Improved Electrochemical Properties of LiMn2O4-Based Cathode Material Co-Modified by Mg-Doping and Octahedral Morphology

In this work, the spinel LiMn2O4 cathode material was prepared by high-temperature solid-phase method and further optimized by co-modification strategy based on the Mg-doping and octahedral morphology. The octahedral LiMn1.95Mg0.05O4 sample belongs to the spinel cubic structure with the space group of Fd3m, and no other impurities are presented in the XRD patterns. The octahedral LiMn1.95Mg0.05O4 particles show narrow size distribution with regular morphology. When used as cathode material, the obtained LiMn1.95Mg0.05O4 octahedra shows excellent electrochemical properties. This material can exhibit high capacity retention of 96.8% with 100th discharge capacity of 111.6 mAh g−1 at 1.0 C. Moreover, the rate performance and high-temperature cycling stability of LiMn2O4 are effectively improved by the co-modification strategy based on Mg-doping and octahedral morphology. These results are mostly given to the fact that the addition of magnesium ions can suppress the Jahn–Teller effect and the octahedral morphology contributes to the Mn dissolution, which can improve the structural stability of LiMn2O4.


Introduction
With the increasingly serious environmental pollution, new energy and environmental technology have caught more and more extensive attention. Under this circumstance, the research and development of lithium-ion batteries are receiving more and more attention at home and abroad since their first commercial application in 1991 [1][2][3]. As an important cathode material, LiMn 2 O 4 possesses a rather high cost advantage because of the abundant manganese resource and this material can be obtained by many preparation technologies [4][5][6][7][8]. Moreover, this material does not involve the use of toxic metal elements. All these advantages can promote large-scale applications of LiMn 2 O 4 . It must be noted, however, that the cycling stability and high temperature performance cannot meet the requirement of long endurance mileage [9][10][11][12].
According to the existing literature, the electrochemical performance of LiMn 2 O 4 can be severely affected by the Jahn-Teller distortion effect and Mn dissolution during the process of discharging and charging due to the fact that the Jahn-Teller distortion and Mn dissolution is closely related to the trivalent state (Mn 3+ ), which can seriously affect the discharging process and the discharged state [13][14][15][16]. In recent years, many optimization strategies (doping, coating, morphology control, etc.) have been developed to address these problems [11,15,[17][18][19][20][21]. Among them, the doping strategy

Materials and Methods
The octahedral LiMn 1.95 Mg 0.05 O 4 (LMMOO) particles were prepared by high temperature solid-phase method with magnesium nitrate and Mn 3 O 4 octahedra as doping agent and manganese source. The octahedral Mn 3 O 4 particles were firstly prepared via a hypothermal approach according to the existing literature [30]. Subsequently, in a typical synthesis process, stoichiometric LiOH·H 2 O, Mn 3 O 4 octahedra, and Mg(NO 3 ) 2 ·6H 2 O were ground to obtain the slurry mixture with the help of absolute ethanol. Then, the homogeneous mixture was dried in a drying oven and sintered at 700 • C for 10 h in air. In order to allow the comparison, the undoped LiMn 2 O 4 (LMO) and LiMn 1.95 Mg 0.05 O 4 (LMMO) particles were prepared by using electrolytic MnO 2 as manganese precursor.
The structure and morphology usually have an important impact on the electrochemical properties of cathode material. The obtained LiMn 2 O 4 and LiMn 1.95 Mg 0.05 O 4 samples were characterized by using XRD and SEM techniques. The effects of Mg-doping and octahedral morphology on the cycling stability of LiMn 2 O 4 were studied by fabricating the coin cells with the obtained spinels as cathode materials. The positive electrode was constituted from 85% synthesized product as cathode material, 10% acetylene black as conductive agent, and 5% polyvinylidene fluoride dissolved in N-methyl-2-pyrrolidone as binder. The metallic lithium foil was used as counter electrode, the polypropylene microporous membrane was used as diaphragm, and the 1 M lithium hexafluorophosphate (LiPF 6 ) solution in a mixture of the ethylene carbonate (EC) and diethyl carbonate (DEC) at a volume ratio of 1:1 was used as the electrolyte. All the electrochemical tests were carried out on LANHE CT2001A system (LANHE, Wuhan, China) and CHI660E electrochemical workstation (CH Instruments, Shanghai, China).

