Introduction

Lithium-ion batteries (LIBs) have been widely used in mobile electronic devices in recent years and have been regarded as one of promising power source candidates for other fields such as electric vehicles and energy storage power station [1,2,3,4,5,6,7,8]. The demands for the next-generation LIBs have focus on the energy density, safety, and cost. Among the LIB cathode materials (LiMn2O4, LiCoO2, LiNiCoMnO2, LiFePO4, et al.), lithium manganese oxides have unique advantages such as safe, cheap, and low toxic and, therefore, can be considered the promising cathode material [9, 10]. Recently, Li-rich Mn-based layered compounds (Li2MnO3·LiMO2 or Li[M1−xLix]O2 (M = Mn, Ni, Co, etc.)), which are solid solutions between the monoclinic Li2MnO3 (C2/m symmetry) phase and trigonal LiMO2 (R\( \overline{3} \)m symmetry) phase, have attracted a significant amount of interest because of their encouraging high reversible capacity (> 200 mAh g−1) when cycled between 2.0~4.8 V [11,12,13,14]. Synthesizing technology has a great impact on the micromorphology and structure and subsequently affects the electrochemical performances. The hydroxide co-precipitation method was the usual synthesis method used to produce these materials [15, 16]. However, the metal Mn2+ ion can easily be oxidized to Mn3+ (MnOOH) or Mn4+ (MnO2) during the hydroxide co-precipitation process, therefore, resulting in inhomogeneous and impure phases in final product [17, 18]. Fortunately, in carbonate precipitation method, the oxidation state of the Mn ions is kept as 2+ in aqueous solution; thereby, it can be used for the synthesis of a more homogeneous and pure Li-rich Mn-based layered material with high electrochemical performance [19].

Co-precipitation of two or more metal ions homogeneous is a challenging process. Besides the polymetallic precipitation reactions, several other co-precipitation conditions such as ligand concentration, pH, stirring rate, and feeding sequence are important to the nucleation and growth of particles and, therefore, to the morphology of the final product [20,21,22,23]. However, no comprehensive study is available in the literature on the carbonate particle growth as a function of reaction time. The present research gives insights into the nucleation and growth mechanism of precursors prepared by the carbonate co-precipitation process in the presence of ethylene glycol. We experimentally monitored the nucleation and growth processes by analyzing the particle morphology and structure as a function of the reaction time. The structure, morphology, and electrochemical performance of the final cathode materials are also analyzed.

Experimental

All the reagents used in the present work were of guaranteed grade and used without further purification. Reagents used in this investigation included nickel sulfate hexahydrate (NiSO4·6H2O), manganese sulfate monohydrate (MnSO4·H2O), cobalt sulfate heptahydrate (CoSO4·7H2O), and ammonium bicarbonate (NH4HCO3), ethylene glycol (EG), and lithium carbonate (Li2CO3). The precursors were prepared as follows: an equal volume of 0.15 mol L−1 NH4HCO3 solution was added into a mixture of 0.15 mol L−1 MSO4 (Ni:Co:Mn 0.15:0.15:0.70) and an appropriate volume of ethylene glycol under vigorous agitation in the temperature of 55 °C. In the process of preparing precursors, the ammonium bicarbonate (NH4HCO3) solution serves as not only a reservoir for precipitating anions but also complexing agents. The precipitated powders were filtered and washed and then dried at 105 °C overnight. The obtained precursors were thoroughly mixed with Li2CO3 and then calcined at 950 °C for 10 h in air to obtain the final cathode material. The synthesis process is illustrated in Fig. 1. The pristine sample (without ethylene glycol) was also prepared for comparison.

Fig. 1
figure 1

Schematic illustration on the formation process of sample

Full size image

Powder X-ray diffraction (XRD) measurements of materials were carried out on a Rigaku 2500 X-ray diffractometer using Cu-Kα radiation. The diffraction data were collected over the range 10° < 2θ < 80°. The morphology of the synthesized samples was observed by a JEOL JSM-5600LV scanning electron microscopy (SEM).

The positive electrodes were prepared by mixing 80 wt% prepared powders, 10 wt% carbon conducting additive, and 10 wt% polyvinylidene difluoride (PVDF) binder in N-methyl-2pyrrolidone (NMP) solvent. Then, the slurry was cast onto a smooth aluminum foil and then dried. The laboratory half-cell was consisted of a positive electrodes (diameter = 14 mm) and a lithium metal anode separated by a Celgard 2400 polyethylene/polypropylene film and then with the electrolyte of 1 mol/L LiPF6 dissolved in EC/DMC/DEC (1:1:1 by volume). The half-cells were assembled and sealed in an argon-filled glove box and then tested on a LAND-CT2001A battery test system (Jinnuo Wuhan Co. Ltd., P.R. China) in the voltage range of 2.0–4.8 V at room temperature. Electrochemical impedance spectroscopy (EIS) was carried out on an electrochemical workstation (CHI660E, Shanghai Chenhua) in the frequency range from 0.1 MHz to 0.01 Hz.

