Introduction

Energy crisis and greenhouse gas emission have brought unprecedented attention all over the world. High energy and high power rechargeable lithium-ion batteries are the most promising energy storage systems for electric vehicles and plug-in hybrids. The development of those systems is in need of battery with higher energy density and longer life (Goodenough 2014; Croguennec and Palacin 2015; Cai et al. 2014, 2015, 2016; Zhang et al. 2018; Fang et al. 2017). LiNi0.5Mn1.5O4 has caused great interests as high-voltage cathode material (Santhanam and Rambabu 2010; Manthiram et al. 2014; Yi et al. 2016; Kim et al. 2004). LiNi0.5Mn1.5O4 has a high theoretical energy density above 600 Wh kg−1 because of its high specific capacity (147 mAh g−1) and high potential plateau (4.7 V). Besides, the facial three-dimensional lithium-ion diffusion pathways of LiNi0.5Mn1.5O4 give rise to its high power density.

The synthesis methods of LiNi0.5Mn1.5O4 mainly include wet chemical reaction and solid-state method. Solid-state method (Lv et al. 2014; Zhu et al. 2013) is used because of its simplification and low cost. But in this method, it is difficult to achieve the even mix of Mn and Ni elements. Wet chemical methods include co-precipitation methods, sol-gel methods (Li and Li 2014; Bauer et al. 2015), and hydrothermal methods (Xue et al. 2014). The co-precipitation method has been reported by many groups (Luo 2015; Yao et al. 2014, Zhang et al. 2012; Gu et al. 2014, 2015; Chemelewski et al. 2013; Liu et al. 2013a, b; Li et al. 2013; Axmann et al. 2016; Liu et al. 2015). Its main advantage is homogeneously mixing Ni and Mn. However, traditional co-precipitation procedure is very inconvenient and complex. Because pH value and temperature conditions of the solution must be controlled precisely to produce precursor using a continuously stirred tank reactor (CSTR). And ammonia is usually used as complexing agent. There are two main systems: hydroxide and carbonate. In hydroxide co-precipitation method, pH is controlled at about 10 with NaOH as precipitant. And argon serves as a shielding gas to protect Mn from oxidation. In carbonate co-precipitation method, pH is controlled at about 8 with Na2CO3 as precipitant. If pH value of solution is too high, transition metal carbonate precipitate will turn into hydroxide precipitate.

Herein, samples LiNi0.5 − xMn1.5 + xO4 (x = 0, 0.05, and 0.1) are prepared by the simple co-precipitation, in which temperature and pH conditions do not need control. And LNMO materials synthesized by the simple co-precipitation method show better electrochemical performance than those of traditional co-precipitation methods. Besides, Ni/Mn ratios in LNMO materials are of great importance to the performances (Wan et al. 2015; Ren et al. 2014; Song et al. 2012). Different Ni/Mn ratios can change the Mn3+ contents in spinel, leading to different electrochemical performances. Here, the effects of different Ni/Mn ratios in spinel are studied in detail.

Experimental

Preparation of LNMO

A NaHCO3 solution (0.4 mol L−1) was poured to the NiSO4 and MnSO4 solution rapidly under stirring. The molar ratios of NiSO4 and MnSO4 are 0.5:1.5, 0.45:1.55, and 0.4:1.6, respectively. The total concentration is 0.04 mol L−1. The mixture was kept stirring for 5 h. Then, the produced precipitate was collected, washed, and dried. The precipitate and LiOH·H2O were dispersed in ethanol and stirred. The ethanol was evaporated under stirring. The obtained mixture was calcined at 800 °C for 12 h in air to obtain LiNi0.5Mn1.5O4, LiNi0.45Mn1.55O4, and LiNi0.4Mn1.6O4 powder, denoted as samples Ni0.5, Ni0.45, and Ni0.4, respectively.

