Chenqiang Dua,
Man Yang*a,
Jie Liua,
Shuting Suna,
Zhiyuan Tang*a,
Deyang Qub and
Xinhe Zhangb
aDepartment of Applied Chemistry, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, China. E-mail: 369469876@qq.com; Fax: +86-769-83195372; Tel: +86-769-83017180
bMcNair Technology Company Limited, Dongguan City, Guangdong 523700, P. R. China
First published on 23rd June 2015
Lithium boron oxide glass (LBO-glass) coated LiNi0.5Mn1.5O4 cathode materials have been synthesized by a solution method to enhance the electrochemical performances. The structure and morphology of the as-prepared materials have been characterized by XRD, SEM, and TEM. The results indicate that the LiNi0.5Mn1.5O4 is coated with a layer of amorphous LBO-glass. The electrochemical properties are characterized by galvanostatic charge–discharge cycling, cyclic voltammetry, electrochemical impedance spectroscopy and a self-discharge test. Differential scanning calorimetry is carried out to confirm the improved safety by LBO-glass coating. The LBO coating can effectively enhance the electrochemical kinetics of the LiNi0.5Mn1.5O4 phase and improve the cycling performance. Among the as-prepared samples, the 1 wt% LBO-glass coated LiNi0.5Mn1.5O4 presents optimal electrochemical behaviors with a capacity retention of 91.4% after 100 cycles at 1 C and a discharge capacity of 105.8 mA h g−1 at 10 C. Besides, the electrochemical impedance spectroscopy analysis shows the 1 wt% LBO-glass coating reduces the electrochemical impedance and improves the ability to conduct Li+ in the cells to a great extent.
However, the poor electrochemical performances of LiNi0.5Mn1.5O4 such as low discharge capacity and depressed cyclability hinder its application to a large extent.3 One reason is the decomposition of electrolyte at high operating voltage,4,5 resulting in the formation of SEI film, which blocks the insertion/extraction of Li+ and hinders the charge transfer, hence the kinetics of the electrochemical processes. Besides, side reactions between the cathode material and the electrolyte restrict its electrochemical performance.6 Also, the minor Mn3+ ions in LiNi0.5Mn1.5O4, which result from high-temperature and long-time calcinations, further lead to Mn dissolution, structure distortion, and capacity fading upon cycling.7 To overcome these disadvantages, many efforts have been made, such as ion doping8–20 and surface coating.21–29 Cation and/or anion doping such as Zn2+,8 Mg2+,9,10 Al3+,11,12 Fe3+,13 Cr3+,11,14,15 Sm3+,16 Ru4+,17,18 Ti4+,18 Nb5+,19 W6+20 and F−
11 have been employed to improve the electrochemical behaviours of LiNi0.5Mn1.5O4. However, doping is not able to prevent the decomposition of the electrolyte and side reactions between cathode and electrolyte. In recent years, various oxides and phosphates have drawn researchers' attention because of facile synthesis methods and positive effects on the electrochemical properties of LiNi0.5Mn1.5O4, such as ZnO,21,22 ZrO2,23 ZrP2O7,23 Al2O3,6,22 ZnAl2O4,24 Bi2O3,6,22 SiO2,25 CuO,26 YBaCu3O7,27 BiOF,28 AlPO4
22 and FePO4.29 These previous literature has reported that the surface-modifications using inorganic materials are effective to enhance the electrochemical performances and the thermal stability of the high-voltage LiNi0.5Mn1.5O4 cathode system.
Nevertheless, these coating materials can form barriers to the movement of lithium ions due to low ion conductivity. According to Amatucci and co-workers' report,30 lithium boron oxide glass (LBO-glass) is particularly suitable for surface modification for the following reasons. First of all, molten LBO compositions exhibit good wetting properties with respect to ceramics. Not only the good wetting properties but also the relatively low viscosity of LBO-glass in the molten state allow easy processing and result in even coverage with the use of a minimal amount of material. Secondly, LBO compositions have already been investigated as solid lithium ion conductors which show good ability to conduct Li+.31,32 Furthermore, electrochemical studies have shown that LBO coating is helpful for the 4 V positive electrode materials used in lithium-ion batteries such as LiNi0.8Co0.2O2,33 LiNi1/3Co1/3Mn1/3O2,34 LiMn2O4,35,36 Li1+xMn2O4.37
In this paper, a surface modification method is conducted through coating LBO-glass on LiNi0.5Mn1.5O4 in order to enhance its electrochemical performances. The results of various structure and electrochemical analyses show that the LBO-glass plays a protecting role on the surface of LiNi0.5Mn1.5O4 cathode, thus improving the electrochemical behaviours of the cathode materials.