Results and Discussion
In order to confirm the structures of the obtained samples, the LiMn 2 O 4 , LiMn 1.95 Mg 0.05 O 4 , and octahedral LiMn 1.95 Mg 0.05 O 4 samples were characterized. It can be seen from Figure 1 that the diffraction peaks of the undoped LiMn 2 O 4 are in good agreement with the standard diffraction peaks of LiMn 2 O 4 (JCPDS No. . No other diffraction peaks of manganese oxide and magnesium oxide can be observed, suggesting the complete transformation of electrolytic manganese dioxide to LiMn 2 O 4 [31]. After introducing a small amount of magnesium ions, the obtained LiMn 1.95 Mg 0.05 O 4 sample was still present in the spinel cubic structure of LiMn 2 O 4 , which indicates that the addition of magnesium ions did not change the crystal structure [32,33]  . No other diffraction peaks of manganese oxide and magnesium oxide can be observed, suggesting the complete transformation of electrolytic manganese dioxide to LiMn2O4 [31]. After introducing a small amount of magnesium ions, the obtained LiMn1.95Mg0.05O4 sample was still present in the spinel cubic structure of LiMn2O4, which indicates that the addition of magnesium ions did not change the crystal structure [32,33]. For the LiMn1.95Mg0.05O4 sample obtained from Mn3O4 octahedra, the characteristic diffraction peaks were indexed to the spinel LiMn2O4. Moreover, the corresponding peak intensities are stronger than that of the LiMn2O4 and LiMn1.95Mg0.05O4 samples prepared from electrolytic manganese dioxide, suggesting the good crystalline quality of the octahedral LiMn1.95Mg0.05O4 sample [9,34]. Table 1 lists the related crystal parameters of these three samples. The addition of magnesium ions leads to the reduction of the lattice parameter and shrinking of unit cell volume, suggesting the more stable structural stability of the LiMn1.95Mg0.05O4 samples.   Figure 2 shows the SEM images of the octahedral Mn3O4, LiMn2O4, and Mg-doped LiMn2O4 samples. As shown in Figure 2a, the Mn3O4 particles prepared by hydrothermal approach present rather good octahedral morphology. For the octahedral LiMn1.95Mg0.05O4 sample shown in Figure 2b,c, it can be seen that it presents a narrow size distribution with regular morphology, which indicates that the LiMn1.95Mg0.05O4 sample inherits the special morphology of Mn3O4 octahedra [30]. Moreover, the particle size belongs to the submicron scale, which agrees with the particle size of Mn3O4. By contrast, the particle morphology of the undoped LiMn2O4 particles (Figure 2d) is irregular with micron grade particle size. Especially, the obvious agglomerated particle can be observed in the undoped LiMn2O4 particles. These unsatisfactory characteristics usually have a greater negative impact on the electrochemical properties of the cathode material [35,36]. For the LiMn1.95Mg0.05O4 sample (Figure 2e), it shows relatively good size distribution, which is closely related to the addition of a certain amount of magnesium ions, which agrees with the research result [23,33]. These results suggest that the combination of Mg-doping and octahedral morphology can be useful in optimizing the morphology and size distribution of LiMn2O4 particles.
To investigate the influence of jointly using the Mg-doping and octahedral morphology on the electrochemical performance, the LiMn2O4, LiMn1.95Mg0.05O4, and octahedral LiMn1.95Mg0.05O4 samples    Figure 2b,c, it can be seen that it presents a narrow size distribution with regular morphology, which indicates that the LiMn 1.95 Mg 0.05 O 4 sample inherits the special morphology of Mn 3 O 4 octahedra [30]. Moreover, the particle size belongs to the submicron scale, which agrees with the particle size of Mn 3 O 4 . By contrast, the particle morphology of the undoped LiMn 2 O 4 particles ( Figure 2d) is irregular with micron grade particle size. Especially, the obvious agglomerated particle can be observed in the undoped LiMn 2 O 4 particles. These unsatisfactory characteristics usually have a greater negative impact on the electrochemical properties of the cathode material [35,36]. of a certain amount of magnesium ions, which agrees with the research result [23,33]. These results suggest that the combination of Mg-doping and octahedral morphology can be useful in optimizing the morphology and size distribution of LiMn 2 O 4 particles. plateaus correspond to the intercalation/de-intercalation processes of lithium ions, which correspond to the two-phase equilibrium of λ-MnO2/Li0.5Mn2O4 and single-phase equilibrium of Li0.5Mn2O4/LiMn2O4, respectively. For the LiMn1.95Mg0.05O4 and octahedral LiMn1.95Mg0.05O4 samples, the initial charge-discharge curves show similar platform characteristics, and the potential interval of the Mg-doped spinels is less than that of the undoped spinel, suggesting the higher reaction kinetics of the Mg-doped spinels [39]. It is important to note, however, that the discharge voltage plateaus of the LiMn1.95Mg0.05O4 samples are slightly higher than that of the undoped LiMn2O4 sample, which may be related to the optimization of the Li + intercalation/deintercalation behaviors due to the addition of other cations in the spinel structure [10,33,39,40].   samples, the initial charge-discharge curves show similar platform characteristics, and the potential interval of the Mg-doped spinels is less than that of the undoped spinel, suggesting the higher reaction kinetics of the Mg-doped spinels [39]. It is important to note, however, that the discharge voltage plateaus of the LiMn 1.95 Mg 0.05 O 4 samples are slightly higher than that of the undoped LiMn 2 O 4 sample, which may be related to the optimization of the Li + intercalation/deintercalation behaviors due to the addition of other cations in the spinel structure [10,33,39,40].
