Structural and morphological characterizations
X-ray diffraction (XRD) was employed to analyze changes in the crystal structures of NCM, NCM@LP-1, NCM@LP-3, NCM@LP-5, NCM@LP-10, and NCM@LP-20. As shown in Fig. 1a, the peak corresponding to (003) does not shift [30], indicating that the addition of a moderate amount of PPA does not alter the layered structure of the NCM raw materials. The distinct splitting of (006)/(102) and (018)/(110) confirms the presence of an undamaged layered structure [31], and a higher value of the peak intensity ratio I(003)/I(104), indicates a favorable layered structure [32]. The position and intensity of diffraction peaks before and after the surface modification with Li3PO4 are consistent with the characteristics of a hexagonal crystal system. The structures of NCM, NCM@LP-1, NCM@LP-3, and NCM@LP-5 exihibit a well-ordered layered structure, α-NaFeO2. No impurity peaks can be observed when the amount of PPA is less than 5%, suggesting that the lattice structures of the electrode materials remain largely unaffected by the minor addition of PPA. However, excessive surface modification (NCM@LP-10 and NCM@LP-20 in Figs. S1a and b) adversely affects the crystal structure. The accumulation of a large amount of PPA on the particle surfaces resulted in complete coverage of some NCM particles and penetration into their interiors through gaps, inducing stress and unfavorable reactions. Moreover, an excessive coating amount can lead to changes in the primary structure [33].
X-ray photoelectron spectroscopy (XPS) was utilized to determine the elemental composition on material surface. The presence of C and P in NCM@LP-3 can be confirmed from Figs. 1b and c, respectively (Ni, Co, and Mn in Figs. S1 c, d and e). The C 1s spectrum in Fig. 1b exhibits peaks at 289.4 and 283.5 eV corresponding to the C = O bond and C-C bond. It indicates the removal of most of the Li2CO3, which is consistent with the Fourier-transform infrared (FT-IR) analysis shown in Fig. 1d. Figure 1c presents the spectrum of P 2p3/2 and P 2p1/2 at 133.8 and 134.1 eV. The presence of P–O bonds from the metal phosphate confirms the formation of Li3PO4 [34]. The XPS results are consistent with the XRD analysis. Moreover, Fig. S1 c shows the Ni 2p3/2 peak at 855.12 eV and the Ni 2p1/2 peak at 872.5 eV, and Fig. S1 d displays the Co 2p3/2 and Co 2p1/2 peaks at 779.5 eV and 794.7 eV, respectively. The Mn 2p3/2 peak at 642.1 eV and the Mn 2p1/2 peak at 654.3 eV are shown in Fig. S1 e.
To investigate the presence and reaction extent of surface Li2CO3, FT-IR analysis was conducted. As depicted in Fig. 1d, the Li2CO3 peak in the original NCM is prominent, whereas it is weakened at 860 cm–1 in NCM@LP-3. Additionally, the presence of P = O bonds was detected by FT-IR, as shown in Fig. 1d. These results indicate the successful removal of Li2CO3 from the NCM surface, with the formation of Li3PO4 through surface chemical reactions. The bond dissociation energies of P = O, Ni–O, Co–O, and Mn–O are 596.6, 391.6, 368, and 402 kJ mol–1, respectively. Among these, P = O exhibits the highest bond dissociation energy. Following the surface modification with Li3PO4, the adverse effect of transition metal dissolution in NCM@LP-3 is mitigated. The stability of P = O on the surface of NCM@LP-3 enhances the stability of Ni–O, Co–O, and Mn–O bonds, thereby reducing the likelihood of lattice oxygen escape [35]. It is well-known that the instability of surface Ni/Co/Mn leads to their dissolution in the electrolyte, resulting in the collapse of layered nickel structures and the separation of lattice oxygen from the interior layered structure.
