a process of formation and stabilization of the solid electrolyte interface (SEI) protective layer (Nisar et al., 2021). Figure 5B displays the cycle performance results obtained from the cell measurements of the bare NMC811 and the LZO coated on NMC811 samples, with LZO content controlled at 1, 2, and 3 wt% with respect to bare NMC811. Measurements were conducted four times for each sample at 25°C; then, the average value of each one was calculated and presented in Figure 5B. Noticeably, all samples showed a decrease in specific capacity with increasing cycle number. When cycled up to 300 cycles, the bare NMC811 cathode material delivers a capacity of approximately 90 mAh/g with a capacity retention of 59%. Among the LZO on NMC811 samples, the 2 wt% sample demonstrated the highest capacity (approximately 112 mAh/g) and a capacity retention of approximately 71%, superior to the results of other oxide-coated NMC811 (Himani et al., 2021; Jia et al., 2022). In comparison, the 1 and 3 wt% samples show capacities of 107 mAh/g and 95 mAh/g, respectively, with capacity retentions of nearly 71% and 63%, respectively. Based on these results, we could confirm that LZO coating materials guard the active material against HF attack, mitigate a problem with the leaching of transition metal ions in the cathode during cycling, and maintain a stable cathode electrolyte interphase (CEI) (Nisar et al., 2021). We also evaluated the cycling performance of an LIB cell with a 2 wt% LZO on NMC811 sample at a higher temperature (55°C) for practical applications in real LIBs. Supplementary Figure S5 demonstrates the superior LIB performance observed at higher temperature compared to lower temperature (25°C). Based on this result, we can confirm the excellent thermal stability of the LZO coating layer under elevated temperatures.

As mentioned in the introduction section, we doped Li into the ZnO structure to enhance the low Li⁺ ion conductivity of the pure ZnO coating layer. To ascertain the successful incorporation of Li⁺ ions into the ZnO structure, the conventional approach in previous reports have involved analyzing the presence of Li within Li-doped ZnO using XPS (Xiaofeng et al., 2019; Jie et al., 2022). However, we faced a limitation to apply this approach to our proposed Li-doped ZnO coating layer in this study. Both NMC811 and 3–5 nm thick LZO thin film coated on NMC811 samples contain Li; therefore, data obtained from analytical techniques can indicate the presence of Li in both NMC811 and LZO on NMC811. As an alternative approach to analytical techniques, we compared the electrochemical performance between 2 wt% ZnO on NMC811 and 2 wt% Li-doped ZnO on NMC811 at 1C for 10 cycles to imply the presence of Li within the ZnO structure in the LZO sample (Supplementary Figure S6). The specific capacity of the 2 wt% LZO on NMC811 surpasses that of 2 wt% ZnO on NMC811 for all the cycles. These data demonstrate that Li should be doped into ZnO and act as a facilitator for Li⁺ ion and electron transport, leading to superior LIB performance compared to pure ZnO.

Furthermore, through 300 charge-discharge cycling test, a reduction in capacity retention is observed in all samples as the cycles increase. This phenomenon can be attributed to the dissolution of transition metals within NMC811 and the structural degradation of NMC811 during cycling processes (Nisar et al., 2021). Noticeably, our proposed LZO coating layer serves as a protective barrier to mitigate these issues. To verify this hypothesis, post-mortem analyses (SEM and ICP-MS) were performed on both bare NMC811 and 2 wt% LZO on NMC811 after completing the charge-discharge testing. As illustrated in Supplementary Figure S7, we observe distinct morphological differences between bare NMC811 and 2 wt% LZO on NMC811 in the SEM images. This observation confirms that the structural integrity of 2 wt% LZO on NMC811 is more preserved compared to that of bare NMC811. Furthermore, ICP-MS analysis was employed to determine the concentrations of Ni, Co, and Mn, the transition metal constituents of NMC811. For bare NMC811, the concentrations of Ni, Co, and Mn are 460, 45, and 68 ppb, respectively. In contrast, for 2 wt% LZO on NMC811, the corresponding concentrations are 370, 46, and 70 ppb, respectively. This higher Ni concentration in bare NMC811 indicates a greater amount of Ni dissolution from bare NMC811 compared to 2 wt% LZO on NMC811. These findings demonstrate the significant role of the LZO coating layer in mitigating the structural degradation of NMC811 and suppressing the leaching of transition metals into the electrolyte during cycling, in line with the LIB performance exhibited in Figure 5B.

To further evaluate the effect of LZO coating materials on NMC811 cathodes, we conducted electrochemical impedance spectroscopy (EIS) analysis. Figure 5C presents the Nyquist plots obtained for each sample after 300 cycle test. A slight interruption at high frequency corresponds to the electrolyte resistance (R_s), which lies within the range of 3–4 Ω for all samples. As the frequency changes from high to medium regions, the charge transfer resistance (R_ct) can be calculated for each sample based on the semicircle. The values of R_ct for each sample follow the sequence: 16.7 Ω (2 wt% LZO), 20 Ω (1 wt% LZO), 20.2 Ω (3 wt% LZO), and 50.7 Ω (bare NMC811). It should be noted that the smaller values of R_ct for all the LZO on NMC811 samples than the bare NMC811 sample can be mainly associated with the significant roles of LZO on NMC811 as mentioned earlier (Song et al., 2011; Shi et al., 2013; Kong et al., 2014). These results are consistent with the aforementioned LIB performance during charge and discharge cycles (see Figure 5B). Moreover, the calculated Li⁺ diffusion coefficient for the bare NCM811 electrode subjected to 300 cycling test is 2.34 × 10⁻¹⁵ cm²/s, while those for the NCM811 samples with 1 wt%, 2 wt%, and 3 wt% LZO coatings are 1.31 × 10⁻¹⁴ cm²/s, 1.99 × 10⁻¹⁴ cm²/s, and 4.66 × 10⁻¹⁵ cm²/s, respectively (Feng et al., 2022). These results confirm that the LZO coating layer can enhance Li⁺ ion kinetics, consistent with the LIB performance at 1 C for 300 cycles (Figure 5B) and previous reports (Tongxin et al., 2022).