Introduction

Rechargeable lithium-ion batteries (LIBs) have been applied widely in portable electronics such as cell phones, lap-tops, and digital cameras, due to their high output voltage, large energy density, and good cycling life [1,2,3]. Currently, with the ever-increasing demands for electric vehicles (EVs) and hybrid electric vehicles (HEVs), LIBs with larger energy and power density are highly desirable [4,5,6]. Among various cathode materials, LiNiO2, which is a nickel-rich layered material, has been considered as one of the most promising candidates, because of its higher capacity (200 mA h g−1) and lower cost compared with the currently used LiCoO2 [7,8,9]. However, LiNiO2 is difficult to be synthesized because of the chemical instability of Ni(III) and it suffers from serious capacity fade when it is cycled at a high cutoff voltage (> 4.2 V) [10, 11]. Furthermore, the redox reaction between the delithiated LiNiO2 and electrolyte can slowly cause the decomposition of the liquid electrolyte to result in the evolution of gases, which leads to the risk of cell explosion [12, 13]. The commercial exploitation of the material is thus limited.

It is now generally recognized that partial substitution of Ni in LiNiO2 with Co and doping with an electrochemically inert metal cation (Al, Mg, Ti, Mn, Ca, etc.) are successful methods to improve its electrochemical properties [14,15,16,17,18]. This is attributed to the suppression of phase transitions or lattice changes during cycling. Specially, Zn with a similar atomic size to the widely used Al dopant is non-toxic and low cost, and becomes a promising candidate for doping lithium nickel cobalt oxides. For instance, Fey et al. have reported Zn-doped LiZnyNi0.8-yCo0.2O2 (0.0000 ≤ y ≤ 0.0100) compositions were synthesized by a conventional solid-state method [19]. The obtained LiZn0.0025Ni0.7975Co0.2O2 exhibits remarkably enhanced electrochemical performance than that of the un-doped material, which is due to the structural stability derived from incorporating the size-invariant Zn(II) ions. However, synthesis of Zn-doped LiNi0.8Co0.2O2 via the conventional solid-state method requires high calcination temperature and prolonged calcination time, leading to large size and irregular particles which limits the improvement of its electrochemical performance. In this work, we explore a more facile sol-gel method to replace the conventional one.

Besides, surface coating is considered to be an another effective approach to enhance the electrochemistry performance of LiNi0.8Co0.2O2. Due to the presence of coating layer, the redox between electrode and electrolyte at delithiated state is depressed. Among those inaction metal oxides, Al2O3-coated LiNi0.8Co0.2O2 cathode show preferably improved cycling stability as the Al2O3 coating layer can protect the LiNi0.8Co0.2O2 particles from reacting with the electrolyte [20,21,22,23,24].

Herein, we report a submicron LiNi0.8Co0.15Zn0.05O2 cathode material, which was prepared by a facile sol-gel method instead of the conventional solid-state method. This solution-based synthesis offers molecular level mixing of the starting materials (Li, Ni, Co, Zn) and thereby gives rise to a high degree of homogeneity with minimum particle size and high surface area. In an effort to further enhance the structural stability of the LiNi0.8Co0.15Zn0.05O2 cathode material, the Al2O3 coating layer is introduced to prevent the contact between the LiNi0.8Co0.15Zn0.05O2 and electrolyte, and thus avert Ni atom dissolving in electrolyte, which leading to the forming of rhombohedral LiNi0.8Co0.2O2 collapse. By employing above the facile sol-gel method and the coating strategy, the Al2O3-coated LiNi0.8Co0.2O2 cathode gives a superior electrochemistry performance in lithium-ion batteries.

Experimental

Materials preparation

Preparation of LiNi0.8Co0.15Zn0.05O2: the samples of LiNi0.8Co0.15Zn0.05O2 were prepared by the sol-gel method using citric acid as a chelating agent. Stoichiometric amount of reactants Zn(NO3)2·6H2O (AR, 99%), Ni(CH3COO)2·4H2O (AR, 99%), Co(CH3COO)2·4H2O (AR, 99%), and CH3COOLi·2H2O were dissolved in deionized water to give a solution with mild stirring. An aqueous solution of citric acid was added at 1:1 molar ratios with the total transition metal ions. The pH of the mixed solution was maintained at 6.7 by adding ammonium hydroxide solution. Thereafter, the mixed solution was constantly shocked at about 80 °C for 4 h. Then, the gel was dried in oven at 120 °C for 12 h, forming the amorphous powders. The resulted amorphous powder was disposed at 500 °C for 5 h in air to remove the organic contents. Then, the as-prepared precursor was ground to fine powders and calcined at 700 °C for 10 h in oxygen to obtain the final product. The disposed and calcined temperatures are determined by the TGA-DSC curve (Supplementary material Fig. S1).

