Enhanced Cycling Stability of LiCuxMn1.95−xSi0.05O4 Cathode Material Obtained by Solid-State Method

The LiCuxMn1.95−xSi0.05O4 (x = 0, 0.02, 0.05, 0.08) samples have been obtained by a simple solid-state method. XRD and SEM characterization results indicate that the Cu-Si co-doped spinels retain the inherent structure of LiMn2O4 and possess uniform particle size distribution. Electrochemical tests show that the optimal Cu-doping amount produces an obvious improvement effect on the cycling stability of LiMn1.95Si0.05O4. When cycled at 0.5 C, the optimal LiCu0.05Mn1.90Si0.05O4 sample exhibits an initial capacity of 127.3 mAh g−1 with excellent retention of 95.7% after 200 cycles. Moreover, when the cycling rate climbs to 10 C, the LiCu0.05Mn1.90Si0.05O4 sample exhibits 82.3 mAh g−1 with satisfactory cycling performance. In particular, when cycled at 55 °C, this co-doped sample can show an outstanding retention of 94.0% after 100 cycles, whiles the LiMn1.95Si0.05O4 only exhibits low retention of 79.1%. Such impressive performance shows that the addition of copper ions in the Si-doped spinel effectively remedy the shortcomings of the single Si-doping strategy and the Cu-Si co-doped spinel can show excellent cycling stability.


Introduction
Lithium-ion batteries have been applied extensively in a lot of power supply fields, like in pure electrical vehicles (EVs), unmanned aerial vehicles and smartphones. As one important part of lithium-ion batteries, cathode materials have played a crucial role in terms of electrochemical performance [1][2][3][4][5][6][7]. Among the existing cathode materials, LiMn 2 O 4 possesses major advantages and great potential for the large-scale commercial application due to the mature production technology, cheap production costs and non-pollution characteristics [8][9][10]. It is important to note, however, that this material shows poor cycling stability and elevated-temperature performance, which produces a serious negative effect on promoting the large-scale commercial application. These unsatisfactory deficiencies are mainly caused by Jahn-Teller distortion and manganese dissolution [11][12][13][14].
Unfortunately, the capacity retention is only 85.1% after 100 cycles. It was obvious that the optimization degree of replacing the Mn 4+ ions with tetravalent cations cannot reach the demand for large-scale application of LiMn 2 O 4 . It has been reported that the Cu-doping strategy can make a positive contribution in enhancing the cycling stability due to the fact that the addition of copper ions in the LiMn 2 O 4 decrease the trivalent manganese ions and cell volume of LiMn 2 O 4 , which can inhibit the Jahn-Teller effect and enhance structural stability [23]. In this work, the LiCu x Mn 1.95−x Si 0.05 O 4 (x = 0, 0.02, 0.05, 0.08) samples have been obtained by a simple solid-state method. The effect of copper doping content on the structures, morphologies and cycling life of the LiCu x Mn 1.95−x Si 0.05 O 4 samples is discussed. The results indicate the addition of copper ions in the Si-doped spinel effectively remedy the shortcomings of the single Si-doping strategy and the Cu-Si co-doped spinel can show excellent cycling stability.

Materials and Methods
The  10:5). The anode material and diaphragm are lithium foil and Celgard 2400 polymer, respectively. The mixture of 1 M LiPF 6 , ethyl methyl carbonate (EMC), ethylene carbonate (EC) and dimethyl carbonate (DMC) was used as electrolyte (EMC:EC:DMC = 1:1:1). The electrochemical measurement was executed on LAND (Wuhan, China) battery testing system. The electrochemical impedance spectroscopy (EIS) were tested by CS-350 electrochemical workstation (Wuhan, China). These tests were investigated by using CR2016 coin-type cells.  [17,33], where lithium and manganese ions occupy the tetrahedral sites (8a) and octahedral sites (16d), respectively. According to the reported research result, the (220) characteristic peak may be observed if other cations occupied the tetrahedral sites [34]. However, there is no (220) characteristic peak in Figure 1, suggesting that the copper ions successfully replaced the manganese ions in octahedral sites. According to the reported literature [35], the intensity ratio of (311)/(400) peaks can be optimized by replacing the Mn ions with some other cation ions in the spinel structure of LiMn2O4. If this intensity ratio is in the range of 0.96-1.10, the obtained samples usually show excellent cycling stability. Table 1 lists this intensity ratio of LiCuxMn1.95-xSi0.05O4 (x = 0, 0.02, 0.05, 0.08) samples. It can be seen that the Cu-doping strategy has played a positive role in optimizing this intensity ratio. The copper and silicon co-doped spinels can present a larger intensity ratio than that of the silicon co-doped spinel. Therefore, it can be inferred that the further addition of copper ions in the silicon-doped sample may greatly enhance the cycling stability. 1.03 Figure 2 shows the SEM images of the LiCuxMn1.95−xSi0.05O4 (x = 0, 0.02, 0.05, 0.08) samples. As shown in Figure 2a, the silicon-doped LiMn2O4 particles present less-than-ideal size distribution. For the copper and silicon co-doped LiMn2O4 samples, the introduction of some copper ions can further optimize the mean diameter and size distribution. When the copper doping content increases, the mean diameter of the LiCuxMn1.95−xSi0.05O4 (x = 0.02, 0.05, 0.08) has a decreasing tendency. It is important to note that the LiCu0.05Mn1.90Si0.05O4 particles shown in Figure 2c present the quite uniform size distribution. The above-mentioned results suggest that introducing some copper ions can effectively improve the crystallinity and optimize the size distribution, which is conducive to the enhancement of cycling stability. According to the reported literature [35], the intensity ratio of (311)/(400) peaks can be optimized by replacing the Mn ions with some other cation ions in the spinel structure of LiMn 2 O 4 . If this intensity ratio is in the range of 0.96-1.10, the obtained samples usually show excellent cycling stability. Table 1 lists this intensity ratio of LiCu x Mn 1.95-x Si 0.05 O 4 (x = 0, 0.02, 0.05, 0.08) samples. It can be seen that the Cu-doping strategy has played a positive role in optimizing this intensity ratio. The copper and silicon co-doped spinels can present a larger intensity ratio than that of the silicon co-doped spinel. Therefore, it can be inferred that the further addition of copper ions in the silicon-doped sample may greatly enhance the cycling stability. 1.03 Figure 2 shows the SEM images of the LiCu x Mn 1.95−x Si 0.05 O 4 (x = 0, 0.02, 0.05, 0.08) samples. As shown in Figure 2a, the silicon-doped LiMn 2 O 4 particles present less-than-ideal size distribution. For the copper and silicon co-doped LiMn 2 O 4 samples, the introduction of some copper ions can further optimize the mean diameter and size distribution. When the copper doping content increases, the mean diameter of the LiCu x Mn 1.95−x Si 0.05 O 4 (x = 0.02, 0.05, 0.08) has a decreasing tendency. It is important to note that the LiCu 0.05 Mn 1.90 Si 0.05 O 4 particles shown in Figure 2c present the quite uniform size distribution. The above-mentioned results suggest that introducing some copper ions can effectively improve the crystallinity and optimize the size distribution, which is conducive to the enhancement of cycling stability. Figure 3a shows the first discharge curves of the LiCu x Mn 1.95−x Si 0.05 O 4 (x = 0, 0.02, 0.05, 0.08) samples. All these samples present characteristic discharge curves, which show two distinct voltage platforms around 4.10-4.15 V and 3.95-4.00 V, suggesting that introducing some copper ions do not change the electrochemical redox reaction mechanism and all these copper and silicon co-doped LiMn 2 O 4 samples processes two extraction/insertion process of Li + ions [14,32]. Figure 3b presents the cycling stability of the LiCu x Mn 1.95−x Si 0.05 O 4 (x = 0, 0.02, 0.05, 0.08) samples. The cycling stability of these co-doped samples were remarkably enhanced as the copper doping content increased, due to the suppressed Jahn-Teller effect and stronger structural stability [23]. Note, however, that the addition of more copper ions has a great negative impact on the reversible capacity of the LiCu 0.08 Mn 