Characterization

Scanning electron microscopy (SEM, Quanta-200) and transmission electron microscopy (TEM, FEI Tecnai-F30FEG) were used to analyze morphology and microstructure of the as-prepared materials. Inductively coupled plasma (ICP, PerkinElmer, Optima 5300DV) was used to analyze Ni/Mn ratios. X-ray powder diffraction analysis was carried out with a D/max-RB diffractometer using a Cu Kα source and recorded with a step of 0.05°.

The 2025 coin cells were assembly in a pure argon-filled glove box. To make the working electrode, 80 wt% active material, 10 wt% conductive acetylene black as conductive agent, and 10 wt% polyvinylidene fluoride as binder with proper amount of n-methyl pyrrolidinone (NMP) as a solvent were put on a piece of aluminum foil. After drying in a vacuum oven at 120 °C, the foil was punched into a circular electrode. The loading weight of the cathode active material was about 3 mg cm−2. Lithium metal was used as anode in the cell. Charge and discharge tests were carried out on a battery testing system (NEWWARE BTS 5 V 10 mA) between 3.5 and 4.95 V. If the current density was higher than 0.2 C, the cells were charged galvanostatically to 4.95 V first and kept at 4.95 V until the current decreased to 0.1 C.

Results and discussion

Co-precipitation methods have been widely used to synthesize LNMO materials. Precipitate is generated simultaneously and Ni/Mn elements distribute uniformly in co-precipitation procedure. However, traditional co-precipitation procedure is very inconvenient and complex. Here, precursors are synthesized by a simple co-precipitation method, in which NaHCO3 solution is poured into transition metal solution rapidly. Traditional co-precipitation methods (Luo 2015; Yao et al. 2014; Zhang et al. 2012; Gu et al. 2014, 2015; Chemelewski et al. 2013; Liu et al. 2013a, b; Li et al. 2013; Axmann et al. 2016; Liu et al. 2015) and simple co-precipitation method used in this work are compared in Scheme 1 and Table 1. During the simple co-precipitation procedure, temperature, and pH conditions of solution do not need control attributed to the following reasons. Complexing agent ammonia is not added, so transition metal cannot complex with NH3 and can be precipitated completely. In the simple co-precipitation procedure, NaHCO3 is used as precipitant instead of Na2CO3 and NaOH. The pH value of NaHCO3 solution is about 8.3, avoiding hydroxide precipitate. The NaHCO3 solution is poured, not dropped, into transition metal solution rapidly. So, transition metal can generate precipitate immediately and simultaneously. Moreover, LNMO materials synthesized using the simple co-precipitation method show better electrochemical performance than those of traditional co-precipitation methods, as shown in Table 1.

Scheme 1
scheme 1

Illustration of a traditional co-precipitation procedure and b simple co-precipitation method

Full size image
Table 1 Comparison of preparation conditions and electrochemical performance of LNMO (“-” means that the data is not reported)
Full size table

To study the effects of different Ni/Mn ratios, samples LiNi0.5Mn1.5O4, LiNi0.45Mn1.55O4, and LiNi0.4Mn1.6O4, denoted as Ni0.5, Ni0.45, and Ni0.4, are prepared by the simple co-precipitation method. The ICP is used to test the molar ratios of Ni/Mn in the samples. The ICP results show that the molar ratios of Ni/Mn in Ni0.5, Ni0.45, and Ni0.4 are 0.50/1.50, 0.44/1.56, and 0.39/1.61, close to the designed formulas in this work.

Figure 1a shows the XRD patterns of transition metal carbonate obtained after co-precipitation process. Peak positions of the three patterns are the same and the XRD patterns can be assigned to the rhombohedral structure of MnCO3 (JCPDS Card No.: 83-1763). Peak intensities of all precursors are low because they are obtained without calcination. Figure 1b shows the SEM images of Ni0.5Mn1.5(CO3)2. As can be seen, the obtained transition metal carbonate precursors are microspheres.