Preparation of Al2O3 coated LiNi0.8Co0.15Zn0.05O2: a typical experiment is as following: 100 mg LiNi0.8Co0.15Zn0.05O2 was dispersed in N,N-Dimethylformamide (DMF) and dispersed by ultrasonic for 60 min. After stirred for 3 h, 15 mg aluminum isopropoxide (AR, 99%) was added in the mixed liquid. Finally, the mixture was dried in oven at 120 °C for 12 h, and calcined at 400 °C in oxygen for 6 h to form Al2O3-coated LiNi0.8Co0.15Zn0.05O2 cathode material (Al2O3-coated LiNi0.8Co0.15Zn0.05O2). The decomposition temperature of aluminum isopropoxide is at about 300 °C (Supplementary material Fig. S2).

Materials characterization

TGA and DSC measurements were performed in oxygen stream from 25 to 1000 °C with a heating rate of 10 °C min−1. Powder X-ray diffraction (XRD) was performed with Cu Kα1 (45 kV, 40 mA, 10° < 2θ < 90°) to identify the crystalline phase of the materials. The X-Ray Fluorescence (XRF, EAGLE III, EDAX Inc.) was used to detect the elements of the materials. The valence states analysis of samples was performed with X-ray electron spectrometer (XPS, AXIS-ULTRA DLD-600W). The particle morphologies of the samples were examined with a scanning electron microscopy (SEM, NOVA 450, FEI) and transmission electron microscopy (TEM, G2 FEI).

Electrochemical test

For the preparation of cathode sheets, a slurry was formed by mixing the active material (80 wt%), acetylene black (10 wt%), and binder (10 wt% polyvinylidene fluoride, PVDF, dissolved in N-methyl-2-pyrrolidone, NMP). Then the slurry was coated onto an aluminum current collector. The electrodes were dried under vacuum at 90 °C for 12 h and then punched and weighed. The mass loading of the electrode is about 2.98 mg/cm2 and the area of electrode is 0.50 cm2. The batteries were assembled in a glove box under a high purity argon atmosphere. A celgard 2300 (polypropylene) was used as the separator and a lithium foil was used as anode. One molar of LiPF6 dissolved in a mixture of ethylene carbonate (EC) and dimethyl carbonate (DMC) (1:1 by volume) was used as the electrolyte. Cyclic voltammetry (CV) was measured with electrochemical workstation (CHI 760D) at a scan rate of 0.1 mV s−1 between 2.5 and 4.5 V. Electrochemical impedance spectroscopic (EIS) is studied in the frequency range from 10 mHz to 100 kHz (Princeton 4000) with a cell, which is discharged to 3.0 V after 10, 50, and 100 cycles, respectively, at room temperature. Charge-discharge performance of the cell was characterized at different current rate between 3.0 and 4.25 V (Land CT 2001).

Results and discussion

Structure and morphology characterizations

A typical experiment is presented in Fig. 1 and the crystallographic structures of the products were examined by XRD (Fig. 2a). The crystal structure of the as-prepared samples (bare and coated LiNi0.8Co0.15Zn0.05O2) corresponds to a layered structure of α-NaFeO2-type (space group, R-3m) and all the diffraction peaks can be assigned to Lithium Nickel Cobalt Oxide (JCPDS card No. 80-1917). The doublets of (006)-(012) and (018)-(110) could be well separated that indicate an ordered of Li and Ni/Co in a layered structure [16, 19]. At the same time, the XRD patterns do not detect impure peaks indexed to any other phases (ZnO, CoO, and NiO). Although the phase of Al2O3 is not detected due to the low content, the X-ray fluorescence (XRF) detects the Al element (Supplementary material Fig. S3a).

