The burgeoning electric vehicle market has catalyzed the advancement of high-performance cathode materials. The nickel-rich layered cathode LiNi_xCo_yMn_1-x-yO₂ (NCM) is distinguished by its high energy density, exemplary performance, and cost-effectiveness, presenting considerable benefits in electric vehicle (EV) applications [1–4]. However, nickel-rich materials are inherently prone to substantial surface residual lithium (LiOH, LiHCO₃, and Li₂CO₃) due to lower calcination temperatures and sensitivity to atmospheric CO₂ and H₂O [5–7]. This residual lithium leads to lattice oxygen and active lithium depletion, expediting the emergence of rock-salt phases at the surface/subsurface of particle [8,9]. Furthermore, the highly oxidative Ni⁴⁺ tends to react severely with the electrolyte, forming a loose and fragile cathode electrolyte interphase (CEI) layer [10]. Consequently, this diminishes the electronic and ionic conductivities of the electrode materials, escalating impedance and impairing rate capabilities [11,12]. Alarmingly, at a charging voltage of 4.2 V, Li₂CO₃ decomposition releases gas, posing severe battery safety risks [13–15].

Addressing the issues inherent in nickel-rich materials, researchers have found success with surface cleansing and coating approaches [16–19]. Washing with water and ethanol solutions effectively removes surface residual alkali from nickel-rich materials. Economical and straightforward water-washing techniques have already been integrated into production processes [20]. Applying a surface coating creates a protective barrier, reducing adverse side reactions. Solid-state methods, wet chemical methods, and atomic layer deposition are commonly used to apply coating layers on cathode material particles. Simple solid-state methods can coat nickel-rich layered cathode materials with layers such as Al₂O₃ [21], LiAlO₂ [22], ZrO₂ [23], B₂O₃ [24], isolating them from environmental moisture and CO₂, preserving stability during storage and greatly mitigating negative interface reactions, thereby improving cycle stability [25–30]. Compared to crystalline coatings, the amorphous layer characterized by short-range order and long-range disorder enhances Li⁺ transport. Moreover, its intimate adhesion to the cathode material surface helps mitigate the formation of interfacial microcracks [31]. The coating layer prevents direct contact between the cathode material and the electrolyte, significantly influencing the composition of species in the inner Helmholtz layer (IHL) at the cathode surface. thereby affecting the formation of the CEI during cycling. Liu et al. [32] adjusted the materials to construct a spinel phase coating layer, enabling robust LiF-rich CEI. Zhang et al. [23] demonstrated that uniformly dispersed ZrO₂ nano-rivets and the rich phase boundaries between ZrO₂ and LCO act as nucleation sites, promoting the formation of a uniform, dense CEI. This effectively suppresses continuous interfacial side reactions and CEI degradation, thereby enhancing interfacial stability during the entire cycling process. Li et al. [33] reconfigured an island-shaped AlPO₄/Li₃PO₄ coating on the surface of LiCoO₂ and observed that the coating transformed into a dense, ion-conductive LiAlF₆-rich CEI layer. This layer effectively mitigates HF corrosion, suppresses surface degradation, enhances Li⁺ transport kinetics, and improves thermal stability.

Borate coatings are widely recognized in academia and industry for significantly enhancing the electrochemical performance of layered cathode materials. The Li_xB_yO_z (LBO) coating layer, characterized by a robust B-O bond (809 kJ/mol), bolsters material stability and curtails lattice oxygen release [13]. Additionally, Three-dimensional framework of LBO favors Li⁺ diffusion [34,35]. In industrial processes, surface residual alkali is in-situ converted to an LBO coating via a straightforward solid-state high-speed mixing followed by heat treatment. This method not only enhances Li⁺ transport kinetics at the surface but also preserves the bulk structure of the particles. However, due to the rich compositional variability of LBO can form different phases under various reaction conditions. Therefore, to enable large-scale commercial production of LBO-coated cathode materials two critical challenges must be addressed. (1) What is the optimal temperature for forming an LBO coating with superior electrochemical performance? (2) How does the LBO surface coating affect material structure to enhance electrochemical performance?