Scanning electron microscope (SEM) images Figs. 2a and b illustrate the presence of Li2CO3 impurities on the surface of the original NCM particle, while Figs. 2d and e exhibit the presence of Li3PO4 on the NCM@LP-3 particle surfaces. Additionally, transmission electron microscopy (TEM) images in Figs. 2c and f reveal the presence of a thin layer of Li3PO4 on the NCM@LP-3 surface. These observations can further support the results obtained from XRD, XPS, and FT-IR analyses, which collectively confirm the successful conversion of Li2CO3 to Li3PO4.
To explore the influence of our temperature-controlled treatment technology, SEM images of NCM@LP-3 and NCM@LP-3-A are shown in Figs. 2e and S1 f, respectively. The surface reaction of NCM@LP-3 after temperature control is uniform, which can be attributed to the improved dispersal of PPA in the solvent, facilitated by its fluid-like behavior at 50°C. Consequently, PPA can more easily react with the surface Li2CO3. In contrast, the particle surface of NCM@LP-3-A exhibits irregularities, with a mixture of large and small, bumpy, and uneven features. Fig. S1 f clearly shows a prominent spot, resulting from the adhesion of undissolved viscous PPA to the surface due to insufficient reaction of PPA with residual Li2CO3. As a consequence, the cycling performance of NCM@LP-3-A begins to decay after 20 cycles,
To gain further insights into the specific composition of particles attached to the surface of NCM@LP-3, energy-dispersive X-ray spectroscopy (EDS) was employed to characterize and analyze the elemental distribution on the NCM@LP-3 surface. Figure 2g shows the presence of Ni, Co, Mn, C, O, and P on the surface of NCM@LP-3. Additionally, as depicted in Fig. 2g, the distribution of the P element appears homogenous on NCM@LP-3 surfaces, which is attributed to the surface modification process involving PPA. The utilization of temperature control at 50°C during PPA stirring ensures the complete dissolution of this sticky substance (PPA) in ethanol. Ethanol, being the solvent, can form hydrogen bonds with PPA, increasing its solubility. As a result, PPA can react uniformly with Li2CO3.
Electrochemical performances
Figures 3a and b depict the charge and discharge curves of NCM and NCM@LP-3 cathodes. At a rate of 0.1 C, NCM@LP-3 exhibits a higher initial discharge capacity of 218.1 mAh g–1 than that of the original NCM (214.3 mAh g–1). This superior performance can be attributed to the active Li in Li3PO4, which facilitates the removal of trace lithium ions during delithiation/lithiation processes. As shown in Fig. 3c, NCM@LP-3 demonstrates excellent long-cycle performance in the potential window of 2.75–4.3 V versus Li/Li+ at 1.0 C. After 100, 200, 300, 400, and 500 cycles, NCM@LP-3 exhibits capacity retention rates of 97.28, 90.84, 84.19, 78.13, and 72.48%, respectively. In comparison, the pristine NCM shows capacity retention of 96.92, 83.13, 70.26, 59.42, and 51.01%. While the capacity retention rate of NCM@LP-3 differs by less than 1% from NCM at 100 cycles, it shows an increase of nearly 20% by 500 cycles. Furthermore, as shown in Fig. 3d, NCM@LP-3 demonstrates the excellent long-cycle performance at 5.0 C, with capacity retention rates of 92.25, 88.48, and 84.19% after 100, 200, and 300 cycles, respectively. In comparison, the original NCM exhibits capacity retention rates of 85.32, 83.27, and 70.26%. NCM@LP-3 showcases an improvement of nearly 12% in capacity retention compared with NCM after 300 cycles.