Fig. 1
figure 1

Schematic illustration of preparation of Al2O3-coated LiNi0.8Co0.15Zn0.05O2 by the sol-gel method using citric acid as the chelating agent

Full size image
Fig. 2
figure 2

a XRD patterns of the as-prepared samples. b–d XPS spectra of Ni(III) 2p, Co(III) 2p, and Zn(II) 2p, respectively

Full size image

The XPS spectra (Fig. 2b–d) also illuminate the structure of Al2O3-coated LiNi0.8Co0.15Zn0.05O2. The fitting XPS peak data (Fig. 2b) of Ni element could match well with the standard binding energy values of Ni(III), indicating that no other impurity peak assigned to Ni(II). Although the satellite peaks of Co 2p1/2 and 2p3/2 are unidentified as the influence of Ni(III), the XPS peaks of 2p1/2 and 2p3/2 could still prove the valence state of Co(III) existing (Fig. 2c).

Meanwhile, the XPS spectra of Zn are fitted (Fig. 2d). Two peaks (1045.4 and 1048.08 eV, the standard of 2p1/2 is 1044.7 eV for Zn element) can be assigned to the orbit of 2p1/2 and another two peaks (1022.24 and 1024.7 eV, the standard of 2p3/2 is 1022.1 eV for Zn element) can be assigned to the orbit of 2p3/2, which manifest the state of Zn element is similar to Zn2O3. These results show that the elements of Ni and Co are completely oxidized and exist with highest oxidation state and no phase of pure ZnO exists. All the facts accord well with the XRD analysis (Fig. 2a).

The morphological and structural features of Al2O3-coated LiNi0.8Co0.15Zn0.05O2 were characterized by field emission scanning electron microscope (FSEM) and TEM. As shown in Fig. 3a, the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 material consists of numerous submicron particles. The sizes of the submicron particles are in the range of 200–400 nm, which leads to a higher surface area than the bulk particles (Supplementary material Fig. S4). The detailed surface structure of Al2O3-coated LiNi0.8Co0.15Zn0.05O2 particle is clarified by TEM. The Al2O3-coated LiNi0.8Co0.15Zn0.05O2 particle surface is obvious encapsulation and the thickness of Al2O3 protective layer is about 15 nm (Fig. 3b), which indicates the Al2O3 phase is preferable coated on the surface of LiNi0.8Co0.15Zn0.05O2. Besides, HR-TEM investigation further confirms the rhombohedra texture of Al2O3-coated LiNi0.8Co0.15Zn0.05O2 particles (Fig. 3c). The lattice fringes are clearly visible, and the calculated d-spacing of 0.47 nm corresponds well to the (003) lattice planes of the LiNi0.8Co0.15Zn0.05O2 phase. Additionally, the composition of sample was checked by EDS mapping (Fig. 3d) and S-TEM (Supplementary material Fig. S5). The elements of oxygen, nickel, aluminum, cobalt, and zinc elements are clearly distinguished, which can further demonstrate the coating effect.

Fig. 3
figure 3

a High-magnification FESEM images of Al2O3-coated LiNi0.8Co0.15Zn0.05O2 submicron particles. b TEM image of Al2O3-coated LiNi0.8Co0.15Zn0.05O2. c Lattice plane spacing of Al2O3-coated LiNi0.8Co0.15Zn0.05O2. d EDS mapping data of Al2O3-coated LiNi0.8Co0.15Zn0.05O2

Full size image

These results imply that the as-prepared samples are LiNi0.8Co0.15Zn0.05O2 phase, which form a solid solution with Zn atoms and the Al2O3 layer is successfully coated on the surface of LiNi0.8Co0.15Zn0.05O2. The Al2O3 protective layer could effectively prevent the cathode material from directly contacting electrolyte and avert the Ni atom dissolving in electrolyte which leading the phase of rhombohedra LiNi0.8Co0.2O2 collapse.