In this work, H₃BO₃ was coated to the surface of 50 kg of NCM material in a production workshop using a high-speed mixer. During the calcination phase, H₃BO₃ underwent a reaction with the superficial residual lithium, in situ transforming into an LBO fast ion-conducting layer. This modification strategy not only diminished the quantity of surface residual lithium compounds but also established a novel protective coating on the surface of cathode material. This study presents a preliminary investigation through a series of pilot-scale experiments at different sintering temperatures to evaluate how these temperatures influence the LBO composition of the interfacial coating on NCM materials and electrochemical performance. Moreover, the study examines the role of the artificial coating layer in the formation of the subsequent CEI layer. The amorphous LiBO₂ layer formed at 300 ℃, through an early-stage self-sacrificial reaction, generates trace amounts of BF₄^⁻ in the IHL, thereby optimizing the thickness and uniformity of the CEI and assisting the growth of a stable, robust, LiF-rich electrode interphase, which effectively mitigates the structural degradation of the cathode material.

As shown in Fig. 1a, a simple solid-state method enables the in situ conversion of residual surface alkali species into a fast lithium-ion conductor (LBO) using glassy H₃BO₃. Furthermore, distinct lithium borate phases were obtained under varying thermodynamic conditions. X-ray diffraction (XRD) analysis revealed that all samples maintained the characteristic hexagonal layered α-NaFeO₂ structure with R-3m space group (Fig. 1b). The distinct splitting of the (006)/(012) and (018)/(110) peaks in all samples indicates well-ordered layered structures [36,37]. Following heat treatment, the (003) peak shifted slightly to lower angles, which may be due to the partial extraction of Li⁺ from the NCM reacting with boric acid. The absence of signal peaks corresponding to boron-containing compounds in the coated samples is attributed to the minimal boron content used. The Rietveld refinement results are shown in Fig. S1 and Table S1 (Supporting information), A slight increase in the Li⁺/Ni²⁺ mixing degree was observed in B-400T and B-500T compared to the pristine and lower-temperature treated samples (B-200T and B-300T). This trend is linked to the increased disorder of transition metal lattice at higher temperatures [38]. The amount of surface residual lithium compounds (Li₂CO₃ and LiOH) was measured using electrochemical titration. Fig. 1c summarizes the data on residual lithium before and after H₃BO₃ coating at different temperatures. For the pristine sample, a certain amount of residual lithium compounds was present on the surface before treatment, with a total residual lithium content of approximately 4434 ppm. After H₃BO₃ surface treatment, the residual lithium content decreased to varying degrees. Notably, B-300T and B-400T exhibited significantly lower surface residual lithium contents, measuring 1645 and 1586 ppm, respectively.

SEM analysis was employed to delineate the morphology of the samples, as depicted in Figs. 2a₁-e₁. Both uncoated and coated materials consisted of analogous spherical secondary particles, with the coating process preserving the granular dimensions and primary particle morphology of the polycrystalline layered materials. The bare NCM surface is rough, with many irregularly shaped residual lithium impurities on the primary particles. For B-200T and B-300T, a marked reduction in surface impurities was observed, with the primary particles acquiring a glass-like substance coating. With elevated coating temperatures, B-400T and B-500T exhibited exceptionally smooth surfaces, showcasing a uniformly distributed coating over the primary particle surfaces. To further characterize the micro-morphology and crystal structure of the pristine and coated samples, HRTEM and corresponding fast Fourier transform (FFT) analyses were conducted (Fig. 2). The pristine NCM surface contains visible fragments of amorphous Li₂CO₃. In contrast, the boric acid-coated material exhibits a uniform coating layer, while the intrinsic crystal structure of the bulk NCM remains intact. As the temperature rises, the melting and migration of H₃BO₃, coupled with its reaction with residual lithium, lead to the thinning of the coating layer. Further temperature increases cause Li⁺ to migrate from the bulk material to the surface, promoting the growth of the coating layer. (Figs. 2a₂-e₂). Specifically, B-200T and B-300T samples only showed amorphous structures on the surface. In the B-400T sample, the presence of Li₃BO₃ was detected, with the lattice spacings of 0.3258 nm, corresponding to the (101) crystal plane of Li₃BO₃. The surface of B-500T sample exhibited lattice fringes and FFT patterns consistent with Li₆B₄O₉ (Figs. 2a₃-e₃). Elemental mapping (Figs. 2a₄-e₄), demonstrated a uniform distribution of Ni and O elements across all samples. Additionally, the uniform distribution of B in the coated samples confirmed that the solid-state method achieved a homogeneous H₃BO₃ coating on the NCM surface.

