1 Introduction

With the wide application of Li-ion batteries in electric vehicles and energy storage systems, the demand of developing the high performance of cathodes for the high energy density and long-life span batteries is gradually increasing [1,2,3]. As a candidate, the conventional Ni-rich layered oxide cathode (LiNixCoyMnzO2 (NCM), x > 0.8) is regarded as the auspicious cathode material, due to its attractive reversible capacity (> 200 mAh·g−1) [4,5,6]. However, the polycrystalline Ni-rich layered oxide cathodes (PC-NCM) are prone to structural degradation and severe side reaction during the long-term cycling process, resulting in battery performance deterioration [7,8,9,10,11,12,13]. In addition, the formation of surface residual lithium species (LiOH and Li2CO3) could block lithium diffusion and decrease the reversible capacity [14,15,16,17,18]. Severe impedance growth during the charging-discharging process ultimately causes the increased polarization and substantial safety issues [19, 20].

To address these issues, extensive efforts have been carried out to improve the cycle performance of NCM in high-voltage operation [11, 21,22,23,24,25,26], including elemental doping (Mg, Al, Ti, Sn, etc.) and surface coatings (Al2O3, AlPO4 and Li3PO4, etc.). However, the vital defects of severe intergranular cracks for the PC-NCM cathode materials are still inevitable, especially under harsh operating conditions (the elevated temperature and high cutoff voltage, etc.). In our previous works, the capacity retention of the single crystal NCM (SC-NCM) cathode is superior to that of PC-NCM cathode, due to its robust structure and fewer grain boundaries [27,28,29,30,31,32]. It means that the micron-sized primary particles of SC-NCM cathodes could effectively suppress the micron scale structural degradation, including the appearance of intergranular cracks. In addition, with the increase in the charging cutoff voltage, SC-NCM cathodes could release more Li+ and posse higher specific capacity value, which is becoming a research hot spot in academia and industry. Nevertheless, the interfacial instability and bulk structural degradation of the Ni-rich SC-NCM cathodes at high operation potentials would cause rapid capacity decay and voltage fading. Therefore, it still has a great challenge to develop the high-voltage Ni-rich SC-NCM (Ni > 0.8) cathodes with superb cycling stability and the high capacity at high voltage (≥ 4.6 V).

The foreign elements doping in the SC-NCM cathode could be an effective way to restrict the internal stress at a deep discharging state [33,34,35]. However, the current doping strategies are commonly executed by the after-treatment, making it challenging to ensure a uniform distribution of foreign element ions in the bulk of SC-NCM material. Meanwhile, the relevant electrochemical performances of the doping cathode materials with the different treatments are seldom discussed in the literature, which plays a crucial role in the enhanced structural stability and electrochemical performances of the SC-NCM cathodes.

In this work, a Zr-doped Ni-rich single-crystalline LiNi0.827Co0.11Mn0.06Zr0.003O2 (Zr@SC-N83) material was synthesized by in situ doping method. The introduction of trace Zr element exhibits the following positive effects: (1) effectively enhancing electrochemical reactivity; (2) bulk doping improves the structural stability; (3) inhibiting the side reactions with the uniformly surface distribution of Zr4+; (4) improving Li+ transportation via increasing the interlayer distance. The positive effects of Zr-doping in the surface and bulk of SC-NCM cathode material could enable better electrochemical performances, compared with the undoped sample.