Electrochemical performance

Motivated by the advantages of coated layer and submicron structures in LIBs, electrochemical measurements were carried out to evaluate the performance of the as-prepared Al2O3-coated LiNi0.8Co0.15Zn0.05O2 cathode material. CV is first employed to study reaction mechanism during electrochemical cycling in half-cell and the data is given in Fig. 4a. The first CV curve is different from the second cycle for anode scan. The oxidation peaks in the initial cycle are at 3.90, 4.03, and 4.23 V, corresponding to the reduction peaks at 3.65, 3.94, and 4.15 V, respectively. The initial structural phase change induced a relatively low columbic efficiency during the first cycle, and a solid electrolyte interface (SEI) film formed on cathode surface during the initial electrochemical oxidation process [25,26,27]. The oxidation peak in the initial cycle was at 3.90 V and moved to 3.79 V in the second curve, indicating that the formation of the SEI film leads to a loss of reversible capacity in the first charge-discharge cycle. In the cathode scan, three reduction peaks are observed, corresponding to the nickel-rich material (LiNiO2) of the three phase transition processes. These include the hexagonal phase H1-monoclinic phase M-hexagonal, the monoclinic phase M-hexagonal phase H2, and the hexagonal phase H2-hexagonal H3 phase transitions [28].

Fig. 4
figure 4

Electrochemical properties of the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 as electrodes in LIBs: a CV traces of Al2O3-coated LiNi0.8Co0.15Zn0.05O2/Li cell. b Charge-discharge voltage profiles at a current rate of 0.5 C. c Discharge capacities versus cycle number at a current rate of 0.5 and 1 C. d Rate capability at various current rates between 0.5 and 32 C

Full size image

Figure 4b depicts the 1st, 2nd, and 100th charge-discharge profiles of the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrodes in the potential window of 3.0–4.25 V at a current rate of 0.5 C. The given specific capacity was calculated based on the total mass of the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 composite. It can be obviously seen that the first discharge curve exhibits a clear potential plateau at about 3.75 V. More importantly, this potential plateau has scarcely changed in the subsequent cycles. The initial charge-discharge capacities of the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrodes are 234.4 and 194.5 mA h g−1, respectively, with a Columbic efficiency of 82.9%. The relatively low Columbic efficiency in the first cycle can be ascribed to the decomposition of the electrolyte and formation of a SEI on the electrode surface. It is shown the second charge and discharge capacities are 200.5 and 193.5 mA h g−1, respectively, indicating the high reversibility for Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrode. The high retention of the charge-discharge curves after 100 cycles shows the enhanced stability of the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrode.

The cycling performance of the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrodes is shown in the Fig. 4c. Reversible capacities as high as 182 and 169.8 mA h g−1 could still be maintained at the current rate of 0.5 and 1 C after 100 cycles, respectively. In addition, the cycling performance of the bare LiNi0.8Co0.15Zn0.05O2 sample is tested, which delivers a capacity of 140 mA h g−1 after 100 cycles. This fact indicates the Al2O3 coating protective layer has effectively enhanced the cycling stability of the LiNi0.8Co0.15Zn0.05O2 cathode. The rate capability of the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrodes is also evaluated by applying various current rates from 0.5 to 32 C (Fig. 4d) after 10 pre-cycle. The Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrode exhibits high reversible capacities of 187, 163, 129, 106, 83, and 48 mA h g−1 at the current rate of 0.5, 1, 4, 8, 16, and 32 C, respectively. When the current rate returned to 0.5 C, the reversible capacity can be recovered to 185 mA h g−1, which is close to the original value, showing good tolerance for high current rate. For comparison, the rate capability of the bare LiNi0.8Co0.15Zn0.05O2 sample is also evaluated, which shows a poor rate capability. These results indicate that the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 cathode shows higher reversibility capacity and better cycling performance than that reported LiNi0.8Co0.15Zn0.05O2 (Supplementary material Table S1) [19, 29, 30].