Results and discussion

The crystals endured two steps during the carbonate precipitation: the formation of crystal nuclei and their growing up which includes each single crystal nucleus development and the agglomeration of some nuclei [17, 24]. The crystallization process could be controlled by a poor solvent [25,26,27]. In this work, ethylene glycol (EG) is employed to obtain crystals with the desired size and distribution. Figure 2(I) shows the SEM images of the precursors collected in the presence of ethylene glycol (EG precursors) with different reaction time. Primarily, small particles are produced and then combine with each other to form irregular-shaped and micron-sized loose agglomerates (Fig. 2(a) and (a’)) before 2.5 h. Thereafter, the particles grow gradually and continue to aggregate (Fig. 2(b) and (b’)). After 3.5 h, obvious segregation is detected within the agglomerations, and particles continued to grow with a round shape. Finally, the particles obtained after aging overnight are spherical with uniform size. The process of particle growth is shown in schematic illustration (Fig. 3). The precursor particles collected without the addition of ethylene glycol (Fig. 2(II)) are composed of large loose agglomerates with no defined shape. The possible reason is that some extra nucleation sites are available at the surface of the formed particles for further agglomeration without the dispersion of ethylene glycol.

Fig. 2
figure 2

SEM images of Ni0.15Co0.15Mn0.7CO3 collected at different reaction times. (I) EG-precursors (precursors collected in the presence of ethylene glycol), 2.5 h (a, a’), 3 h (b, b’), 3.5 h (c, c’), aging overnight (d, d’). (II) The pristine (precursors collected without ethylene glycol), 2.5 h (e, e’), aging overnight (f, f’)

Full size image
Fig. 3
figure 3

Schematic illustration for the overall formation process of microspheres of the EG-precursor

Full size image

The XRD patterns of precursors with different precipitation time are given in Fig. 4. The diffraction patterns of all samples can be indexed as MnCO3 (JCPDS no. 44-1472) with hexagonal structure (R\( \overline{3} \)c space group). The peaks of sample c and d are relatively broad which can be attributed to the small grain size of the precursor particles [28], in agreement with the morphology results. As can be seen from the precursors’ XRD results, the phase composition of the precursor does not change as the precipitation reaction proceeds, and the addition of the ethylene glycol in the precursor’s process does not affect the formation of the carbonate phase.

Fig. 4
figure 4

X-Ray diffraction patterns of Ni0.15Co0.15Mn0.7CO3 collected at different reaction times. a–d EG-precursors: 2.5 h (a), 3 h (b), 3.5 h (c), aging overnight (d). e, f The pristine: 2.5 h (e) and aging overnight (f)

Full size image

Figure 5 shows the SEM images of the final powders after calcination with Li2CO3 at 900 °C. After lithiation, the final Li1.2Ni0.12Co0.12Mn0.56O2 particles retain the morphology of the carbonate precursors even after being recrystallized with lithium sources (Fig. 5). Compared with the particles of the pristine sample (more than 1.5 μm in diameter) (Fig. 5b), the primary particle size of the EG-assisted sample is less than 0.5 μm (Fig. 5a), indicating good electronic ion conductivity. The morphology difference between two samples can be attributed to the usage of novel organic agent (ethylene glycol).

Fig. 5
figure 5

SEM images of the final powders. a EG-assisted sample. b The pristine

Full size image

Figure 6 describes the XRD patterns of final powders after calcination with Li2CO3 at 900 °C. All main diffraction patterns of the materials can be indexed as hexagonal α-NaFeO2 structure based on the R\( \overline{3} \)m space group, except for a broad peak around 21°, which is belonging to the Li2MnO3-like (C2/m space group) phase [29,30,31]. The clear splitting of the (006)/(102) and (108)/(110) peaks in the XRD patterns indicates that the material has a highly ordered layered structure [32]. The lattice parameters of the final materials are calculated by the Rietveld refinements based on the R\( \overline{3} \)m space group, and the values are summarized in Table 1. The c/a ratio of all samples is greater than 4.9, which indicates the formation of layered structure [32]. The intensity ratio (R) of I(003)/I(104) is sensitive to the cation distribution in the lattice. It is reported that the undesirable cation mixing would appear when R is smaller than 1.2 [33]. In our experiment, all samples have an R value higher than 1.2, which means that all samples have a well-ordered hexagonal structure without undesirable cation mixing. The EG-assisted sample exhibits a higher degree value (1.76) than the pristine one (1.37); it reveals that the former has a better layered structure than the latter, which implies that the former might have better electrochemical performance.