2 Experimental

2.1 Synthesis of Zr@SC-N83 material

The spherical precursor Ni0.827Co0.11Mn0.06Zr0.003(OH)2 of Zr@SC-N83 samples were synthesized via a co-precipitation method [28, 36]. A stoichiometric amount of NiSO4·6H2O, CoSO4·7H2O, and MnSO4·5H2O (the concentrations for all samples are 2 mol·L−1) solution were pumped into the reaction tank. Zr(SO4)2·4H2O (0.5 mol·L−1), NaOH (5 mol·L−1, the precipitation agent), and NH3·H2O (4 mol·L−1, chelating agent) solutions were separately pumped into the tank reactor as mentioned above under N2 atmosphere. After filtering the precipitations, washing, and drying, the as-prepared precursor was finally heated in a vacuum oven at 120 °C for 12 h. The precursor was thoroughly mixed with LiOH·H2O (M:Li ratio = 1:1.06, where M represents Ni/Co/Mn ions) and subsequently calcined at 500 °C for 6 h. Finally, the Zr@SC-N83 materials are obtained to calcine at 830 °C for 10 h in an oxygen atmosphere. For comparison, the bare LiNi0.83Co0.11Mn0.06O2 (SC-N83) samples were fabricated by a similar method without adding the Zr containing salt. Besides, the after-treated Zr@SC-N83 samples were prepared by doping after precursor synthesis.

2.2 Material characterization

Chemical compositions of Zr@SC-N83 particles were tested by inductively coupled plasmaoptical emission spectrometry (OPIMA 8300, Perkin Elmer). Crystalline structures of all samples were determined by X-ray diffraction measurement (XRD, PANalytical Empyrean), where 2θ was corrected from 5° to 120° at 2 (°)·min−1 and transmission electron microscopy (TEM, JEOL 2100F, JEOL). In addition, X-ray photoelectron spectroscopy (XPS) tests of all samples were conducted by Thermo Fisher ESCALAB 250Xi. Before XPS testing, the cycled cathodes were washed with dimethyl carbonate (DMC) three times.

2.3 Electrochemical measurements

The cathode slurry containing the active materials, acetylene black, and polyvinylidene fluoride (PVDF) at a weight ratio of 8:1:1 was coated on aluminum foil, where the area loading of the active material is about 8.0 mg·cm−2. Coin-type cell (CR2032) was assembled with a lithium metal anode, the prepared Zr@SC-N83 cathode, and a commercial electrolyte of 1.0 mol·L−1 LiPF6 in ethylene carbonate/ethyl methyl carbonate/diethyl carbonate (EC/EMC/DEC, LB037 obtained from DoDo Chemical Co., Ltd, Soochow, China). A LAND battery cycler was applied to test the cycling performances of the cells at 2.75–4.6 V (vs. Li/Li+) and 25 °C. A CHI660D electrochemical station (Chenhua, Shanghai) was employed to acquire electrochemical impedance spectra (EIS) over the frequency range of 1×10–2–1×105 Hz at 25 °C. The state of charge of the cells for EIS, XPS, SEM and TEM measurements are the fully discharging state.

3 Results and discussion

3.1 Morphology and composition

The representative electron microscopy images of the as-prepared Ni0.827Co0.11Mn0.06Zr0.003(OH)2 precursor and Zr@SC-N83 material are shown in Fig. 1. The precursor consists of aggregated particles and shows an apparent secondary particle of 3.5–5.5 μm (Fig. 1a). After the modified solid-state reaction at high temperature, Zr@SC-N83 cathode material exhibited micron-sized polyhedral particles with a smooth surface (Fig. 1c), similar to bare SC-N83 particles (Fig. 1b). To verify the uniformity of element distribution, the Zr@SC-N83 sample was examined by EDS mapping (Fig. 1d). Ni, Co, Mn and Zr elements were uniformly distributed on the surface of Zr@SC-N83 particles. SEM–EDS line scanning of the cross-section for the Zr@SC-N83 primary particle further demonstrates that Zr ions are doped into the bulk particle (Fig. 1e, f). It means that the transition metal ions in Zr@SC-N83 lattice were substituted by Zr4+, which could mitigate the parasitic reaction at the electrode/electrolyte interface and improve the structural stability. The corresponding HRTEM images in Fig. 1g, h show a precise and uniform interplanar distance of 0.47 nm, corresponding to (003) crystal planes of the layered structure of NCM.