The LBO-glass coated LiNi0.5Mn1.5O4 was prepared using a solution method. LiOH·H2O (A.R., Guangdong Guanghua Sci-Tech Co., Ltd) and H3BO3 (A.R., Guangdong Guanghua Sci-Tech Co., Ltd) were mixed in ethanol (A.R., Guangdong Guanghua Sci-Tech Co., Ltd) with a molar ratio of 1:
2. Then bare LiNi0.5Mn1.5O4 was added into the solution, and the mixture was stirred and heated at 80 °C until the solvent was completely evaporated. The powders were grinded, pressed into pellets and heated at 500 °C for 10 h. The melting points of H3BO3 and LiOH·H2O were 186 °C and 445 °C, respectively. Therefore, during heating the liquid H3BO3 and LiOH·H2O could wet the surface of LiNi0.5Mn1.5O4 and reacted with each other, producing the LBO-glass. After the mixture cooled to room temperature, the solid Li2O–2B2O3 glass modified LiNi0.5Mn1.5O4 powders were obtained. The weight ratios of the Li2O–2B2O3 glass to the LiNi0.5Mn1.5O4 powders in the modified samples were 0.5%, 1%, 2% and 3%, which were denoted as 0.005LBO–LNMO, 0.01LBO–LNMO, 0.02LBO–LNMO and 0.03LBO–LNMO, respectively.
The thermal stability of the LNMO and 0.01LBO–LNMO was studied using differential scanning calorimetry (DSC-Perkin-Elmer Pyris 1 differential scanning calorimeter). After 2 charge–discharge cycles at 0.1 C rate, the cells were fully-charged and subsequently subjected to 1 h of trickle charge at constant voltage of 5.0 V to ensure the cathode material is completely delithiated. Then, the cells were introduced into a dry glove box and opened carefully and the cathode materials were recovered. The electrodes were slightly dried and then scratched to obtain the cathode material. 5–6 mg materials from each kind of electrodes were packed into small stainless steel DSC capsules and then hermetically sealed. The samples were scanned in the DSC equipment under nitrogen purging from room temperature to 400 °C at a heating rate of 10 °C min−1.
The charge–discharge tests were performed in the voltage range of 3.5–5.0 V (vs. Li+/Li) on an automatic batteries tester (Land CT 2001A, Wuhan, China). The cycling stability was measured at 1 C (1 C = 147 mA h g−1) for 100 cycles. And to test the rate performance of the samples, the cells were charged at 0.5 C and discharged at different rates. Cyclic voltammetry (CV) tests were conducted on an electrochemical workstation (CHI1040B, ChenHua, China) with a voltage range of 3.5–5.0 V (vs. Li+/Li) at a scanning rate of 0.1 mV s−1. The electrochemical impedance spectroscopy (EIS) analysis of LNMO and 0.01LBO–LNMO was carried out after the cells were charged and discharged at 1 C for 1 cycle and 30 cycles respectively, on an electrochemical workstation (Gamry PC14-750) in the frequency range of 100 kHz to 10 mHz. Self-discharge of LNMO and 0.01LBO–LNMO was investigated on an electrochemical workstation (CHI1040B, ChenHua, China). The cells made by LNMO and 0.01LBO–LNMO were charged and discharged at 0.1 C (1 C = 147 mA h g−1) in the voltage range of 3.5–5.0 V for 5 cycles, then charged to 5 V and left at open-circuit voltage for 10 days. All electrochemical measurements have been performed at room temperature.