Fig. 1
figure 1

a XRD patterns and b SEM micrographs of precursors

Full size image

XRD patterns of LNMO samples are displayed in Fig. 2. The XRD patterns of all powders are representative of cubic spinel (JCPDS Card No.: 80-2162). Minor peaks at 37.5°, 43.6°, and 63.4° related to LixNi1 − xO impurity (Song et al. 2012; Liu et al. 2013a, b) are not observed in these patterns. It means that LNMO samples have very little impurity, beneficial for performance of LNMO (Yi and Hu 2007). The superstructure reflection at 15.3° is typical for an ordered LNMO (Axmann et al. 2016; Lee et al. 2012). The peak intensity at 15.3° of sample Ni0.4 is the strongest, meaning that sample Ni0.4a has the highest level of cation ordering.

Fig. 2
figure 2

a XRD patterns and b magnified XRD patterns of LNMO samples

Full size image

The morphologies and structures of different LNMO samples were investigated by TEM and SEM as shown in Fig. 3. The SEM images with low magnification are shown in Fig. 3k. All samples are consisted of spherical particles with uniform dispersion. The particle sizes of Ni0.5, Ni0.45, and Ni0.4 are about 4.1, 3.0, and 2.2 μm, respectively. Solubility product constant of NiCO3 and MnCO3 are 6.6 × 10−9 and 2.2 × 10−11. MnCO3 is more easily to be precipitated. Nucleation rate is increased with the decreasing of Ni/Mn ratios, leading to small LNMO particle size. SEM images with higher magnitude are shown as Fig. 3d–f. The LNMO particles are composed of ~ 200-nm primary particles. EDS mapping of sample Ni0.45 are shown in Fig. 3g–i. Mn and Ni are distributed uniformly in particles, meaning that the simple co-precipitation method can mix Mn and Ni homogeneously. The EDS results show that the atomic ratios of Ni/Mn in Ni0.5, Ni0.45, and Ni0.4 are 0.49/1.51, 0.42/1.58, and 0.38/1.62. As shown in Fig. 3l, LNMO samples have hollow structures. The hollow structures may be caused by Ostwald ripening during calcining process (Xue et al. 2017). The hollow structures are beneficial for electrochemical performance. The hollow structures can reduce Li+ ions transport path, leading to good rate performance. Besides, the hollow structures can buffer the volume change caused by Li+ insertion/extraction, which are helpful to improve cycling performance (Fang et al. 2018; Tong et al. 2018). The spinel crystal structure of LNMO is further confirmed by the high-resolution TEM images as shown in Fig. 3m, n, in which the interplanar spacing is consistent with the (111) atomic planes.

Fig. 3
figure 3

SEM micrographs of sample a, d Ni0.5, b, e Ni0.45, c, f Ni0.4, g–i EDS mapping of Ni0.45, k low-magnification SEM image, and l–n TEM micrographs of sample Ni0.45

Full size image

The electrochemical performances of LNMO samples were tested in coin cells with lithium metal as anode. The charge-discharge curves of 0.2 C are showed in Fig. 4. The capacities of LNMO are mainly generated from Ni4+/Ni2+ redox. So, the decreases of Ni/Mn ratios in samples reduce specific capacities. The specific discharge capacities of Ni0.5, Ni0.45, and Ni0.4 are 130, 127, and 122 mAh g−1, respectively. All samples show two distinct discharge plateaus at ∼ 4.7 V, attributed to the reduction of Ni4+ to Ni3+ and Ni3+ to Ni2+, respectively (Lee et al. 2012; Kunduraci and Amatucci 2007). Sample Ni0.4 shows larger voltage gap between the two plateaus at ∼ 4.7 V due to its high level of cation ordering, as showed in the insert in Fig. 4 (Kunduraci and Amatucci 2007). A minor plateau in 4-V region is also observable, which originates from the reduction of Mn4+ to Mn3+ due to the presence of small amount of Mn3+ (Kunduraci and Amatucci 2006). The length of 4-V plateau in discharging curve can be used to calculate the relative amount of Mn3+ ions in LNMO (Xiao et al. 2012; Moorhead-Rosenberg et al. 2012). Therefore, the capacity percentages were calculated from the capacity between 3.8 and 4.2 V divided by the total discharge capacity to evaluate the amount of Mn3+ ions in spinel. The capacity percentages of 4 V plateau of Ni0.5, Ni0.45, and Ni0.4 are 5.4%, 6.0% and 8.5%, respectively. The amount of Mn3+ increases with the decreasing of Ni/Mn ratios. This is because part of Mn is reduced to 3+ to keep charge neutrality in materials.

