Introduction

Nowadays, the development of lithium-ion batteries with high power density and high energy density has been accelerated due to the trend of miniaturization of portable electronic devices as well as the development of electric vehicles. At this time, layered LiMO2 (M = Mn, Co, Ni), LiNi0.3Mn0.3Co0.3O2, olivine-type LiFePO4, spinel-type LiMn2O4, lithium-rich material xLi2MnO3·(1 − x)LiMO2 (M = Co, Fe, Ni1/2Mn1/2), etc. are the main cathode materials for rechargeable lithium-ion batteries. Among the various choices of cathode materials, the spinel lithium manganese oxide (LiMn2O4) is considered to be one of the promising cathode materials for lithium-ion batteries due to its advantages such as the abundance of raw materials, cheap resources, and environment friendliness. So it has been extensively considered as an alternative positive electrode material [1–4]. Nevertheless, the capacity fading of LiMn2O4 is a deadly obstacle which limits its commercial use. The reasons for fast capacity fading have been investigated, and they were mainly ascribed to the following three factors: (І) the dissolution of manganese into an electrolyte during cycling [5, 6], (II) Jahn-Teller distortion [7], and (III) the decomposition of the electrolyte solution [8].

There are two kinds of strategies to overcome the aforementioned problems. One way is substitution of heterogeneous atoms into the host spinel LiMn2O4 structure to improve the structural stability. This strategy, however, reduces the initial discharge capacity. The other one is surface modification. Coating the cathode material with various particles or films including SiO2, ZnO, CeO2, LaF3, FePO4, etc. can effectively minimize the direct contact area between the LiMn2O4 electrode and electrolyte and further inhibit the Mn dissolution during the cycling process [9–13]. Researches showed that surface modification techniques and the types of coating materials can affect the characteristics of electrode materials [14, 15]. Nowadays, most surface modification approaches of spinel lithium manganese oxide reported in the literature are carried out with sol-gel method [10], chemical deposition method [16], precipitation method [17], polymeric process [18], melting impregnation method [19], atomic layer deposition (ALD) method [20], and so on. ZnO was regarded as a desirable surface-coating material due to its excellent chemical and thermal stability. Some studies of ZnO-coated LiFePO4, LiMn1.5Ni0.50O4, LiNi0.5Co0.2Mn0.3O2, and LiMn2O4 have been reported by using different coating techniques to improve their electrochemical performance [21–25]. Tu et al. [15] coated ZnO on LiMn2O4 particles via the melting impregnation method. The ZnO coating sample showed a capacity retention rate of 96.36 % at 0.5 C after 100 cycles. Liu et al. [10] reported ZnO-coated LiMn2O4 by the sol-gel method. The 2 wt% ZnO-coated LiMn2O4 delivered an average capacity loss of 0.19 % per cycle in 50 cycles under a current rate of 0.5 C. Recently, Zhao et al. [20] used the ALD method to deposit ultrathin and highly conformal ZnO coatings onto LiMn2O4 cathodes. They have obtained a better cycling performance compared to the pristine LiMn2O4. However, to the best of our knowledge, there are no reports on ZnO-coated LiMn2O4 cathode materials prepared by a combustion method so far.

In our previous paper, we have successfully synthesized a single phase of spinel LiMn2O4 powders using a combustion method [26]. However, the LiMn2O4 did not achieve the desired capacity retention. In this paper, ZnO was selected for coating over the surface of LiMn2O4 cathodes to further improve their capacity and cycling performance. The effects of coating on the structure, morphology, and electrochemical performance of cathode materials are investigated in detail.

Experimental

Sample preparation

Synthesis of LiMn2O4 powders

The synthesis method of LiMn2O4 was referred to the literature [26]. Stoichiometric amounts of lithium nitrate (LiNO3, AR, aladdin) and manganese acetate (Mn(CH3COO)2·4H2O, AR, aladdin) were dissolved in 9 mol L−1 nitric acid and then placed in a 300-mL crucible. The mixture was calcined at 600 °C for 3 h in a muffle; then, the product was taken out and cooled down to room temperature naturally to get LiMn2O4 powders.