1 Figure 3a shows the first discharge curves of the LiCuxMn1.95−xSi0.05O4 (x = 0, 0.02, 0.05, 0.08) samples. All these samples present characteristic discharge curves, which show two distinct voltage platforms around 4.10-4.15 V and 3.95-4.00 V, suggesting that introducing some copper ions do not change the electrochemical redox reaction mechanism and all these copper and silicon co-doped LiMn2O4 samples processes two extraction/insertion process of Li + ions [14,32]. Figure 3b presents the cycling stability of the LiCuxMn1.95−xSi0.05O4 (x = 0, 0.02, 0.05, 0.08) samples. The cycling stability of these co-doped samples were remarkably enhanced as the copper doping content increased, due to the suppressed Jahn-Teller effect and stronger structural stability [23]. Note, however, that the addition of more copper ions has a great negative impact on the reversible capacity of the LiCu0.08Mn1.87Si0.05O4 sample in spite of the improvement of cycling life (Figure 3c). These unsatisfactory results are principally because introducing more copper ions can cause the reduction of Mn 3+ , which is unfavourable to the Mn(III)-Mn(IV) interconversion. Figure 3d shows the long cycling performance of the LiCuxMn1.95−xSi0.05O4 (x = 0, 0.05) samples. For the LiCu0.05Mn1.90Si0.05O4 sample, the reversible capacity peaked at 127.3 mAh g −1 , which is slightly lower than that of the LiMn1.95Si0.05O4 sample. After 200 cycles, the LiCu0.05Mn1.90Si0.05O4 sample can still exhibit 121.8 mAh g −1 with outstanding retention of 95.7%. Unfortunately, the LiMn1.95Si0.05O4 sample shows lower capacity with worse cycling life. After 200 cycles, this sample only delivers 108.3 mAh g −1 with low retention of 81.6%. According to the reference [32], the undoped LiMn2O4 only delivers a discharge capacity of 48.3 mAh g −1 with capacity retention of 37.8% after 100 cycles, which is much lower than that of the LiSi0.05Mn1.95O4 sample. Although the silicon-doping enhance the cycling performance, the further addition of copper ions can significantly enhance the cycling stability of LiMn2O4.   Figure 4a shows the corresponding discharge curves of the representative LiCu 0.05 Mn 1.90 Si 0.05 O 4 sample at varying rates. It can be seen that there are two voltage platforms which are obvious at 0.2 C and 0.5 C, suggesting the diffusion process of lithium ions [36]. When the rate further increases, these two potential plateaus gradually show ambiguous boundary and shifted toward lower voltage. This result has a lot to do with the polarization effect and ohmic drop [37]. Figure 4b Figure 4a shows the corresponding discharge curves of the representative LiCu0.05Mn1.90Si0.05O4 sample at varying rates. It can be seen that there are two voltage platforms which are obvious at 0.2 C and 0.5 C, suggesting the diffusion process of lithium ions [36]. When the rate further increases, these two potential plateaus gradually show ambiguous boundary and shifted toward lower voltage. This result has a lot to do with the polarization effect and ohmic drop [37]. Figure 4b shows the cycling stability of the LiCu0.05Mn1.90Si0.05O4 and LiMn1.95Si0.05O4 samples at varying rates. When cycled at 0.2 C, the capacities of these two samples reached up to 138.5 and 131.4 mAh g −1 , respectively. However, what is important to pay attention to is the reversible capacity of the LiCu0.05Mn1.90Si0.05O4 sample, which showed much more obvious difference at high rates of 5.0 C.

Results and Discussion
To further study the cycling performance at a high rate, the LiCu0.05Mn1.90Si0.05O4 and LiMn1.95Si0.05O4 samples were tested at 10 C. For the LiCu0.05Mn1.90Si0.05O4 sample, the two characteristic voltage plateaus shown in Figure 4c become blurred to a certain extent. By contrast, the LiMn1.95Si0.05O4 presents lower voltage plateau and corresponding to this, the capacity of this material shows severe degradation. Figure 4d presents the cycling life of these two spinels at 10 C. The LiMn1.95Si0.05O4 sample shows unsatisfactory capacity retention of 85.7% with low initial capacity of 68.4 mAh g −1 , while the LiCu0.05Mn1.90Si0.05O4 sample can display a higher capacity of 82.3 mAh g −1 .
