2.1. Oxalate Precursor Characterization
To see the results of the precipitates and detect the presence of impurities, X-ray diffraction was used to characterize oxalate precursors at an early stage.
Figure 1 depicts the entire diffraction pattern of the prepared samples. Both CoOX, MnOX, and NiOX are in orthorhombic crystal systems, as well as NMCOX. CoOX and NiOX are indexed with β-M”C
2O
4·2H
2O, while MnOX is denominated as γ-MnC
2O
4·2H
2O [
35,
36]. Therefore, all the peaks are similar to FeC
2O
4·2H
2O. Individual phases of NiOX, MnOX and CoOX are absent in NMCOX, corresponding to the homogeneous mixing and distribution of Ni-Co-Mn at an atomic level. However, the peaks in the NMCOX, which are less sharp than that of NiOX, indicated smaller crystal sizes and fluorescence effect by the presence of cobalt and manganese atoms [
36]. Testing samples with high amounts of cobalt might result in a pattern with high background noise due to fluorescence effect caused by interaction between X-rays from Cu with the cobalt-containing sample. Therefore, the XRD pattern of LCO and NMCOX111 show high noise. The lattice parameters for all samples can be seen in
Table 2. The a, b, and c values follow the values reported by Oh et al. [
37,
38]. Nickel oxalate and cobalt oxalate have similar lattice parameters, however, manganese oxalate has different values; thus, as the nickel content of NMC-oxalate increases, the lattice parameter values will be similar to nickel oxalate.
Figure 2 shows the mid-IR region (4000–400/cm) absorbance spectra of all the precursor samples as supporting data for the X-ray diffraction pattern. The absorbance of all the samples was the same. The bending vibrations of O-H are responsible for the broad peaks at around 3300–3500/cm and 1600/cm. The sharp peak located at about 1300/cm corresponds to the C-O bond [
39,
40,
41]. The peaks between 600 and 800/cm indicate C-O and O-C = O bonds. The presence of C-O and O-H peaks indicates that there was direct contact between the sample and air during preparation [
21]. The metal oxide (M-O) bond, which is in the metal oxalate phase, also has a peak around 490/cm [
42]. The FTIR results of the NMCOX samples are comparable to those of FeC
2O
4·2H
2O, indicating that the precipitation of NMC-oxalate occurs with the reaction given in Equations (1) and (2), as indicated by the X-ray diffraction results.
In the absorbance spectra test, it was discovered that the sample contains an O-H group in the form of water. Thermogravimetric analysis (TGA) was used to perform a thermal analysis to determine the amount of water in the sample.
Figure 3 shows weight loss through two stages, at temperatures around 200–250 °C and 320–400 °C. The weight loss at stage one is associated with dehydration of the precursor to anhydrous compound [
43]. Because the sample loses 19.7% of its weight, the precipitants formed are NMC-oxalate dihydrate and nickel oxalate dihydrate, which can be confirmed [
43,
44,
45]. Then, in stage two, the weight loss is 40% due to the decomposition of the NMC-oxalate to the NMC-oxide [
38]. The reactions in these two stages are shown in Equations (3) and (4) [
44]. This phenomenon is also explained in
Table 3, which shows the temperature of the dehydration and decomposition stages of the precursor, including those reported in previous studies. The calcination process began at a slightly different temperature for each sample, but the nickel content tended to differ. The calcination temperature drops as the nickel content (Ni-rich material) rises [
46].
2.2. Cathode Material’s Characterization
The lithiated and heated product from the coprecipitation process is then used. The final product in the cathode material is then characterized in order to determine the material that has been formed. The first test, as in the previous characterization, is X-ray diffraction, as shown in
Figure 4. X-ray Diffractometer (XRD) patterns of LCO, NMC, and LNO samples prepared after heat treatment showed a hexagonal structure, such as layered α-NaFeO
2 (space group: R3 m). All peaks conform to the JCPDS LiNiO
2 card with no observable phase impurities [
47,
48]. However, for NMC 333, LNO, LCO and LMO, Li
2CO
3 was detected in the material. To compensate the Li loss during high-temperature lithiation (at 800 °C), an excess of Li source was added. In samples with high content of Co and Mn, which usually form a layered structure at high temperatures, it can be predicted that the formation of layer-structured material is not completed due to nonoptimal sintering temperature. On the contrary, the formation of Ni-rich, especially Ni-based, cathode material, is optimal in lower temperature. Therefore, at 800 °C, the formation of rock-salt NiO was favorable, and this hampered the lithiation process which predicted the unsuccessful lithiation process. These phenomena cause the formation of Li
2CO
3 phase on the sample. Meanwhile, LMO shows a cubic spinel structure with space group Fd3m (the lattice parameter a = b = c) [
49]. The double peak at (006)/(102) and (018)/(110) corresponds to a well-ordered layered structure [
22,
23]. Especially for NMC, a double peak can be clearly seen at 2θ 38.4°–38.5° for (006)/(102) and 64.5°–65.5° for (018)/(110).