Synthesis of ZnO and ZnO-coated LiMn2O4 powders

ZnO-coated LiMn2O4 cathode materials were synthesized using glucose as fuel by a combustion method. Firstly, zinc acetate (Zn(CH3COO)2·2H2O) (AR, Tianjin, China Chemical Co., Ltd.) was dissolved in a water medium, and glucose (10 wt%) was added into the solution followed by continuous stirring for 15 min. Subsequently, previously synthesized LiMn2O4 powders were dispersed into the mixture. The mixture was stirred for 1 h to form a homogeneously dispersed suspension. The precursor was dried and calcined at 500 °C for 3 h to obtain the ZnO-coated LiMn2O4. The expected amounts of the ZnO coating are 1, 2, and 5 wt% of the total LiMn2O4 powder weight, respectively.

In order to investigate the crystal structure of ZnO, the powder of ZnO was prepared via thermal decomposition of zinc acetate. Firstly, zinc acetate was dispersed in the water medium, and then glucose (10 wt%) was added into the solution. The mixture was stirred for 1 h to obtain a homogeneously dispersed suspension. The precursor was dried and calcined at 500 °C for 3 h to obtain the ZnO powders.

Characterization of ZnO-coated LiMn2O4 cathode materials

The phase identification of the samples is performed on a diffractometer (X-ray diffraction (XRD), D/Max-TTRIII, Japan) using Cu Kα radiation; the scanning range of diffraction angle (2-theta) is 10°–70°. The scanning rate was 4° min−1 and the step width was 0.02° with 30 mA operation current and 40 kV voltage. The lattice parameters of the samples are calculated by using MDI Jade software. The morphology of the samples was observed by a scanning electron microscope (SEM, QUANTA 200, America FEI) equipped with an energy-dispersive spectrometer (EDS) and a transmission electron microscope (TEM, JEM-2100, JEOL). The value of d-spacing is measured by DigitalMicrograph software. X-ray photoelectron spectroscopy (XPS, PHI 5000 Versaprobe II, ULVAC-PHI, Japan) using monochromatic Al Kα (1486.6 eV) radiation was used to analyze chemical valence states of the samples, and the C 1s XPS binding energy at 284.6 eV was used as a calibration of the spectra.

Electrochemical measurements

Electrochemical properties of as-prepared powders were measured with CR-2025 coin-type cells. The electrodes were fabricated by mixing the active materials of the powder, super P carbon black as a conductive, and polyvinylidene fluoride (PVDF) as a binder in a weight ratio of 8:1:1 in N-methylpyrrolidinone solvent. Then, the mixture was spread onto an aluminum current collector and vacuum-dried at 120 °C for 10 h. The cells consisted of the as-fabricated electrodes as the positive electrode, metallic lithium as the counter electrode, a polypropylene film (Celgard 2320) as the separator, and the electrolyte 1 mol/L LiPF6 in EC and DEC (1:1 v/v). Cells were fabricated in an argon-filled glove box. The charge-discharge tests, cyclic performance, and rate performance tests were examined using a LAND battery program control cell tester (Wuhan Land Electronics Co., Ltd, China) between 3.0 and 4.5 V (vs. Li/Li+) at room temperature. Cyclic voltammetry (CV) was performed on a CHI604E electrochemical workstation (Shanghai Chenhua Instruments Co., Ltd, China) using a standard three-compartment cell in the potential range of 3.6–4.5 V and at a scan rate of 0.05 mV s−1 (vs. Li/Li+). Electrochemical impedance spectroscopy (EIS) spectra of cells were measured using the same electrochemical workstation in a frequency range of 0.1 Hz to 100 KHz with the applied AC signal amplitude of 10 mV.

Results and discussion

Characterization of the coating structures and morphology

Figure 1a shows the XRD patterns for un-coated LiMn2O4 and ZnO-coated (1, 2, and 5 wt%) LiMn2O4 powders, respectively. As expected in Fig. 1a, all the diffraction peaks could be indexed in a spinel structure of LiMn2O4 (JCPDS card no. 35-0782) with a space group of Fd-3m, corresponding to the eight crystal planes of (111), (311), (222), (400), (331), (511), (440), and (531), where Li ions occupy tetrahedral sites (8a), Mn atoms octahedral sites (16d), and O2− ions octahedral sites (32e) [27]. No impurity phase can be found in Fig. 1a(a–c). While the ZnO peak was detected in Fig. 1(d), no diffraction peaks from ZnO are observed in Fig. 1a(a–c), suggesting that the content of ZnO may be too low to be detected. Compared with the un-coated LiMn2O4, 1 wt% ZnO-coated LiMn2O4, and 2 wt% ZnO-coated LiMn2O4, the 5 wt% ZnO-coated LiMn2O4 sample obviously displays a broadening of FWHMs, indicating that the density of stacking faults increased after ZnO was surface-modified to a certain extent. This result was similar to an earlier report [19]. The lattice parameters are summarized in Table 1. The lattice constants of the coated materials compared with that of the un-coated LiMn2O4 show no obvious change. Arumugam et al. [28] showed that the lattice parameter of samples increased with an increase in the amount of the dopant Zn. This indicated that ZnO coating does not change the structure of LiMn2O4 and ZnO is just coated on the surface of LiMn2O4 particles.