Fig. 1
figure 1

SEM images of a precursor, b bare SC-N83 and c Zr@SC-N83 materials; d SEM image and corresponding elemental mappings of a single Zr@SC-N83 particle; e cross-section SEM image of a single Zr@SC-N83 particle and f corresponding EDS linear scan mapping of transition metal elements; g, h HRTEM images of a Zr@SC-N83 particle

Full size image

Moreover, XPS spectrum and XRD diffraction patterns of the bare SC-N83 and Zr@SC-N83 samples are shown in Fig. 2 and Table S1. Two peaks of Zr 3d of Zr@SC-N83 material at different etch depths (0, 200 and 400 nm) from particle surface are located at 181.59 and 183.98 eV, respectively (Fig. 2a). And the intensity of Zr from the outer surface to the interior of Zr@SC-N83 is almost the same. It suggests that Zr is homogeneous distribution, which is consistent with by cross section SEM–EDS mapping and liner scanning. In addition, XRD patterns of Zr@SC-N83 powder (Fig. 2b, c and Table S2) could be corresponding to the typical layered hexagonal structures with a space group R-3m (JCPDS No. 87-1562), which exhibits the marked splitting of pair reflections (006)/(102) and (108)/(110) [37, 38]. In general, Zr4+ doping with the relatively large ion radius (0.072 nm) could stable the structural stability of NCM materials and increase the lithium ion transportation during cycling process. Therefore, it could be a good method to improve the cycle performance of the Zr@SC-N83 sample, especially at the elevated cutoff voltage.

Fig. 2
figure 2

a XPS results of Zr 3d for Zr@SC-N83, b, c XRD pattern and Rietveld refinement results of bare SC-N83 and Zr@SC-N83 materials

Full size image

3.2 Electrochemical test

To reveal the fundamental effects of Zr4+ doping, a series of electrochemical tests were performed at 2.75–4.6 V in 2032-type coin cells with the Zr@SC-N83 and the bare SC-N83 cathodes in Fig. 3. Figure 3a displays the initial charge-discharge profiles of the Zr@SC-N83 and the bare SC-N83 cathodes at 2.75–4.6 V and 0.1C. In detail, the charge-discharge curves for both cathodes display similar shapes, indicating that no extra electrochemical reaction occurred after in situ Zr4+-doping treatment. Both the Zr@SC-N83 and bare SC-N83 cathodes exhibit the similar initial discharge capacity (about 218 and 210.5 mAh·g−1, respectively) at 2.75–4.6 V/0.1C. Moreover, the Coulombic efficiency of the Zr@SC-N83 cathode at initial cycle is slightly larger than that of the bare SC-N83 cathode, which is attributed to the enhanced reversibility of phase transition between H2 and H3 after Zr4+-doping. Notably, the Zr@SC-N83 cathode exhibits a higher discharge capacity of 171.9 mAh·g−1 after 100 cycles at 1.0C and 2.75–4.6 V in Fig. 3b, maintaining satisfactory capacity retention of 85.5%, which is superior to that of the bare SC-N83 cathode (75.7 mAh·g−1 and 36.2%, respectively). The significantly enhanced cycling stability of the Zr@SC-N83 cathode might be ascribed to Zr4+ doping, providing additional Li+ deintercalation channels that reduce the irreversible specific capacity losses and electrochemical polarization voltages [39]. Furthermore, the different contents of the Zr-doping in Zr@SC-N83 cathodes and the after-treated Zr-doped SC-N83 (noted as the Zr/SC-N83) were adjusted to achieve the optimized electrochemical performance in Fig. S1, which shows that the best cycling stability is obtained by 0.3 wt% Zr doped into the SC-N83 cathode. It also exhibits the comparable results at such high voltage in literature (Table S4), demonstrating the best cycling stability at 4.6 V. Besides, the Zr@SC-N83 cathode shows the fast lithium-ion diffusion kinetics, compared to the bare SC-N83 cathode, according to the EIS fitting results (Fig. 3f and Table S3). It is well known that the more considerable \(D_{\text{Li}^{+}}\) (lithium-ion diffusion coefficients) value indicates the faster lithium-ion diffusivity. Therefore, the Zr@SC-N83 cathode could significantly reduce the electrochemical resistance of batteries during cycling process, which is also consistent with the changes in midpoint voltage values during 100 cycles for these two samples. As shown in Fig. 3c, the SC-N83 cathode exhibits the fast mid-voltage values decay as increasing cycles, speculating the severe structural degradation. In contrast, the smaller polarization voltage is achieved after the Zr-doping (ΔV = 0.833 V, ΔV stands for the midpoint voltage difference between the SC-N83 cathode and the Zr@SC-N83 after 100 cycles at 2.75–4.6 V), indicating the better cycling stability for the Zr@SC-N83 cathode. Additionally, the corresponding differential capacity (dQ/dV) profiles under different voltage platforms are analyzed in Fig. 3e. For the bare SC-N83 electrode, the cathode peak sharply moves out of the voltage window with the decrease of peak intensity. Finally, it disappears during the cycling process, which might be caused by electrode polarization. At the same time, with the continuously left shift of the cathodic peak at 3.75–4.0 V, a rapid decrease of the initial discharge voltage can be observed during the cycle, indicating that the bare SC-N83 cathode had a severe polarization behavior and rapid capacity decline. On the contrary, the cathodic peak of the Zr@SC-N83 electrode exhibits a negligible attenuation after 100 cycles, as shown in Fig. 3d. It is further demonstrated that the polarization behavior of Zr@SC-N83 is effectively inhibited, significantly reducing the parasitic reactions at the electrode/electrolyte interface and improving the structural integrity of the Zr@SC-N83 electrode at ultrahigh cutoff voltage of 4.6 V.