In the Li-B-O ternary system includes several possible phases such as LiBO₂ (Li₂O·B₂O₃), Li₃BO₃ (3Li₂O·B₂O₃), and Li₆B₄O₉ (3Li₂O·2B₂O₃), which can exist in either crystalline or amorphous forms [39]. To analyze the types of interfacial coating substances at different sintering temperatures, a mixture of H₃BO₃ and Li₂CO₃ was calcined to simulate the reactions between H₃BO₃ and residual lithium on the surface of NCM at various temperatures. The surplus Li₂CO₃ in the blend serves to represent the disproportionately high lithium content within the NCM bulk relative to the H₃BO₃ coating content, resulting in XRD patterns dominated by Li₂CO₃ peaks (Fig. 3a). Magnified views in Figs. 3b and c reveal that at 200 ℃, H₃BO₃ did not completely decompose. Given the difficulty in crystallizing H₂B₄O₇, it is inferred that amorphous H₂B₄O₇ formed in the material surface treated at 200 ℃ [40]. The sample treated at 400 ℃ showed characteristic peaks of Li₃BO₃ and Li₂B₄O₇. The sample treated at 500 ℃ exhibited characteristic peaks of Li₆B₄O₉, consistent with TEM results. However, at 300 ℃, no specific Li-B-O peaks appeared, and since LiBO₂ can be synthesized by melting Li₂CO₃ and H₃BO₃ at around 300 ℃ [41], it is hypothesized that the sample treated at 300 ℃ formed amorphous LiBO₂.

In addition, the XPS spectra (Fig. 4) were analyzed to obtain surface information of the pristine NCM and coated samples. In the B 1s spectra of the modified samples, the peak positions of B-O bonds varied with different sintering temperatures. The peak at 192.3 eV in the B-200T sample corresponds to H₂B₄O₇ and H₃BO₃, while the peak at 191.8 eV in the B-300T sample is characteristic of LiBO₂ [42]. Peaks indicative of Li₃BO₃ and Li₂B₄O₇ were located around 191.5 [43] and 191.8 eV [44], respectively. Integrating TEM findings and XRD results of H₃BO₃/Li₂CO₃ simulated reactions at different heating temperatures, the peak for B-400T at 191.75 eV was assigned to a combination of Li₃BO₃ and Li₂B₄O₇. Finally, the peak at 191.6 eV in the B-500T sample is considered to be the characteristic peak of Li₆B₄O₉. Based on the different compositions observed in the coating layer of NCM at various calcination temperatures, we infer that as the temperature increases, the following chemical reactions and occur on the surface of material (Table S2 in Supporting information):

Notably, it was observed that as the sintering temperature increased, the B peaks intensity decreased, suggesting that higher temperatures accelerated the flow of H₃BO₃, leading to a more uniform coating on the particle surfaces. The C 1s spectra showed a peak around 289.4 eV, corresponding to O-C=O bonds, reflecting the surface content of Li₂CO₃. The intensity of the Li₂CO₃ peak decreased in 200, 300, and 400 ℃ coated samples, each for distinct reasons. For B-300T and B-400T samples, the surface Li₂CO₃ is transformed into LiBO₂ and a coating layer of Li₃BO₃ and Li₂B₄O₇, respectively. In contrast, the diminished Li₂CO₃ peak intensity for the B-200T resulted from H₂B₄O₇ and H₃BO₃ coverage, as evidenced by the weakened intensity of both the Li₂CO₃ and lattice Li peaks in the Li 1s spectrum. Although the consumption of surface residual lithium during the formation of Li₆B₄O₉ in the 500 ℃ coated sample, the intensity of the Li₂CO₃ peak does not diminish, suggesting that the elevated temperature prompted lithium migration from the NCM bulk, forming residual lithium impurities on the surface again. This is supported by the heightened peak intensity at 55.2 eV in the Li 1s spectrum of B-500T. In the Li 1s spectrum (Fig. 4c), the peak positions of various LBO compounds are very close to that of Li₂CO₃, resulting in overlap. Specifically, the peak positions of Li₂CO₃ and LiBO₂ are at approximately 55.12 and 55.2 eV, respectively [43,45], and the significant reduction in peak intensity in the Ni 2p spectra of all coated samples (Fig. 4d) further confirms the presence of a coating layer on the surface of the NCM particles.