The lattice parameter and the R value presented in
Table 4 can be used to evaluate the well-ordered, layered material. The lattice parameters of each sample were calculated using the least-squares regression method [
21]. The c/a value of the lattice parameter indicates the crystallization of the material. The ideal value of c/a is above 4.899 [
50]. Both NMC and LNO samples had values above the ideal c/a ratio, which meant that the samples had a well-ordered layered structure. The relative intensity of certain peaks in XRD indicates the degree of antisite interference of Ni
2+ and Li
+. Furthermore, a good hexagonal ordering of the lattice has an R value below 0.45 [
50]. All samples had R values less than 0.45, with the exception of NMC 811. Electrochemical performance can be improved by lowering antisite interference values [
51].
Calculating the intensity ratio (IR) can reveal the presence of cation mixing, which is common in layered transition-metal oxide cathode materials. IR for LCO, NMC, and LNO uses peaks (003) and (104), while for LMO it uses peaks (111) and (311), and the results are listed in
Table 4. The IR value of each sample is more than 1.2, except for LCO, which has a low degree of cation mixing [
52]. The occurrence of cation mixing mostly occurs at higher Ni content. In the transition-metal lattice, the Ni
2+ will substitute the Li-ion due to their similarity in radii size [
53]. In this case, the IR value does not significantly change as the Ni content increases. This could be due to the presence of Co and Mn, which are still capable of allowing NMC samples to be structurally stabilized.
Figure 5 shows an FT-IR spectrum analysis of the presence of anionic impurities on the surface of cathode material samples. On the basis of spectra, the carbonate compound is only visible (~900/cm and ~1400/cm). This can be attributed to Li
2CO
3 as a result of excess Li source. The peak should indicate the oxide phase only, but there is another phase that can be observed that is not attributed to the metal oxide phase. These phase is the result of the absorption of molecules in the atmosphere in the sample. During the heating process, Li
2CO
3 is melted and decomposed, releasing CO
2. However, residual Li was converted back into Li
2CO
3 as the sintering process was finished [
21]. The equation depicts the formation of Li
2CO
3 on the surface of samples (5,6). At NMC 333 and LNO, where the peak is sharp and clear, these spectra support the suspicion of impurity. Since the presence of Li
2CO
3 is considered low, it can be neglected.
At temperatures of 46–200 °C and 250–400 °C,
Figure 6 depicts weight loss in two stages. The loss of adsorbed water on the material’s surface and the intercalation of several water and alcohol molecules are responsible for 9.8% of the weight loss in stage one. In stage two, the weight loss of 25% is due to the reaction between lithium oxide, nickel oxide, manganese oxide, and cobalt oxide. Then, the sample weight starts to be constant at ~450 °C. This weight loss can confirm that the formation of NMC crystals has occurred [
54]. It is also reinforced by the DSC curve in
Figure S1, to determine the optimal calcination temperature. Three peaks between 60–360 °C indicate an exothermic state that affects the loss of water in the material due to evaporation [
55]. To confirm the TG curve, the lithiation process begins at 460 °C. Based on the curve, the final mass is consistent with the theoretical value of ~42%.
Scanning Electron Microscope (SEM) testing was used to determine the morphology of the sample after the heating process as shown in
Figure 7. Most samples have nanosized-to-submicron-sized primary particles clustered together to form microsized secondary particles. Compared to before the heating and lithiation process, the particle shape did not change significantly, for example in NMCOX 622 and NMCOX 811 in
Figure S2. Small agglomerated cubes were found in the NMC 622, NMC 811, and LCO samples. The particles in NMC samples 333, 424, 442, 523, and 532 were quasispherical. LNO has a chunk morphology and LMO has an irregular shape. All samples have a primary particle size ≤1.5 µm, while for secondary particles, the variations are presented in
Table 5.
Table 6 presents the XRF result of the as-prepared samples for NMC. This XRF data is based on three atoms, namely Ni, Mn and Co. The nickel content of NMC 333, NMC 523, NMC 532, and NMC 622 is a little far from the expected nickel content. Meanwhile, NMC 424, NMC 442, NMC 622 and NMC 811 are close to the expected values. Similarly, except for the NMC 622 sample, which had a lower Co content, the Mn and Co content in all samples was close to the expected value.