Fig. 3
figure 3

Electrochemical performances of Zr@SC-N83 and bare SC-N83 cathodes at 2.75–4.6 V: a initial charge–discharge curves at 0.1C; b cycle performance at 1.0C; c–e corresponding midpoint voltage and dQ/dV curves during 100 cycles at 1.0C; f EIS results before cycling and after 100 cycles at 1.0C (Z’ stands for real part impedance, Z” stands for imaginary part impedance); g rate performance of these cathodes at 0.1C to 5.0C

Full size image

To investigate the interfacial reaction kinetics of the Zr@SC-N83 and the bare SC-N83 cathodes, EIS measurements results of these two samples before and after 100 cycles are shown in Fig. 3f and EIS results are simulated using the equivalent circuit plots (Fig. S2). All Nyquist plots consist of a depressed semicircle at high frequency and a slope at low frequency, corresponding to the charge-transfer resistance at the electrode–electrolyte interface (Rct) and Warburg impedance, respectively. The values of the fitting resistance for the bare SC-N83 and Zr@SC-N83 cathodes before and after 100 cycles are listed in Table S3. Rct values of the SC-N83 and Zr@SC-N83 cathodes are 151.4 Ω/95.54 Ω (before cycle) and 461.22 Ω/194 Ω (after 100 cycles), respectively. The more negligible impedance of the Zr@SC-N83 cathode confirms that the effect of Zr-doping can reduce the resistance at the electrode–electrolyte interface, thus enhancing the electrochemical performance. The corresponding \(D_{\text{Li}^{+}}\) are calculated by the following formula and listed in Table S3, where the detail values of the bare SC-N83 and Zr@SC-N83 cathodes are 4.53 × 10–13 and 1.50 × 10–12 before cycles, and 6.17 × 10–14 and 9.69 × 10–14 cm2·s−1 after 100 cycles, respectively. It means that the Zr-doped modified SC-N83 material exhibits a higher Li-ion diffusion coefficient than the bare Ni-rich NCM material before and after cycling, which is consistent with the results of the cycling performance.