The cycling performance is a very important index sign for the practical application of LiMn 2 O 4 . Figure 4 shows the cycling stability of the LiMn 2 O 4 , LiMn 1.95 Mg 0.05 O 4 , and octahedral LiMn 1.95 Mg 0.05 O 4 samples at 1.0 C. For the undoped LiMn 2 O 4 sample, it exhibits an initial capacity of 116.6 mAh g −1 with unsatisfactory cycling stability. After 100 cycles, the discharge capacity presents much decrease with the 100th capacity of 83.5 mAh g −1 . Such poor performance is mainly attributed to the wide size distribution and large agglomerated particle [35]. When adding some magnesium ions, the LiMn 1.95 Mg 0.05 O 4 sample shows higher capacity retention than that of undoped LiMn 2 O 4 sample. Although the addition of magnesium ions decreases the initial discharge capacity, the capacity retention of the LiMn 1.95 Mg 0.05 O 4 sample is enhanced greatly. After 100 cycles, the discharge capacity can maintain 100.1 mAh g −1 with high retention of 89.5%. The improvement in cycling stability is attributed to the fact that the introduction of magnesium ions can strengthen the structural stability of LiMn 2 O 4 by inhibiting the Jahn-Teller effect and reducing the cell volume [23,33]. It is important to note that the octahedral LiMn 1 [29,36].   The cycling performance is a very important index sign for the practical application of LiMn2O4. Figure 4 shows the cycling stability of the LiMn2O4, LiMn1.95Mg0.05O4, and octahedral LiMn1.95Mg0.05O4 samples at 1.0 C. For the undoped LiMn2O4 sample, it exhibits an initial capacity of 116.6 mAh g −1 with unsatisfactory cycling stability. After 100 cycles, the discharge capacity presents much decrease with the 100th capacity of 83.5 mAh g −1 . Such poor performance is mainly attributed to the wide size distribution and large agglomerated particle [35]. When adding some magnesium ions, the LiMn1.95Mg0.05O4 sample shows higher capacity retention than that of undoped LiMn2O4 sample. Although the addition of magnesium ions decreases the initial discharge capacity, the capacity retention of the LiMn1.95Mg0.05O4 sample is enhanced greatly. After 100 cycles, the discharge capacity can maintain 100.1 mAh g −1 with high retention of 89.5%. The improvement in cycling stability is attributed to the fact that the introduction of magnesium ions can strengthen the structural stability of LiMn2O4 by inhibiting the Jahn-Teller effect and reducing the cell volume [23,33]. It is important to note that the octahedral LiMn1.95Mg0.05O4 sample can show more excellent cycling performance. Compared with the undoped LiMn2O4 and LiMn1.95Mg0.05O4 samples, the capacity retention of the octahedral LiMn1.95Mg0.05O4 sample can reach up to 96.8% after 100 cycles with the 100th capacity of 111.6 mAh g −1 . Such excellent cycling stability mainly benefits from the synergistic effect of the Mgdoping and octahedral morphology. The Mg-doping can suppress the Jahn-Teller effect and the octahedral morphology can contribute to inhibit the Mn dissolution, which can improve the structural stability of LiMn2O4 [29,36]. To investigate the effect of jointly using the Mg-doping and octahedral morphology on the rate capability, the LiMn2O4, LiMn1.95Mg0.05O4, and octahedral LiMn1.95Mg0.05O4 samples were successively cycled at 0.5, 1.0, 2.0, and 5.0 C, respectively. Figure 5a presents the characteristic discharge curves of the octahedral LiMn1.95Mg0.05O4 sample (the representative of these three samples) at different cycling rates. As shown here, the discharge capacity and voltage platform are significantly affected by the high cycling rate. When the cycling rate gradually increases, the boundary of the two voltage plateaus become smooth and fuzzy and the discharge capacity gradually decreases due to the increased polarization, which are in accordance with the existing literature [41][42][43]. Figure 5b shows the corresponding cycling performance of the LiMn2O4, LiMn1.95Mg0.05O4, and octahedral LiMn1.95Mg0.05O4 samples at varying cycling rates. The undoped LiMn2O4 sample exhibits a discharge capacity of 126.7 mAh g −1 at low cycling rate of 0.5 C. With the increase of the cycling rates, the discharge capacity is influenced greatly. As the cycling rate increases to 5.0 C, the undoped LiMn2O4 sample only exhibits 55.0 mAh g −1 with rather low retention of 43.4%. By contrast, the Mg-doped LiMn2O4 