To evaluate the electrochemical performance, cyclic and rate capability were conducted within a voltage range of 2.8–4.3 V (Fig. 5a). It is evident that the surface boron coating significantly enhanced the cyclic stability of NCM cathode. After 200 cycles, the discharge capacity retention rate of the pristine NCM cathode was only 61.58%, while the retention rates for B-200T, B-300T, B-400T, and B-500T were 92.46%, 88.51%, 87.81%, and 87.42%, respectively. Fig. S2 (Supporting information) displays the initial cycle capacities and coulombic efficiencies of all cathodes. The in-situ converted LiBO₂ coating on B-300T effectively reduces polarization, thereby enhancing sodium storage performance. However, its initial coulombic efficiency is slightly lower, which may be attributed to the self-sacrificial reaction of the LiBO₂ coating and the formation of a uniform, dense CEI on the surface. The rate performance was evaluated at various current density from 0.1 C to 5 C (inset in Fig. 5a). The coated cathode demonstrated superior rate capabilities compared to the pristine NCM. Notably, the B-300T cathode maintained a high discharge capacity of around 183 mAh/g even at a high current rate of 5 C. Among the samples, B-300T exhibited the best electrochemical performance, balancing outstanding cycling stability, high capacity, and excellent rate performance. The discharge curves across varying cycle counts (Fig. 5b), indicate that post 200 cycles, B-300T exhibited the minimal decline in midpoint voltage (Fig. S3 in Supporting information) and retained a reversible capacity of 179.67 mAh/g, whereas the pristine NCM only provided 109.25 mAh/g. This finding further demonstrates that LiBO₂ coating layer obtained by low-temperature H₃BO₃ treatment at 300 ℃ significantly enhances the practical performance of NCM cathode materials. The initial charging curves delineated in Fig. 5c showcase a pronounced elevation in the voltage plateau for B-300T, divergent from the other samples. Moreover, its capacity increase during charging is the slowest, resulting in a higher capacity at the end of the charge state. This suggests that the LiBO₂ coating layer contributes to the suppression of polarization. The cyclic voltammetry (CV) curves at varying scan rates (Fig. S4 in Supporting information). At a scan rate of 1 mV/s, the potential differences between the oxidation and reduction peaks for B-200T, B-300T, B-400T, and B-500T electrodes are 0.268, 0.216, 0.199, and 0.301 V, respectively, which are considerably smaller than the 0.537 V of the pristine NCM. This indicates that surface boron modification enhances electrochemical reversibility and reduces interfacial polarization. Furthermore, there is a linear relationship between the peak current and the square root of the scan rate in the CV tests (Fig. 5d). The slope reflects the diffusion-controlled behavior of Li⁺ in the electrode material. The results indicate that the B-300T electrode has the highest Li⁺ diffusion coefficient, demonstrating the best Li⁺ kinetic. The electrochemical impedance spectroscopy (EIS) spectra fitting results (Fig. S5 in Supporting information) obtained before cycling further confirm that B-300T exhibits the lowest charge transfer impedance. This superior ion transport property is the primary reason for its excellent rate performance. Additionally, a comparative analysis of our work with previous studies on boric acid coating reveals the superior overall electrochemical performance of our samples (Table S3 in Supporting information and Fig. 5e). This comparison provides further insights into how temperature influences the composition of the coating layers. The superior electrochemical performance is attributed to the amorphous LiBO₂ coating on B-300T surface. LiBO₂ offers several advantages over other lithium-rich borates. Its lower viscosity reduces hygroscopicity and lithium carbonate retention [46]. Moreover, its amorphous structure facilitates the formation of a functional CEI layer during cycling, enhancing the overall electrochemical performance.