As the current density increases, the Zr@SC-N83 cathode delivers a better rate performance at 0.1C, 0.2C, 0.5C, 1.0C, 2.0C and 5.0C, compared with the bare SC-N83 cathode (Fig. 3g). Both samples delivered similar discharge capacities at 0.1C and 0.2C. Nevertheless, with the lifting of the current density, the difference in discharge capacity value for the bare SC-N83 and Zr@SC-N83 cathodes is increasing rapidly. Especially at the current rate of 5.0C, the Zr@SC-N83 sample also exhibits a higher capacity of 175.6 mAh·g−1, compared to that of the bare SC-N83 cathode (143.4 mAh·g−1). It is worth noting that the foreign Zr-doping modification strategy could effectively improve the rate capability of the SC-N83 cathode due to its more significant lithium diffusion coefficient and the relatively low mitigated side reactions at the electrode/electrolyte interface. Therefore, the excellent comprehensive electrochemical performance of the Zr@SC-N83 cathode is contributed to improving the structural stability and inhibiting side reactions.

It is well known that crack formation and surface deterioration would play a critical role in the battery performance, especially during the charging-discharging process at high voltage [27,28,29]. First of all, nano/micro-cracks may quickly occur at the vast volume shrinkage and expansion during the H2-H3 phase transition beyond 4.2 V, influencing the structural integrity. On the other hand, high-voltage induced electrolyte decomposition may attack the cathode material surface, forming a vulnerable cathode-electrolyte interface. Zr-modification strategy could provide an effective way to address these issues. The morphologies of the cycled bare SC-N83 and Zr@SC-N83 particles after 100 cycles at 4.6 V and 1.0C reveal this positive effect of the Zr-doping (Fig. 4). As shown in Fig. 4e–g, the microstructure of the bare SC-N83 particle becomes quite different after cycling, where some severe intergranular cracks are observed. It implies that the high reactivity of Ni3+ would lead to structural instability, causing the intergranular cracks associated with the surface structural degradation. On the other hand, this phase evolution and structural deformation are almost immunized by the Zr@SC-N83 electrodes. In detail, the intergranular cracks for the Zr@SC-N83 electrode have practically disappeared on both the internal and external surfaces of the particles (Fig. 4a–c). It is demonstrated that the structural integrity of the micron-sized particle for the Zr@SC-N83 electrode is well maintained, indicating that Zr can serve as a substitution of transition metal and inhibit the undesired crack formation and surface deterioration. Moreover, the SEM–EDS mapping signals of the C, O, F and P elements of the Zr@SC-N83 electrode are slightly weaker than that on the surface of the bare SC-N83 electrode, displaying the relatively low content of the electrolyte decomposition after Zr-doping. Meanwhile, the presence of the Zr with a larger ionic radius in the SC-N83 electrode kinetically facilitates the Li+ diffusion, thus improving the excellent cell-level electrochemical performance, even exceed to the reported SC-NCM at high-voltage in literature (Table S4).

Fig. 4
figure 4

SEM images and EDS mapping images of a–d bare SC-N83 and e–h Zr@SC-N83 cathodes after 100 cycles at 2.75–4.6 V and 1.0C

Full size image

3.3 Morphologies and composition after long-term cycling

To further demonstrate the inhibiting effect of the electrolyte decomposition on the Zr@SC-N83 cathode surface, the cycled cells with the bare SC-N83 and Zr@SC-N83 cathodes were disassembled in the glove box, and the washed cathodes were tested by XPS measurement (Fig. 5). The cathode/electrolyte interfacial (CEI) chemistry on both samples includes the carbon, oxygen, fluorine, and phosphorous species, where analysis of individual spectra reveals the slight differences in concentrations. In Fig. 5a, C 1s data show the evidence of four distinct carbon-containing compounds with binding energies at 284.2, 286.5, 287.7 and 290.5 eV, respectively, corresponding to C–C, C–O, C=O and OCO2 species from the decomposition of carbonate solvents (such as EC, EMC and DEC) [40]. The concentration of carbon-containing species reduces after Zr-doping, indicating that the decomposition ratio of the carbonate solvent is small, which is well consistent with the O 1s spectra (Fig. 5b). F 1s and P 2p data for these two cathodes are mainly dominated by LiF (684.8 eV) and LiPOxFy (133.2 eV) compounds [41, 42]. Moreover, the lower concentration of these two species in F 1s and P 2p spectra for the Zr@SC-N83 cathode is observed in Fig. 5c, d, demonstrating the lower decomposition concentration of LiPF6. It indicates that the distribution of trace Zr on the surface and bulk of the SC-N83 cathode can reduce the electrolyte decomposition deposits on the cathode surface, forming a thin and low impedance (Fig. 3f) of CEI film.