To understand the Al2O3 coating effect in depth, EIS is carried out using the cells with the bare LiNi0.8Co0.15Zn0.05O2 and the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 cathodes, respectively. Figure 5a, b show Nyquist plots obtained from the bare LiNi0.8Co0.15Zn0.05O2 and the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrodes after 10th, 50th, and 100th cycle tests. An intercept at the Zreal axis in high frequency corresponds to the ohmic resistance (RΩ), which consists of the total resistance of the electrolyte, separator, and electrical contacts. The depressed semicircle in the high frequency range is related to the Li-ion migration resistance (Rf) through the SEI film formation on the electrode or another coating layer. Second semicircle in the middle frequency range indicates the charge transfer resistance (Rct). The slope line at low frequency reflects the Li-ions diffusion into the electrode material, which represents the Warburg impedance (Wo) [31, 32]. The equivalent circuit model for the EIS is shown in Fig. 5e. Due to the existence of the Al2O3 protective layer, the Rct of Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrode (115.1 Ω) is larger than the bare LiNi0.8Co0.15Zn0.05O2 electrode (102.3 Ω) in the 10th cycle. In spite of this, the Rct of bare LiNi0.8Co0.15Zn0.05O2 electrodes becomes greater because of the structural instability and the Rct reaches to 464.4 Ω in the 100th cycle. However, the Rct of Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrodes could be stable at about 170 Ω due to the existence of Al2O3 protective layer. The Al2O3 protective layer could also affect the Li-ion diffusion coefficient (D), which determines the rate performance of electrode material. The calculation formulas of D value are presented in depiction 1 (Supplementary material depiction 1) and is directly determined by Warburg factor (σ) [33]. The D value is negative correlation with the σ value. The values of σ are clearly shown in Fig. 5c, d and reveal two main facts: (a) the σ value of LiNi0.8Co0.15Zn0.05O2 electrode is 1093, which is obvious lower than that of Al2O3-coated LiNi0.8Co0.15Zn0.05O2 (1120) electrode at the 10th cycle, signifying the Al2O3 protective layer has a barrier for Li-ion diffusion of the first 10 cycles. (b) In the 100th cycle, the σ values of the LiNi0.8Co0.15Zn0.05O2 and Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrodes are 1237 and 426.7, respectively, which indicate the coated layer has few influence on the Li-ion diffusion coefficient when the cell experiences long-time cycling due to the enhanced structure. The main reasons are as follows: (a) the Al2O3 protective layer shows a blockade for Li-ions diffusion in the first 10 cycles; (b) the layered-structure collapse of the bare LiNi0.8Co0.15Zn0.05O2 electrode reduces Li-ion diffusion in the subsequent cycles. Meantime, the Li-ion diffusion coefficient of Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrodes shows an escalating trend, which can be ascribed to the coating layer avoids the electrode material to directly contact with the electrolyte and prevent the Ni atoms dissolving into the electrolyte in chare-discharge process. The Raman spectra are further to illustrating the coating effect as shown in Fig. S6 (Supplementary material). At short wavelength, the absorption is obviously due to the Al2O3 coating layer. As a result, the Al2O3-coated LiNi0.8Co0.15Zn0.05O2 electrodes could provide superior cycling and rate performance.

Fig. 5
figure 5

a, b Nyquist plots of LiNi0.8Co0.15Zn0.05O2 cathode/Li metal and Al2O3-coated LiNi0.8Co0.15Zn0.05O2 cathode/Li metal half-cells measured in the frequency region of 105–10−2 Hz after 10, 50, and 100 cycles. c, d The real part of the complex impedance versus ω−1/2 at open circuit voltage for LiNi0.8Co0.15Zn0.05O2 and Al2O3-coated LiNi0.8Co0.15Zn0.05O2 cathodes after 10, 50, and 100 cycles. e Equivalent circuit model for the simulation of the Nyquist plots

Full size image

Conclusions

LiNi0.8Co0.15Zn0.05O2 cathode material is prepared by a facile sol-gel method and Al2O3 coating layer is applied to enhance its electrochemical performance. The Al2O3-coated LiNi0.8Co0.15Zn0.05O2 material consists of submicron particles with the size of 200–400 nm. The Al2O3 coating layer could effectively improve the electrochemical performance of the LiNi0.8Co0.15Zn0.05O2 by protecting the LiNi0.8Co0.15Zn0.05O2 particles from reacting with the electrolyte. The Al2O3-coated LiNi0.8Co0.15Zn0.05O2 cathode material can deliver a capacity of 182 mA h g−1 after 100 cycles at a current rate of 0.5 C. Specially, EIS measure is used for understanding the coating effect of Al2O3 in depth. The Al2O3 protective layer could enhance the structural stability of LiNi0.8Co0.15Zn0.05O2 and increase the Li-ion diffusion coefficient after a certain cycles. The synthesis strategy demonstrated herein is simple and versatile for the fabrication of other metal oxides coated and metal-doped cathode materials.