The differences among NCM, NCM@LP-1, and NCM@LP-5 are not substantial after 200 cycles, but NCM@LP-3 exhibits the best performance in terms of capacity retention rate and discharge capacity (Fig. S2 b). It can be attributed to the moderate amount of Li3PO4 incorporated in NCM@LP-3[36]. The presence of Li3PO4 enables the utilization of lithium ions by converting inert lithium from Li2CO3 into active lithium in Li3PO4. The discharge capacities of NCM, NCM@LP-1, NCM@LP-3, and NCM@LP-5 at 0.1, 0.5, 1.0, 2.0, and 5.0 C are shown in Fig. 3f. Among these, NCM@LP-3 demonstrates superior performance across all the tested rates, with specific capacities of 201.5, 196.4, 185.7, 177.6, 168.1, and 146.0 mAh g–1. In comparison, the original NCM exhibits specific capacities of 194.0, 192.0, 178.5, 166.8, 152.8, and 126.2 mAh g–1. The specific capacity of NCM@LP-3-A below 90 mAh g–1 at 100 cycles (Fig. 4e).
The performance of NCM@LP-1 is comparable to that of NCM (Fig. S2 b). However, NCM@LP-5 shows poorer performance than NCM. The superior performance of NCM@LP-3 can be attributed to the addition of a moderate amount of PPA. The Li3PO4 layer can react with a trace amount of HF, preventing direct reactions between the electrolyte and the transition metals (Ni, Co, or Mn) [37], which alleviates side reactions [36]. The phosphate radical in Li3PO4 has a tetrahedral structure, with each P atom connected to oxygen through a double bond. The large pore framework structure of phosphate radicals facilitates the transmission of lithium ions. Additionally, phosphate radicals can form covalent bonds with nickel ions, thereby stabilizing the electrode–electrolyte interface and hindering the removal of phosphate from the surface. Furthermore, the release of oxygen from the lattice is mitigated, preventing the collapse of the layered structure. In summary, the consumption of HF by Li3PO4 and the stabilization of the electrode–electrolyte interface, along with the favorable spatial structure of Li3PO4 for lithium ions transmission and stability of the phosphorus–oxygen bond, collectively contribute to the enhanced performance of NCM@LP-3.
To investigate the influence of lithium-ion diffusion rate after surface modification with PPA, electrochemical impedance spectroscopy (EIS) tests were conducted on NCM and other samples. The equivalent circuits were shown in the inset of Fig. 4a and b. R1, R2, R3, CPE, and W1 represent the ohmic resistance (the intercept on the real axis), contact resistance (the first semicircle at high frequencies), charge transfer resistance (the second semicircle at low frequencies), capacitance of the electrode/electrolyte double layer, and the Warburg resistance (the oblique line at low frequencies), respectively. As shown in Figs. 4a and b, R2 of NCM@LP-3 is lower than that of the original NCM. This is because the tetrahedral structure of Li3PO4 reduces the impedance for transmission between the electrolyte and cathode interface.
As shown in Fig. 4b, both R2 and R3 of NCM@LP-3 are lower than those of NCM after 500 cycles. However, as the cycle number increases, R2 and R3 tend to increase. It can be attributed to the presence of impurities on the surface of the cathode material, which dissolves into the electrolyte during cycling. These impurities can block the transmission of lithium ions. The increase in R2 and R3 is more pronounced in NCM compared to NCM@LP-3, which is due to the formation of microcracks in the particles. As illustrated in Fig. 4c, the microcracks in NCM are more severe compared to those in NCM@LP-3 (Fig. 4d).
Figure 5a shows that all samples exhibit three pairs of redox peaks from 2.75 to 4.3 V at 0.1 mV s–1, indicating the moderate surface modification with PPA does not affect the layered structure and electrochemical behaviors. To further investigate the reasons for the excellent electrochemical performance of NCM@LP-3, cyclic voltammetry (CV) was conducted at various sweep speeds (ranging from 0.1 to 1 mV s–1) within the voltage window of 2.75–4.3 V. Figures 5b and d show that NCM@LP-3 maintains the presence of the three pairs of redox peaks even at higher sweep speeds, implying minimal alteration in its electrochemical behaviors at high scan rates. Figure 5c demonstrates that the peak current of NCM@LP-3 changes roughly linearly with the sweep speeds ranging from 0.1 to 1 mV s–1, exhibiting a pseudo-capacitance effect.