samples present outstanding cycling stability at high cycling rate. Especially, the octahedral LiMn1.95Mg0.05O4 sample can exhibit a higher capacity of 91.8 mAh g −1 at a high cycling rate of 5.0 C. To further explore the high-rate cycling stability, the octahedral LiMn1.95Mg0.05O4 samples were cycled at 10 C. The corresponding characteristic discharge curves are shown in Figure 6a. It can be found that the  Figure 5a presents the characteristic discharge curves of the octahedral LiMn 1.95 Mg 0.05 O 4 sample (the representative of these three samples) at different cycling rates. As shown here, the discharge capacity and voltage platform are significantly affected by the high cycling rate. When the cycling rate gradually increases, the boundary of the two voltage plateaus become smooth and fuzzy and the discharge capacity gradually decreases due to the increased polarization, which are in accordance with the existing literature [41][42][43].  Figure 6a. It can be found that the characteristic voltage plateaus in the discharge curves become blurred to a large extent, which agrees with the research result [39,44,45]. Figure 6b presents the cycling performance of the octahedral LiMn 1.95 Mg 0.05 O 4 samples at 10 C. It can show satisfactory retention of 97.5% after 100 cycles with initial capacity of 72.1 mAh g −1 . The above results suggest that the co-modification strategy based on Mg-doping and octahedral morphology is an important means for effectively improving the rate capability of LiMn 2 O 4 .
Materials 2019, 12, x FOR PEER REVIEW 6 of 10 characteristic voltage plateaus in the discharge curves become blurred to a large extent, which agrees with the research result [39,44,45]. Figure 6b presents the cycling performance of the octahedral LiMn1.95Mg0.05O4 samples at 10 C. It can show satisfactory retention of 97.5% after 100 cycles with initial capacity of 72.1 mAh g −1 . The above results suggest that the co-modification strategy based on Mg-doping and octahedral morphology is an important means for effectively improving the rate capability of LiMn2O4.   Figure 7a presents the high-temperature cycling stability of the LiMn2O4 and octahedral LiMn1.95Mg0.05O4 samples at 1.0 C. It can be seen that the cycling stability of the undoped LiMn2O4 sample is much poorer than that of the octahedral Mg-doped LiMn2O4 sample. The initial discharge capacity of the undoped LiMn2O4 sample is comparable to the test results shown in Figure 4, but the capacity retention is rather poor. After 50 cycles, this sample only presents low capacity retention of 80.5%. It is important to note that the octahedral LiMn1.95Mg0.05O4 can present excellent capacity retention of 92.5% with a satisfactory 50th discharge capacity of 106.6 mAh g −1 . Figure 7b shows the rate capability of the LiMn2O4 and octahedral LiMn1.95Mg0.05O4 samples at 55 °C. As shown here, the octahedral LiMn1.95Mg0.05O4 sample shows more stable high-temperature cycling stability at different rates, especially the high cycling rate. When tested at 5.0 C, the octahedral LiMn1.95Mg0.05O4 sample can maintain the discharge capacity of 95.6 mAh g −1 . Unfortunately, the undoped LiMn2O4 sample characteristic voltage plateaus in the discharge curves become blurred to a large extent, which agrees with the research result [39,44,45]. Figure 6b presents the cycling performance of the octahedral LiMn1.95Mg0.05O4 samples at 10 C. It can show satisfactory retention of 97.5% after 100 cycles with initial capacity of 72.1 mAh g −1 . The above results suggest that the co-modification strategy based on Mg-doping and octahedral morphology is an important means for effectively improving the rate capability of LiMn2O4.   Figure 7a presents the high-temperature cycling stability of the LiMn2O4 and octahedral LiMn1.95Mg0.05O4 samples at 1.0 C. It can be seen that the cycling stability of the undoped LiMn2O4 sample is much poorer than that of the octahedral Mg-doped LiMn2O4 sample. The initial discharge capacity of the undoped LiMn2O4 sample is comparable to the test results shown in Figure 4, but the capacity retention is rather poor. After 50 cycles, this sample only presents low capacity retention of 80.5%. It is important to note that the octahedral LiMn1.95Mg0.05O4 can present excellent capacity retention of 92.5% with a satisfactory 50th discharge capacity of 106.6 