LiNi0.5Mn1.5O4 → αLixNi1−xO + βLiNi0.5−yMn1.5+yO4 + γO2 | (1) |
Fig. 2 shows the SEM images of the pristine and LBO-glass coated LiNi0.5Mn1.5O4. It can be seen that the pristine LiNi0.5Mn1.5O4 possesses a distribution of secondary particles sizing between 2–8 μm, while the LBO-glass coated materials possess a part of smaller secondary particles and slighter particle agglomerations. Thus, the LBO-glass coating prevents the aggregation of nanoparticles, which benefits the improvement of the electrochemical properties.
![]() | ||
Fig. 2 SEM images of (a) LNMO, (b) 0.005LBO–LNMO, (c) 0.01LBO–LNMO, (d) 0.02LBO–LNMO, (e) 0.03LBO–LNMO. |
The HRTEM image of the 0.01LBO–LNMO is displayed in Fig. 3. It is observed that a thin film with a thickness in the range of 5–10 nm is coated on the surface of LiNi0.5Mn1.5O4 cathode material. In addition, the fast Fourier transform (FFT) images are shown in the insets of Fig. 3, which confirm the crystalline characteristics of the LiNi0.5Mn1.5O4 and the amorphous structure of the coated LBO-glass layer, in accordance with the XRD analysis. The lattice fringe with a d-spacing value of 0.4765 nm corresponds to the (111) plane of LiNi0.5Mn1.5O4. From the TEM image and the XRD patterns, it is concluded that the LBO-glass layer is amorphous rather than crystalline.33,36
![]() | ||
Fig. 4 Cycling performances for 100 cycles at 1 C (a) and rate performances (b) of the pristine and LBO-glass coated LiNi0.5Mn1.5O4. |
To characterize the rate capability of the samples, electrodes made from the as-prepared materials are charged to 5.0 V at 0.5 C and discharged to 3.5 V at different rates from 0.5 C to 10 C and then back to 0.5 C. It can be clearly seen that the 0.01LBO–LNMO possesses the best rate capability and cyclability. A discharge capacity of 105.8 mA h g−1 can be retained at a discharge rate of 10 C for 0.01LBO–LNMO, while the LNMO, 0.005LBO–LNMO, 0.02LBO–LNMO and 0.03LBO–LNMO deliver lower capacities of 98.7, 102.3, 95.6 and 85.4 mA h g−1, respectively. When the current density is reduced to 0.5 C, the discharge capacities could recover to their initial value, which indicates that rapid lithiation/delithiation and a large current density did not result in structural damage of the crystals. Moreover, both the stable structure and the LBO-glass with fast lithium ion conduction can reduce the polarization especially at high current density, which leads to an improvement of the rate capability. The worse discharge capacity of the 0.02LBO–LNMO and 0.03LBO–LNMO may be induced by the electrochemical inactivity of the excessive LBO-glass. It can be concluded that 1 wt% is a proper coating amount for the LNMO cathode material.
Typical charge–discharge profiles at 0.5 C for the 10th cycle and cyclic voltammogram (CV) curves for the pristine and LBO-coated LiNi0.5Mn1.5O4 at a scanning rate of 0.1 mV s−1 are shown in Fig. 5. As observed in Fig. 5a, two dominant plateaus around 4.7 V belonging to the charge curves are attributed to the redox reactions of Ni3+/Ni2+ and Ni4+/Ni3+.29 While for the discharge curves, only one distinct plateau is found around 4.7 V. As shown in Fig. 5b, two pairs of redox peaks (around 4.76/4.66 V and 4.81/4.67 V) are ascribed to the redox reactions of Ni3+/Ni2+ and Ni4+/Ni3+, which is in accordance with the above results. Besides, a weak peak at around 4.0 V indicates the Mn4+/Mn3+ redox reaction, which is in agreement with the magnified region in the CV curves. What's more, the 0.01LBO–LNMO presents a larger current density than other samples, suggesting a lower polarization of the material. On the basis of previous reports,31,32 the LBO-glass is not a good electronic conductor, but a very good Li+ ionic conductor. The thin coating layer may promote the Li+ transfer process through the interfaces between the LBO-glass and the LNMO materials. As a result, the electrochemical polarization is decreased.