Fig. 1
figure 1

XRD patterns of a the pristine LiMn2O4 and various wt% ZnO-coated LiMn2O4: the (a) pristine, (b) 1 wt%, (c) 2 wt%, and (d) 5 wt% ZnO-coated LiMn2O4 cathode materials. b XRD patterns of ZnO

Full size image
Table 1 Lattice parameters of the pristine LiMn2O4 and various wt% ZnO-coated LiMn2O4 cathode materials
Full size table

Figure 1b shows the XRD patterns of ZnO; all the diffraction peaks could be indexed with a wurtzite hexagonal structure of the bulk ZnO (JCPDS card no. 36-1451) confirmed by the peaks at 31.8°, 34.5°, and 36.3° and correspond to the (1010), (0002), and (1011) planes of ZnO, respectively. And no impurity phases were observed in the spectrum, indicating that the pure ZnO powders are obtained via thermal decomposition of zinc acetate and the synthesized powder is single crystalline.

Figure 2a–d shows the SEM images of un-coated LiMn2O4 and ZnO-coated (1, 2, and 5 wt%) LiMn2O4 powders, respectively. The inset is the EDS profile of the 2 wt% ZnO-coated LiMn2O4 sample. All the materials display agglomeration and irregular-shaped particles. The particle size increased from around 0.2 to 0.4 μm as the ZnO coating content. It demonstrates that a small layer of ZnO coated on the surface of the LiMn2O4 enlarges the particle size of the pristine LiMn2O4. Zinc, manganese, and oxygen can be clearly observed in the EDS profile (Fig. 2e). The existence of Zn suggests that ZnO may deposit onto the particle surface of LiMn2O4.

Fig. 2
figure 2

The SEM images of pristine LiMn2O4 and various wt% ZnO-coated LiMn2O4: the a pristine, b 1 wt%, c 2 wt%, and d 5 wt% ZnO-coated LiMn2O4 cathode materials and the e EDS spectrum of the 2 wt% ZnO-coated LiMn2O4 sample and the element content of the powder

Full size image

TEM is used to further research the surface morphology of the bare LiMn2O4 and ZnO-coated LiMn2O4. Figure 3a shows the TEM images of the un-coated LiMn2O4 materials with a measured d-spacing value of 0.48 nm, corresponding to the (111) crystalline plane and possessing a smooth surface. Figure 3b shows the image of the 5 wt% ZnO-coated LiMn2O4 cathode material. This image indicates that a physically distinct phase is attached to the LiMn2O4 surface, which is a distinctly different surface structure compared with the pristine LiMn2O4 sample, and the particles have a lattice spacing close to 0.51 nm, corresponding to the (0001) crystal planes of the wurtzite ZnO [29, 30]. A uniform ZnO coating formed over the un-coated LiMn2O4 particles with a coating layer with about 6.7 nm thickness, indicating that the oxide ZnO coating layer is crystalline, which is consistent with the XRD result.

Fig. 3
figure 3

The TEM images of a pristine LiMn2O4 and b 5 wt% ZnO-coated LiMn2O4

Full size image

XPS measurement was performed to examine surface compositions of the ZnO-coated LiMn2O4 samples. As shown in Fig. 4a, the signals of Li, Mn, O, and Zn can be observed in the spectrum. In the Zn 2p spectrum (Fig. 4b), Zn 2p3/2 and 2p1/2 peaks at about 1021.2 and 1044.3 eV with a binding energy difference of 23.1 eV can be assigned to Zn(II), which is consistent with the reported papers on ZnO [31], which indicates a thin layer of ZnO forms on the surface of the LiMn2O4 particles.