mAh g −1 . Figure 7b shows the rate capability of the LiMn2O4 and octahedral LiMn1.95Mg0.05O4 samples at 55 °C. As shown here, the octahedral LiMn1.95Mg0.05O4 sample shows more stable high-temperature cycling stability at different rates, especially the high cycling rate. When tested at 5.0 C, the octahedral LiMn1.95Mg0.05O4 sample can maintain the discharge capacity of 95.6 mAh g −1 . Unfortunately, the undoped LiMn2O4 sample   Figure 4, but the capacity retention is rather poor. After 50 cycles, this sample only presents low capacity retention of 80.5%. It is important to note that the octahedral LiMn 1.95 Mg 0.05 O 4 can present excellent capacity retention of 92.5% with a satisfactory 50th discharge capacity of 106.6 mAh g −1 . Figure 7b  and Figure 8b shows the corresponding equivalent circuit model. According to the research result [9,10,23], the charge transfer resistance (R2) in the high-frequency region has strong ties to the electrochemical properties. Therefore, we mainly studied the R2 value to confirm the effect of both the Mg-doping and octahedral morphology on the electrochemical performance. As shown in Figure  8a, the combination of Mg-doping and octahedral morphology produces an important influence on the R2 value. The addition of magnesium ions can suppress the Jahn-Teller distortion effect to improve the structural stability, and the octahedral morphology of LiMn2O4 octahedra suppresses the dissolution of Mn in electrolyte [30,36]. Moreover, the uniform particle size distribution also contributes to the diffusion efficiency of lithium ions [33,39,43]. As a result, the octahedral LiMn1.95Mg0.05O4 sample presents lower initial charge-transfer resistance than that of the undoped spinel, which suggests excellent electrochemical properties.     Figure 8b shows the corresponding equivalent circuit model. According to the research result [9,10,23], the charge transfer resistance (R 2 ) in the high-frequency region has strong ties to the electrochemical properties. Therefore, we mainly studied the R 2 value to confirm the effect of both the Mg-doping and octahedral morphology on the electrochemical performance. As shown in Figure 8a, the combination of Mg-doping and octahedral morphology produces an important influence on the R 2 value. The addition of magnesium ions can suppress the Jahn-Teller distortion effect to improve the structural stability, and the octahedral morphology of LiMn 2 O 4 octahedra suppresses the dissolution of Mn in electrolyte [30,36]. Moreover, the uniform particle size distribution also contributes to the diffusion efficiency of lithium ions [33,39,43]. As a result, the octahedral LiMn 1.95 Mg 0.05 O 4 sample presents lower initial charge-transfer resistance than that of the undoped spinel, which suggests excellent electrochemical properties.

Conclusions
Materials 2019, 12, x FOR PEER REVIEW 7 of 10 presents unsatisfactory rate capability, which further confirms the synergistic effect of the Mg-doping and octahedral morphology. Figure 8a presents the Nyquist plots of the LiMn2O4 and octahedral LiMn1.95Mg0.05O4 samples, and Figure 8b shows the corresponding equivalent circuit model. According to the research result [9,10,23], the charge transfer resistance (R2) in the high-frequency region has strong ties to the electrochemical properties. Therefore, we mainly studied the R2 value to confirm the effect of both the Mg-doping and octahedral morphology on the electrochemical performance. As shown in Figure  8a, the combination of Mg-doping and octahedral morphology produces an important influence on the R2 value. The addition of magnesium ions can suppress the Jahn-Teller distortion effect to improve the structural stability, and the octahedral morphology of LiMn2O4 octahedra suppresses the dissolution of Mn in electrolyte [30,36]. Moreover, the uniform particle size distribution also contributes to the diffusion efficiency of lithium ions [33,39,43]. As a result, the octahedral LiMn1.95Mg0.05O4 sample presents lower initial charge-transfer resistance than that of the undoped spinel, which suggests excellent electrochemical properties.

Conclusions
To summarize, the octahedral LiMn 1.95 Mg 0.05 O 4 sample was prepared by high temperature solid-phase method with magnesium nitrate and Mn 3

Conflicts of Interest:
The authors declare no conflict of interest.