In order to further highlight the enhancements in material performance provided by the amorphous LiBO₂ coating, we evaluated the air stability of the materials and their cycling performance under extreme conditions. Fig. S6 (Supporting information) demonstrates that after seven days of air exposure, the Li₂CO₃ impurities on the surface of B-300T particles were significantly reduced compared to those on pristine NCM particles. Furthermore, the value of R_ct derived from EIS fitting was lower. As shown in Fig. 5f, when the exposed materials were assembled into coin cells and cycled 200 times at 1 C, B-300T achieved a reversible capacity of 143.5 mAh/g and a retention rate of 69.2%, markedly higher than the pristine NCM electrode (97.3 mAh/g and 54.9%). The amorphous LiBO₂ coating effectively improves the air stability of the material, which is expected to significantly lower storage costs during large-scale production. High temperatures accelerate oxygen release and structural degradation, leading to detrimental interfacial side reactions, harmful phase transitions, and lattice oxygen release. While all materials exhibited enhanced capacity at 50 ℃ compared to room temperature, they also experienced a faster performance decline (Fig. 5g). Notably, B-300T maintained a capacity retention rate of 76.0%, significantly higher than 49.8% of the pristine NCM, further validating the protective effectiveness of the amorphous LiBO₂ coating. We also assessed the cycling stability of the material in pouch full batteries relative to graphite anodes, further exploring the practical application of this interfacial modification approach (Fig. 5h and Table S4 in Supporting information). After 400 cycles at a current density of 1.5 C, the B-300T||Gr cell, with a capacity of 0.8 Ah, retained 91.12% of its capacity, significantly outperforming the commercial pristine NCM cell, which retained only 82.51%. As shown in Fig. 5i, the B-300T||Gr cell maintained a nearly stable median voltage around 3.5 V throughout the cycling process, while the pristine NCM||Gr cell exhibited more considerable voltage instability early in the cycles, with a significant drop in voltage after 400 cycles. High mass loading electrochemical measurements (Fig. S7 in Supporting information) demonstrate that B-300T delivers superior electrochemical performance relative to the pristine NCM sample. Moreover, differential scanning calorimetry (DSC) analysis was conducted to assess structural stability under high temperatures (Fig. S8 in Supporting information). When subjected to heat, materials experienced structural degradation and phase changes, leading to Me-O bond cleavage and lattice oxygen release, which produced considerable heat and posed safety risks. The exothermic onset temperature of B-300T was higher than that of the pristine NCM (240.8 ℃ vs. 220.5 ℃), with a lower heat release.

To elucidate the impact of the amorphous LiBO₂ coating on the degradation mechanisms of cycling performance, we conducted further analyses of the cathode materials post-cycling. Fig. 6a presents the XRD patterns of fresh and cycled (200 times) pristine NCM and B-300T. Both materials retained their layered structure after cycling; however, the (003) peak for B-300T was stronger and narrower than that of pristine NCM, indicating reduced lattice distortion during the 200 cycles. The adjacent magnified image clearly shows that the (003) diffraction peak of B-300T shifted less than that of pristine NCM, suggesting that diminished lattice distortion facilitates more reversible phase transitions, attributed to the protective effect of the amorphous LiBO₂ coating. Fig. 6b reveals noticeable microcracks in the cycled pristine NCM particles, while B-300T particles remained intact, demonstrating that the surface coating alleviates stress and suppresses particle cracking. XPS analysis was performed to explore the composition of the CEI (Figs. 6c and d, Fig. S9 in Supporting information). The C=O and O-C-O signals correspond to organic compounds generated from the decomposition of carbonate solvents, while metal salts composed of F and P primarily arise from the decomposition of LiPF₆. After the initial cycle, the F 1s peak of pristine NCM was stronger than that of B-300T, indicating more severe side reactions on the pristine NCM surface during early cycling (Fig. S10 in Supporting information). After 200 cycles, Fourier transform infrared (FTIR) analysis (Fig. S11 in Supporting information) reveals that B-300T exhibits stronger RCO₂Li/ROCO₂Li peaks, indicating that the amorphous LiBO₂ layer effectively suppresses the further decomposition of organic compounds species. No significant differences were observed in the components corresponding to P 1s between the coated material and pristine NCM. Notably, The F 1s spectrum of B-300T prominently displays a strong LiF peak, a finding further corroborated by consistent results from FTIR analysis. This indicates a high content of beneficial LiF in the CEI layer, which enhances its mechanical strength and structural stability. Further analysisof the B 1s XPS spectra of B-300T after the first and 200^th cycles (Fig. 6d) reveal that, during initial cycling, direct interaction between the electrolyte and amorphous LiBO₂ forms trace BF₄^⁻ anions in the IHL [47]. These anions facilitate the development of a robust and LiF-rich CEI [48]. As cycling progresses, the dense CEI layer gradually envelops the entire particle surface, with the XPS spectrum eventually showing only B-O peaks. The dynamic evolution mechanism of the CEI for B-300T can be summarized as follows: During the initial cycling stage, the LiBO₂ coating undergoes a self-sacrificial reaction upon direct contact with the electrolyte, generating trace amounts of BF₄⁻ anions within the IHL. This interfacial chemical process induces the subsequent enrichment of robust inorganic LiF. During long-term cycling, a uniform and dense LiF-rich CEI fully covers the particle surface. At this stage, the LiBO₂ coating and the stabilized CEI form a dual-passivation barrier, ensuring a stable electrode interface.