Fig. 6
figure 6

XRD patterns of the final powders. a EG-assisted sample. b The pristine

Full size image
Table 1 Lattice parameters of samples
Full size table

Figure 7 shows the initial charge-discharge curves of the Li/Li1.2Ni0.12Co0.12Mn0.56O2 at a current density of 12.5 mA g−1 (0.05 C). All charge curves are composed of a slope region below 4.5 V and a long plateau at ~ 4.5 V. The slope region is attributed to the extraction of Li+ ions from the layered LiNi0.3Co0.3Mn0.4O2 component corresponding to Ni2+/Ni4+ and Co3+/Co4+ reaction [34]. The long plateau is assigned to the removal of Li2O from the Li2MnO3 structure [34, 35]. This shows the typical electrochemistry characteristic of lithium- and manganese-rich composite material. As shown in Fig. 7, the discharge capacity of the sample prepared without using ethylene glycol (the pristine) is as low as 259.6 mAh g−1 when compared with the EG-assisted sample (272.5 mAh g−1). This improved capacity resulted from the addition of ethylene glycol during the carbonate precipitation, which facilitated the formation of small size particles with a uniform size distribution (which was confirmed by the SEM analyses), thereby improving the electrochemical performance.

Fig. 7
figure 7

The first charge-discharge profiles of the Li/Li1.2Ni0.12Co0.12Mn0.56O2 in a CR-2032 lithium cell. a EG-assisted sample. b The pristine

Full size image

Figure 8 shows the rate capabilities of EG-assisted sample and the pristine from 0.05 to 1 C between 2 and 4.8 V. It can be obviously observed that the EG-assisted sample exhibited higher rate capability than the pristine. These improved rate capability can be attributed to the short Li+ diffusion path with the small particle size for the EG-assisted sample. When the electrodes are cycled at high rates up to 1.0 C and then recovering back to 0.1 C, the capacities are returned to the initial value, which implied that the high rate cycling (0.2, 0.5, 1.0) did not have any adverse effect on the low rate (0.1 C) capacity. As can be seen, the reversible capacity of the EG-assisted sample is also 258.2 mAh g−1 after the 30th charge-discharge cycles with capacity retention of 97.2%, while the pristine one delivered 230.6 mAh g−1 with a capacity retention of 93.1% after 30 cycles. Compared to the EG-assisted sample, the large particles of the pristine possibly limited the Li+ diffusion and led to electrochemically inactivated core, further inflecting the capacity of the material.

Fig. 8
figure 8

Rate capabilities of EG-assisted sample (a) and the pristine (b) at 0.05, 0.1, 0.2, 0.5 and 1.0 C rate

Full size image

In order to comprehensively study the electrochemical properties of the materials, the electrochemical impedance spectroscopies (EIS) of both the EG-assisted sample and the pristine are measured, as shown in Fig. 9. These obtained plots are composed of a semicircle arc and a quasi-straight line, which are assigned to the charge-transfer reaction and the diffusion of Li+ through the bulk of the material, respectively. These plots are well fitted using the equivalent circuit model (Fig. 9 inset). In the equivalent circuit, Rs and Rct represent the resistance of the solution and charge-transfer resistance in the electrode/electrolyte interface, respectively, CPE indicates the double-layer capacitance, and Wo represents the diffusion-controlled Warburg impedance [36,37,38,39,40]. The EG-assisted sample shows a relatively small charge-transfer resistance (260.4 Ω) by comparison with the value of the pristine (308.4 Ω), which may benefit from the spherical and dispersed morphology.

Fig. 9
figure 9

Nyquist plots of EG-assisted sample (a) and the pristine (b)

Full size image

Conclusion

Mn-based carbonate precursor (Ni0.15Co0.15Mn0.7CO3) is synthesized by the carbonate precipitation in the presence of ethylene glycol. The nucleation and growth of particles are investigated during the carbonate process by monitoring particle morphologies and structures with reaction time. The additive of ethylene glycol shows significant effects on the morphology of the prepared Ni0.15Co0.15Mn0.7CO3 and the electrochemical performance of lithiated Li1.2Ni0.12Co0.12Mn0.56O2. The SEM results show that the particles of the precursors with the presence of ethylene glycol display more spherical and dispersed morphology than those prepared without using ethylene glycol. The final lithiated cathode materials prepared in the presence of ethylene glycol reveal an initial discharging capacity of 272.5 mAh g−1 and exhibited about 77.4% capacity retention with regard to the capacity obtained at 0.05 C rate when cycled at 1.0 C rate and more than 97.2% capacity retention after 30 cycles at the 0.1 C rate.