Fig. 4
figure 4

a XPS spectra of 2 wt% ZnO-coated LiMn2O4 and b XPS spectra of Zn 2p

Full size image

Improvement in the cycling and rate performance of the spinel cathode

Figure 5 shows the (a) initial charge-discharge and (b) cycling performance profiles of the un-coated and various wt% ZnO-coated LiMn2O4 electrodes at 0.5 C (1 C = 148 mA g−1) between 3.0 and 4.5 V. As can be seen from the discharge profiles, the samples with two plateaus at around 4.05/4.20 V in the charge curve and at 3.9/4.1 V in the discharge curve indicate the two-step insertion/extraction reaction for the spinel LiMn2O4, which also corresponds to redox peaks in the CV curves (Fig. 6) [32]. The initial discharge capacities of the 1, 2, and 5 wt% ZnO-coated LiMn2O4 were 116.1, 112.8, and 102.9 mAh g−1, respectively. Whereas that of the bare LiMn2O4 fades from 120.4 to 69.2 mAh g−1 after 500 cycles, 57.5 % of the original capacity is only reserved. However, the capacity retentions are 62.8 % for 1 wt%, 84.1 % for 2 wt%, and 77.1 % for 5 wt% ZnO after 500 cycles, respectively. Based on the above analysis, the 2 wt% ZnO-coated LiMn2O4 sample is the optimum composition to enhance the stability and cycling performances of LiMn2O4; as ZnO increases beyond the optimal coating content, the lower capacity retention and discharge capacity of composite electrodes may be ascribed to the blocking of lithium ion transport through a slightly thicker coating layer (herein ZnO), which results in more polarization rather than an electrochemical enhancement from the increased electronic conductivity [33]. It means that the fading in capacity suppression is due to modification with ZnO, which may minimize the surface area to suppress the side reactions between electrolyte solutions and cathode materials and to further inhibit the Mn dissolution by the generation of acids such as HF during the cycling process [34].

Fig. 5
figure 5

a Initial charge-discharge curves of pristine LiMn2O4 and various wt% ZnO-coated LiMn2O4. b Cycling performance curves of pristine LiMn2O4 and various wt% ZnO-coated LiMn2O4, after 500 cycles at 0.5 C. c Rate performance of pristine LiMn2O4 and 2 wt% ZnO-coated LiMn2O4 from 0.5 to 5 C

Full size image
Fig. 6
figure 6

Cyclic voltammogram curves of the 100th, 200th, 300th, and 500th cycles for the a pristine LiMn2O4 and b 2 wt% ZnO-coated LiMn2O4 samples are tested at a scan rate of 0.05 mV s−1 between 3.60 and 4.50 V

Full size image

Figure 5c displays the rate performance of un-coated LiMn2O4 and 2 wt% ZnO-coated LiMn2O4 from 0.5 to 5 C in the voltage range of 3.0–4.5 V. It can be observed that the discharge-specific capacity all decreases with the increasing rate due to electrochemistry polarization. The sample of LiMn2O4/2 wt% ZnO possessed the highest discharge capacity of the electrodes at 5 C of 79.1 mAh g−1, while the pristine LiMn2O4 was only 43.4 mAh g−1 at the same condition. The cycling performance of 2 wt% ZnO-coated LiMn2O4 is better than the ZnO-coated LiMn2O4 samples prepared by ALD, because the ZnO-coated LiMn2O4 samples prepared by atomic layer deposition delivers a capacity retention of 76 % after 100 cycles in 1 C at 25 °C and delivers a relatively low initial capacity (65.1 mAh g−1) in 1 C [14]. However, the 2 wt% ZnO-coated LiMn2O4 sample prepared by this combustion method delivers 79.1 mAh g−1 at a big current rate (5 C). Therefore, it is proposed that 2.0 wt% ZnO coating is an outstandingly effective way for improving the rate capability of LiMn2O4.

Cyclic voltammetric and electrochemical impedance spectroscopic behaviors of un-coated LiMn2O4 and 2 wt% ZnO-coated LiMn2O4

To further investigate the effect of ZnO coating on the mechanism of spinel LiMn2O4, cyclic voltammetry tests are conducted on the pristine LiMn2O4 and 2.0 wt% ZnO-coated sample. Figure 6 shows the cyclic voltammogram curves of the 100th, 200th, 300th, and 500th cycles for the un-coated LiMn2O4 and 2 wt% ZnO-coated LiMn2O4 samples tested at a scan rate of 0.05 mV s−1 between 3.60 and 4.50 V. Two pairs of clear oxidation/reduction peaks can be observed from the CV curves associated with the typical two-step reversible intercalation/deintercalation of Li+ in spinel LiMn2O4 [35], which shows good agreement with the corresponding two potential plateaus of the initial discharge curves from Fig. 5a discussed above. The peak currents of the un-coated LiMn2O4 sample significantly decreases with the increasing charge-discharge cycling. However, the cyclic voltammogram curves present in Fig. 6b of the 2 wt% ZnO-coated LiMn2O4 sample shows a distinct difference. With increasing of the charge-discharge cycling, the corresponding peak currents only incur a minor decrease before the 300th cycle. These results indicated that ZnO coating on the surface of LiMn2O4 is an effective way to improve the electrochemistry performance, which corresponds to the good cycling performance of the ZnO-coated LiMn2O4 sample shown in Fig. 5.