To illustrate the influence of the amorphous LiBO₂ coating on CEI formation, TEM was employed to examine the interfacial conditions of the cycled materials (Figs. 6e and f). The CEI layer on pristine NCM was thick and uneven, impeding Li⁺ transport. In contrast, the CEI on B-300T particles was more uniform and significantly thinner, measuring just 6.6 nm. This suggests that the amorphous LiBO₂ coating on B-300T promotes the uniform growth of the CEI during cycling. Additional, on the particle surface, Region Ⅰ of pristine NCM exhibits a rock-salt NiO phase, indicating the occurrence of detrimental irreversible phase transitions, which is likewise supported by the XPS results showing more Ni²⁺ peaks (Fig. S12 in Supporting information). In contrast, Region Ⅲ of B-300T retains a well-preserved layered structure. Fourier transform analysis of the image also revealed that while pristine NCM exhibited amorphous structural features, the coated sample retained its layered structure, further demonstrating the protective efficacy of the coating in preserving the crystal structure. Furthermore, Fig. S13 (Supporting information) presents the EIS spectra after 200 cycles, along with the equivalent circuit model and fitting results. Compared to the pristine NCM, which has an R_CEI of 22.07 Ω and an R_ct of 124.30 Ω, the coated samples show varying degrees of reduction in both R_CEI and R_ct. Among them, the B-300T sample exhibits the lowest R_ct (62.70 Ω), indicating that the amorphous LiBO₂ coating assists the growth of a multifunctional CEI layer during the cycling process, which lowers the energy barrier for Li⁺ migration, thus improving charge transfer kinetics [9]. Overall, the amorphous LiBO₂ coating on the surface of B-300T facilitates the formation of a robust and uniform CEI layer through the following mechanisms (Fig. 7). First, the LiBO₂ layer acts as a protective barrier and enhances Li⁺ kinetics, reducing side reactions between the highly oxidative nickel-rich cathode and the electrolyte during early cycling stages. This effectively suppresses excessive CEI growth. Second, when the LiBO₂ surface reacts with the electrolyte, it forms BF₄⁻ anions in the IHL, which promote the formation of a robust and homogeneous multifunctional CEI layer enriched with LiF.

In summary, this study utilized a solid-state method to achieve in-situ coating with H₃BO₃, converting residual lithium on the material surface into a mechanically robust and ionically conductive LBO coating layer. The composition of the surface LBO coating layer varies with different sintering temperatures: At 200, 300, 400, and 500 ℃, the surface coating layers consist of amorphous H₂B₄O₇ and H₃BO₃, amorphous LiBO₂, Li₃BO₃ and Li₂B₄O₇, Li₆B₄O₉, respectively. Notably, the amorphous LiBO₂ coating layer on the surface of B-300T maintains particle integrity, mitigates irreversible phase transitions and interfacial side reactions, and induces the formation of a stable functional CEI layer during cycling. Consequently, after 200 cycles, the B-300T electrode retained a reversible capacity of 179.67 mAh/g with a discharge capacity retention rate of 88.51% and a 0.8 Ah pouch-type full cell exhibited a negligible 8.88% capacity loss after 400 cycles. This study elucidates the optimal phase of the LBO coating layer and provides insights into the impact of surface coating layers on the CEI layer. It offers a simple and effective strategy for enhancing the electrochemical performance of nickel-rich layered oxide cathodes, making it suitable for large-scale commercial production.