Fig. 5
figure 5

XPS spectra of a C 1s, b O 1s, c F 1s and d P 2p for cycled bare SC-N83 and Zr@SC-N83 cathodes after 100 cycles at 2.75–4.6 V and 1.0C

Full size image

TEM analysis further revealed the suppression of structural degradation of the Zr@SC-N83 electrode (Fig. 6a–g). As shown in Fig. 6a, there are severe cracks in the surface of the bare SC-N83 particle (Fig. 6a). Meanwhile, the structural integrity is observed in TEM images of the Zr@SC-N83 particle after long-term cycling at 4.6 V (Fig. 6d). This phenomenon is also revealed by SEM images of the cycled electrodes in Fig. 4, which further confirms the better mechanical stability after Zr-doping. HRTEM images display the surface of the bare SC-N83 (Fig. 6b) and Zr@SC-N83 (Fig. 6e) particles after 100 cycles. The surface region of the bare SC-N83 particle (yellow, cubic region) changes into a NiO phase with a thickness of more than 21 nm (Fig. 6b). The transition-metal ions have migrated to the lithium layer. The thick rock-salt phases of the cycled SC-N83 electrode are also confirmed by fast Fourier transformation (FFT) in Fig. 6c. By contrast, the slight surface degradation (NiO layer is about ~ 7 nm) of the cycled Zr@SC-N83 cathode is confirmed by HRTEM and FFT images (Fig. 6e–g), further indicating that the Zr-doping method can effectively inhibit the reversible phase transition.

Fig. 6
figure 6

TEM images of a, b bare SC-N83 and d, e Zr@SC-N83 cathodes after 100 cycles; corresponding FFTs of c Region 1 in bare SC-N83 and; f, g Regions 2, 3 in Zr@SC-N83 cathodes; h operation mechanism of bare SC-N83 and Zr@SC-N83 samples at high-voltage of 4.6 V

Full size image

The operation mechanism of the bare SC-N83 and Zr@SC-N83 cathodes at a high-voltage of 4.6 V is elaborated in Fig. 6h. For the SC-N83 electrode, Zr doping can not only enhance the Li+ transportation, but also enhance the structural stability, thus improving the electrochemical performance of the battery. However, the bare SC-N83 electrode underwent serious microstrains during the long-term cycling under high-voltage operation, leading to the formation of the micro-cracks and the surface structural degradation. More importantly, the electrolyte decomposition is significantly increased at such high voltages (≥ 4.5 V), significantly eroding along the surface of these cracks. Therefore, the robustness of the SC-N83 material is greatly boosted by adopting the Zr-doping strategy, which could effectively resist the corrosion of electrolyte and the considerable volume change at deep charging-discharging process.

4 Conclusion

In summary, we fabricate in situ doping strategy for single-crystal Ni-rich cathode by the combination of co-precipitation and calcination method. The cross-sectional SEM liner scanning and mapping and etch depth XPS reveal that Zr has homogeneous distribution both in surface and bulk of single-crystal LiNi0.83Co0.11Mn0.06O2. The introduction of trace Zr element could effectively reduce electrochemical polarization. More importantly, the Zr@SC-N83 electrode can maintain the structural integrity after long-term cycling even under a high range of 2.75–4.6 V. The in situ doping strategy significantly improves electrochemcial performance operating under high-voltage testing condition, which also can provide a guideline for achieving high structural stability of single-crystal Ni-rich cathodes.