To clarify the effect of the ZnO coating on the electrochemical properties of LiMn2O4, Fig. 7 shows the Nyquist plots of (a) the un-coated LiMn2O4 and (b) the 2 wt% ZnO-coated LiMn2O4 samples investigated at different cycles at room temperature. The EIS of pristine LiMn2O4 and 2 wt% ZnO-coated LiMn2O4 are fitted by the ZSimpWin analysis software, and the EIS are simulated using the equivalent circuit shown in the inset of Fig. 7c. The values of parameters are listed in Table 2. A small interruption at the (Z′re) axis in the high-frequency region is assigned to the solution resistance (R s), the semicircle in the middle-frequency region is assigned to the charge transfer resistance (R ct) at the interfacial capacitance between the electrolyte and electrodes [36], and the short line in the low-frequency region is caused by the diffusion of Li+ and is named Warburg impedance (W) [37]. C int is the reflection of intercalation capacitance [38]. The constant phase element (Q) represents the capacitance of the double layer. The R ct of both samples increases with charge-discharge cycling. The R ct of un-coated LiMn2O4 increases from 288.3 to 480.9 Ω, and the R ct of 2 wt% ZnO-coated LiMn2O4 only increases from 114.7 to 208.6 Ω. Obviously, surface modification by coating ZnO can retard the increase of charge transfer resistance during the cycling test. A smaller R ct is conducive for the electrochemical reaction of lithiation/delithiation, and the surface modification by coating ZnO guarantees a better cycling behavior than the pristine sample [39, 40], which agrees well with the electrochemical test results.

Fig. 7
figure 7

Nyquist plots of the a pristine LiMn2O4 and b 2 wt% ZnO-coated LiMn2O4 samples investigated at the 100th, the 200th, the 300th, and the 300th cycles at room temperature. c The corresponding equivalent circuit

Full size image
Table 2 The fitting values of EIS at different cycles
Full size table

XRD test after cycling

To further compare the crystal structure of 2 wt% ZnO-coated LiMn2O4 before cycling and after 500 charge-discharge cycles, XRD was performed on the un-cycled and cycled electrodes for 500 cycles at 0.5 C as shown in Fig. 8. A small amount of an impurity phase is detected in the XRD patterns (2-theta = 28.6°), because the working electrode includes 80 wt% active materials, 10 wt% super P carbon black, and 10 wt% PVDF. It can be seen that the powder also maintains a spinel LiMn2O4 structure after 500 charge-discharge cycles, implying that ZnO coating can improve the structural stability of LiMn2O4 during cycling, retard the side reactions between the electrode and electrolyte, further minimize the Mn dissolution during the cycling process, and result in a better cyclability and rate capability.

Fig. 8
figure 8

XRD patterns of 2 wt% ZnO-coated LiMn2O4 a before cycling and b after 500 charge-discharge cycles at 0.5 C between 3.0 and 4.5 V at room temperature

Full size image

Conclusions

ZnO-coated LiMn2O4 was successfully synthesized by a combustion method using glucose as fuel. Based on the XRD, EDS, TEM, and XPS analysis results, we have proved ZnO formation on the surface of the LiMn2O4 particles. Compared to the pristine LiMn2O4, the 2 wt% ZnO-coated LiMn2O4 sample as a cathode for LIBs exhibited a higher capacity retention of 84.1 % after 500 cycles at 0.5 C and a better rate capacity, still delivering 79.1 mAh g−1 at 5 C. The improved performance of the surface-coated sample is because the ZnO coating on the surface of LiMn2O4 can effectively minimize electrochemical polarization and charge transfer resistance during charge-discharge cycling. The XRD results of the 2 wt% ZnO-coated LiMn2O4 material after 500 cycles revealed that ZnO coating improves the stability of the spinel structure, making lithium ions effectively diffuse in the electrode material. ZnO coating improves the electrochemical performances of LiMn2O4 comparing with the pristine sample in terms of cycling and rate capability.