Fig. 4
figure 4

Charge and discharge capacity curves at 0.2 C of LNMO samples

Full size image

Cyclic voltammetry is tested to study the oxidation/reduction behavior of LNMO samples as shown in Fig. 5. Samples show similar main peaks at 4.7 V ascribed to the Ni2+/Ni4+ redox couple. As shown, the gap between the oxidation peak and reduction peak of sample Ni0.4 is smaller than the others, implying good electrochemical reversibility. Besides, the small peaks at around 4.0 V are related to the Mn3+/Mn4+ redox couple.

Fig. 5
figure 5

Cyclic voltammetry curves of LNMO samples

Full size image

To evaluate the rate capabilities of the materials, the cells were cycled at current densities from 0.2 to 15 C. Figure 6 shows the specific capacities at different discharge rates. The discharge capacities decrease as the current density increases. At 15-C rate, sample Ni0.4 delivers discharge capacity of 106.1 mAh g−1, reaching 84.5% of that at 0.2-C rate, higher than 75.1% and 73.8% of Ni0.5 and Ni0.45. The best rate performance of Ni0.4 sample is attributed to its highest Mn3+ content. With the presence of Mn3+, samples have higher electronic and lithium-ion conductivity (Xiao et al. 2013; Kunduraci et al. 2006).

Fig. 6
figure 6

Rate capacity of LNMO samples

Full size image

Then, the cycling performance of LNMO was tested at 1 C for 500 cycles and 5 C for 1000 cycles as shown in Fig. 7. With the increasing of Ni/Mn ratios in spinel, their cycling performances are improved. Sample Ni0.5 shows the best cycling stability, keeping 94.4% and 90.4% of the highest discharge capacities after 500 cycles at 1 C and 1000 cycles at 5C. It is better than other recent report LiNi0.5Mn1.5O4 (Yin et al. 2018; Mou et al. 2018). The Mn3+ ion in LNMO tends to undergo a disproportionation reaction. The produced Mn2+ ion dissolves into the electrolyte, leading to a capacity fading during cycling (Park et al. 2011; Pieczonka et al. 2013). So, with the lowest Mn3+ contents, sample Ni0.5 shows the best cycling performance.

Fig. 7
figure 7

Cycling performances at different current densities of a 1 C and b 5 C of LNMO samples

Full size image

Conclusions

Samples LiNi0.5 − xMn1.5 + xO4 (x = 0, 0.05, 0.1) are synthesized by the simple co-precipitation method, in which NaHCO3 solution is poured into transition metal solution rapidly to produce precursor. It is very simple that temperature and pH conditions do not need control. With the decreasing of Ni/Mn ratios in spinel, their specific capacities are reduced and amount of Mn3+ is increased. Sample LiNi0.4Mn1.6O4 shows the best rate performance attributed to its high Mn3+ contents, which delivers a high discharge capacity of 106 mAh g−1 at 15 C, keeping 84.5% of that at 0.2 C rate. Cycling performance of sample is improved with the increasing Ni/Mn ratios in spinel. Sample LiNi0.5Mn1.5O4 shows the best cycling stability, keeping 94.4% and 90.4% of the highest discharge capacities after 500 cycles at 1 C and 1000 cycles at 5 C.