More importantly, after 100 cycles, the corresponding retention can reach up to 94.0% with the 100th cycle with a capacity of 77.4 mAh g −1 . The above discussion indicates that the introduction of copper ions has great value in the optimization of the rate capability.  1% with a lower capacity of 106.4 mAh g −1 after 100th cycle. Such low discharge capacity after 100 cycles is mostly given to the fact that the high temperature accelerates the dissolution of manganese and undermines the structural stability of LiMn 2 O 4 . Note, however, that the LiCu 0.05 Mn 1.90 Si 0.05 O 4 sample can still show much better cycling stability although these two samples show low discharge capacity after 100 cycles. These results suggest that introducing some copper ions can be favorable for enhancing the cycling stability at high-temperature. Figure 5b shows the rate capability of these two samples at 55 • C. When cycled at low rates, the LiCu 0.05 Mn 1 Figure 5 shows the electrochemical properties of the LiCu0.05Mn1.90Si0.05O4 and LiMn1.95Si0.05O4 samples at 55 °C. As shown in Figure 5a, the LiCu0.05Mn1.90Si0.05O4 exhibits an initial capacity of 127.2 mAh g −1 at 0.5 C. After 100 cycles, this sample still maintains a high capacity of 119.6 mAh g −1 with excellent retention of 94.0%. However, the LiMn1.95Si0.05O4 sample shows much lower retention than that of the LiCu0.05Mn1.90Si0.05O4. The capacity retention of the LiMn1.95Si0.05O4 sample is only 79.1% with a lower capacity of 106.4 mAh g −1 after 100th cycle. Such low discharge capacity after 100 cycles is mostly given to the fact that the high temperature accelerates the dissolution of manganese and undermines the structural stability of LiMn2O4. Note, however, that the LiCu0.05Mn1.90Si0.05O4 sample can still show much better cycling stability although these two samples show low discharge capacity after 100 cycles. These results suggest that introducing some copper ions can be favorable for enhancing the cycling stability at high-temperature. Figure 5b shows the rate capability of these two samples at 55 °C. When cycled at low rates, the LiCu0.05Mn1.90Si0.05O4 and LiMn1.95Si0.05O4 samples show similar capacities. However, as the cycling rate increased, these two samples gradually show some difference. When cycled at 5.0 C, the LiCu0.05Mn1.90Si0.05O4 sample can show 103.4 mAh g −1 while the LiMn1.95Si0.05O4 only shows 91.7 mAh g −1 . The above-mentioned results suggest that the addition of copper ions can further improve the rate capability of LiMn1.95Si0.05O4 at high-temperature.  Figure 6a,b show the EIS results of the LiCu0.05Mn1.90Si0.05O4 and LiMn1.95Si0.05O4 samples. It has been reported previously that the charge transfer resistance (R2) corresponds to the high-frequency semicircle, which has a connection with the cycling life [14,34]. Therefore, we mainly determine the R2 values to confirm the effect of introducing copper ions on the cycling stability.   It has been reported previously that the charge transfer resistance (R 2 ) corresponds to the high-frequency semicircle, which has a connection with the cycling life [14,34]. Therefore, we mainly determine the R 2 values to confirm the effect of introducing copper ions on the cycling stability. However, this value increases to 158.1 Ω cm 2 with growth rate of 69.6%. Through the above analysis, it is concluded that replacing some trivalent manganese ions with copper ions can have a constructive role in decreasing the R 2 value and enhancing the diffusion of lithium ions [33,38,39].  Figure 6a,b show the EIS results of the LiCu0.05Mn1.90Si0.05O4 and LiMn1.95Si0.05O4 samples. It has been reported previously that the charge transfer resistance (R2) corresponds to the high-frequency semicircle, which has a connection with the cycling life [14,34]. Therefore, we mainly determine the R2 values to confirm the effect of introducing copper ions on the cycling stability. Table 2 lists the fitting values of R2. For the LiCu0.05Mn1.90Si0.05O4 sample, the original R2 value only reach 70.2 Ω cm 2 but increase to 116.0 Ω cm 2 after 200 cycles. The R2 value increase was relatively small with low growth rate of 64.5%. By contrast, the LiMn1.95Si0.05O4 shows a higher original R2 value (93.2 Ω cm 2 ). However, this value increases to 158.1 Ω cm 2 with growth rate of 69.6%. Through the above analysis, it is concluded that replacing some trivalent manganese ions with copper ions can have a constructive role in decreasing the R2 value and enhancing the diffusion of lithium ions [33,38,39].

Conclusions
The LiCu x Mn 1.95−x Si 0.05 O 4 (x = 0, 0.02, 0.05, 0.08) samples have been obtained by a simple solid-state method. The further addition of copper ions in the LiMn 2 O 4 can decrease the trivalent manganese ions and cell volume of LiMn 2 O 4 , which can inhibit the Jahn-Teller effect and enhance structural stability. As the optimal Cu-Si co-doped spinel, the LiCu 0.05 Mn 1.90 Si 0.05 O 4 sample possessed even size distribution. More importantly, it showed much better cycling stability and elevated temperature performance than the Si-doped LiMn 2 O 4 sample. When cycled at 0.5 C, the LiCu 0.05 Mn 1.90 Si 0.05 O 4 sample exhibited 127.3 mAh g −1 , which is slightly lower than that of the LiMn 1.95 Si 0.05 O 4 sample. After 200 cycles, the LiCu 0.05 Mn 1.90 Si 0.05 O 4 sample could exhibit 121.8 mAh g −1 with outstanding retention of 95.7% at 0.5 C. Moreover, this co-doped sample can show outstanding rate capability and high-temperature performance. All these results suggest that the further addition of copper ions can produce an obvious effect in enhancing the cycling stability of the silicon-doped LiMn 2 O 4 .
Author Contributions: H.Z. and J.S. conceived and designed the experiments; H.Z. and F.L. performed the experiments; all authors analyzed the data; H.Z. wrote the paper; all authors discussed the results and commented on the paper.