2.3. Electrochemical Performance Test
A cylindrical battery cell was used to test the NMC cathode material electrochemically. As the anode, mesocarbon microbeads (MCMB) (MTI, America) were used to test the charge–discharge analysis. Before grading, the formation cycle of the NMC battery sample was completed; data are not shown. Then, the specific capacity is determined after three cycles. The charge–discharge curves created using a current density of 0.5 C are depicted in
Figure 8.
The specific discharge capacity of NMC 333, NMC 424, NMC 442, NMC 523, NMC 532, NMC 622 and NMC811 is 120.20 mAh/g, 129.43 mAh/g, 127.30 mAh/g, 140.04 mAh/g, 134.22 mAh/g, 153.60 mAh/g and 106.30 mAh/g, respectively, as shown in
Figure 8. NMC622 delivers a higher specific discharge capacity due to its high nickel content; meanwhile, NMC811 needs different synthesis conditions. NMC cathode material has active electrochemical ions, Ni
2+ and Co
3+. The oxidation of Ni
2+ must follow charge compensation during lithium extraction (i.e., charging) to Ni
3+ and Ni
4+, and Co
3+ to Co
4+, respectively [
56]. Since the Ni
3+ and Ni
4+ ions are converted to Ni
2+ ions when heated, oxygen release from the structure is unavoidable to maintain charge neutrality. The temperature ranges for the LiMn
2O
4-type and Mn
3O
4-type spinel for each sample are similar to the temperature range for oxygen release spreading. The lower the onset temperature of oxygen release and the tighter the oxygen distribution, the higher the nickel content [
57]. Thus, NMC811 needs an extra oxygen environment in the heating process. The same method is used to make LCO, which has a specific capacity of 92 mAh/g (data not shown). As a result, when compared to LCO, it can be concluded that the sample performs exceptionally well.
Using a standard research protocol,
Figure 9 depicts the rate ability of various NMC samples. The real specific capacity (C’) rather than the theoretical capacity (C) is used to determine the charging and discharging rate. After three cycles of the formation process (not shown), the cells were charged at 0.5 C’ and discharged at different discharging rates. Even after being discharged at high currents, the graph shows that all samples have a small capacity reduction, as reported by Schmidt et al. [
58]. This behavior indicates that the prepared sample has a good cycling performance. Since NMC622 has a higher specific discharge capacity,
Figure 10b shows its cycling behavior for 100 cycles at 1 C’ rates, and its capacity retention of 70.9%. The cycling performance of NMC532 and NMC442 are also shown in
Figure 10a and have capacity retention of 86.1% and 72.6%, respectively. More research is needed to obtain better and more stable capacity and lower capacity retention.
NMC622 is a highly recognizable and well-studied material due to its excellent electrochemical performance. Several studies on NMC622 and its electrochemical performance are listed in
Table 7. Compared to full-cell, half-cell electrochemical studies are reported more frequently. However, this study uses full cells because they are closer to commercial applications. This research can be a reference for the development of a simpler, faster, and more effective synthesis of NMC, especially NMC622, in the future.
2.4. Postmortem Analysis
To determine the state of the material after the cycle, a postmortem analysis was performed. Cycling causes capacity degradation, as shown in
Figure 10. Loss of active material due to structural degradation, loss of reversible lithium, and increased resistance, all contribute to capacity degradation [
64]. The structural degradation can be seen through XRD and FTIR analysis presented in
Figure 11. In
Figure 11a, impurities in the form of Li
2CO
3 are seen at 2θ 27.8°. The formation of Li
2CO
3 occurs due to a side reaction between the electrolyte and the cathode material. Furthermore, due to a decrease in crystallinite, the postcycle sample appears to be wider than the precycle sample [
21]. The emergence of Li
2CO
3 impurities is also supported by the FTIR analysis shown in
Figure 11b. FTIR spectra show Li
2CO
3 at wavenumbers 838/cm, 1397/cm and 1480/cm. ROCO
2Li (1634/cm) and ROLi (1073/cm) were also found on the surface of the material. The large amount of Li
2CO
3 formed causes a large reversible Li loss [
64]. In terms of morphology, the presence of a solid electrolyte interphase layer covering the particle surface can be seen in postcycle in
Figure 12. The decomposition of organic electrolytes results in the formation of this layer [
65].