![]() | ||
Fig. 5 Typical charge–discharge profiles (a) and cyclic voltammogram (CV) curves at a scanning rate of 0.1 mV s−1 (b) for the pristine and LBO-glass coated LiNi0.5Mn1.5O4. |
To further characterize the improved electrochemical behaviour of the 0.01LBO–LNMO, the EIS test results of the LNMO and 0.01LBO–LNMO are shown in Fig. 6. The resistances are evaluated using the equivalent circuit model shown in the inset of the Fig. 6a,23,44,45 and the fitted values are listed in Table 1. According to the literatures,23,45 Re represents the solution resistance; RSEI and CPE1 signify the diffusion resistance of Li+ through the solid-electrolyte interface (SEI) layer and the corresponding constant phase element (CPE); Rct and CPE2 correspond to the charge transfer resistance and the corresponding CPE, while RW (not calculated here) is related to the solid-state diffusion of Li+ in the active materials corresponding to the slope of the line at low frequency. From the fitted data, it can be concluded that the resistances of the electrolyte (Re) are similar for both the LNMO and the 0.01LBO–LNMO. However, after 1 cycle at 1 C, both the RSEI and Rct of the 0.01LBO–LNMO are much lower than those of the LNMO. Further, after the cells are charged and discharged for 30 cycles, RSEI and Rct show an increasement, which of the 0.01LBO–LNMO are still much smaller than that of that LNMO, indicating that the LBO-glass coating layer plays an important role in reducing the resistances. What's more, the EIS can also be used to calculate the lithium-ion diffusion coefficient (D) through employing the following equations:19,26
Zre = Re + Rct + σω−1/2 | (2) |
![]() | (3) |
![]() | ||
Fig. 6 Nyquist plots (a) and graph of Zre plotted against ω−1/2 (b) of the LNMO and 0.01LBO–LNMO after cycled at 1 C for 1 cycle and 30 cycles. |
Samples | Re/Ω | RSEI/Ω | Rct/Ω | σ/Ω cm2 s−1/2 | D/cm2 s−1 |
---|---|---|---|---|---|
LNMO@1st | 4.61 | 269.2 | 223.4 | 100.4 | 9.63 × 10−13 |
LNMO@30th | 4.01 | 295.3 | 333.1 | 190.7 | 5.07 × 10−13 |
0.01LBO–LNMO@1st | 5.48 | 60.4 | 81.1 | 65.6 | 1.47 × 10−12 |
0.01LBO–LNMO@30th | 4.43 | 183.5 | 170.3 | 75.6 | 1.28 × 10−12 |
Self-discharge profiles of LNMO and 0.01LBO–LNMO cells are illustrated in Fig. 7. After storage for 10 days, the voltage of the charged LNMO cell decreases from 4.96 V to 4.70 V, while that of the 0.01LBO–LNMO decreases from 4.96 V to 4.74 V. It is apparent that the charged 0.01LBO–LNMO suffers less self-discharge than LNMO. Thus, LBO coating is an effective method to suppress this self-discharge, which may be attributed to the fact that LBO layer is able to protect the active material form reacting with the electrolyte and prevent the self-discharge.
![]() | ||
Fig. 7 Self-discharge profiles of LNMO and 0.01LBO–LNMO cells after 5 cycle at 0.1 C in the potential range of 3.5–5.0 V. |
DSC spectra of over charged LNMO and 0.01LBO–LNMO are illustrated in Fig. 8. As shown in the Fig. 8, peaks around 100 °C may be attributed to the decomposition of the trace of electrolyte. When the temperature reaches around 200 °C, the cathodes undergo an exothermic reaction with large enthalpy and the reaction of LNMO gets completed at a much lower temperature than that of 0.01LBO–LNMO. Pristine LiNi0.5Mn1.5O4 cathode is found to be thermally unsafe, while the coated one gain more thermal stability, which may be attributed to the fact that LBO glass improves the safety of the cathode material.
![]() | ||
Fig. 8 DSC spectra of over charged LNMO and 0.01LBO–LNMO cathodes with traces of 1.0 mol L−1 LiPF6 in EC–EMC (3![]() ![]() |
This journal is © The Royal Society of Chemistry 2015 |