English

• Layered oxide LiNi_1-x-yCo_xMn_yO₂ (NCM) is widely studied for lithium-ion battery cathode materials due to its high energy density, cost-effectiveness, and favorable low-temperature performance [1,2]. Enhancing NCM cathode materials performance remains a major goal, pursued through strategies such as doping [3,4], coatings [5,6], cobalt-free compositions [7,8], high-nickel formulations [9,10], single-crystal structures [11,12], and other modifications. However, the current manufacture of NCM cathode materials is primarily based on wet co-precipitation, which usually involves complex, multi-step procedures including mixing, separation, purification, and drying, requires substantial equipment investment, has long production times, and generates liquid by-products [13-17]. Recently, flame-assisted spray pyrolysis (FASP) has gained attention as a promising alternative for preparing NCM cathode materials due to its fewer steps, reduced costs achieved by utilizing low-cost salts and minimizing equipment usage, as well as enhanced environmental friendliness through the elimination of waste liquid by-products [18-20].

  FASP has been successfully applied to prepare various NCM cathode materials. Ogihara et al. [21] prepared NCM111 via FASP and conducted scale-up experiments. Abram et al. [22] utilized FASP to prepare NCM111 and NCM811, examining the influence of preheating temperature and flame synthesis temperature on electrochemical performance. Yan et al. [23] used the same setup to investigate the Dy doping concentration effects on electrochemical properties and thermal stability of NCM811. Zhang et al. [24] demonstrated that urea addition during FASP synthesis of NCM811 significantly reduces required annealed time. In another study, Zhang et al. [25] established that appropriate preheating temperature helps maintain structural and electrochemical stability of NCM811 cathode materials. The above analyses show that FASP is capable of preparing NCM cathode materials. Current studies on utilizing FASP for preparing NCM cathode materials primarily focus on elucidating the effects of operational parameters such as additives [24], preheating temperature [22,25], synthesis temperature [26], annealed temperature [27], and dopant [23] on the structural, morphological, and electrochemical performance.

  However, the in-depth understanding of mechanism involving FASP for preparing NCM cathode materials remains inadequate, despite its critical importance. Typically, in FASP, the fine liquid droplets containing the precursor(s) are introduced into a flame (with or without preheating), where evaporation, decomposition, nucleation, growth, and agglomeration are sequentially occurring, ultimately yielding nanostructured particles. Comparing to the earlier flame aerosol pyrolysis synthesized metal oxide nanostructured materials, for instance, SiO₂ [28], TiO₂ [29], ZrO₂ [30], are relatively simple (though not easy), which usually involved a single metal element precursor dissolved in solution that was then converted to stable oxide by hydrolysis and thermal decomposition in the flame [31,32]. On the contrary, in the FASP synthesis process of NCM cathode materials, the precursor solution consists of a mixture of multiple metal elements, by dissolving the salts of Li, Ni, Co, and Mn [33-35]. Crucially, these four metal elements exhibit significantly different physicochemical properties, including solubility, diffusion, and precipitation behavior. For example, the solubility values of LiNO₃, Ni(NO₃)₂, Co(NO₃)₂, and Mn(NO₃)₂ in 100 mL of water at 298 K are 1.48, 0.54, 0.56, and 0.90 mol, respectively. These differences significantly impact the FASP process. Abram et al. [22] observed that the low solubility and high concentration of nickel nitrate led to the easier formation of irregular nanoparticles from NCM811. Using molecular dynamics simulations, Hou et al. [33] demonstrated that during evaporation of Li, Ni, Co, and Mn nitrate solutions, Ni, Co, and Mn nitrates preferentially deposit at the droplet surface, generating core-shell structures, while Li distributes more evenly throughout the evaporating droplet due to its higher diffusivity. Further demonstrating the complexity, Bhat et al. [34] found that evaporating droplets containing Li, Ni, Co, and Mn multiple precursors yielded particles with radial elemental gradients, 2.5 wt% urea decomposition reaction perturbed the evaporation process, leading to particles with a uniform elemental distribution.

  Furthermore, the distinct chemical reactivities and speciation of precursor metal elements significantly influence FASP processes. Lithium, as an alkali metal, exhibits high reactivity, and its salts undergo decomposition and readily react with flame components. In contrast, first-row transition metals (Ni, Co, Mn) display lower reactivity, forming stable oxides. This reactivity disparity among these four metal components may bring in great complex variations in the flame. While Zhang et al. [24] observed lithium compound nanoparticle clusters on precursor surfaces during NCM811 synthesis via FASP, though the chemical forms of Li, Ni, Co, and Mn were not yet clarified. Lithium speciation exemplifies this complexity, with multiple possible form in the flame. Studies confirm lithium salts critically impact NCM cathode materials properties. Yang et al. [36] demonstrated strong dependence of NCM111 morphology, density, and performance on lithium salt selection. Kong et al. [37] identified Li₂CO₃ as particularly suitable for NCM523 synthesis. Zhu et al. [38] achieved significant reduction in single-crystal NCM811 annealed temperature using LiNO₃. The above studies show that lithium salts significantly affect the properties of NCM cathode materials. However, the specific role of Li on NCM cathode materials during FASP is not clear.

  The main aim of the present study is to clarify the transformation routes of Li, Ni, Co, and Mn elements in multi-component precursor solutions in the flame and evaluate their impact on NCM cathode materials. The present study first employed thermogravimetric analysis and thermodynamic calculations to determine decomposition temperatures of Li, Ni, Co, and Mn nitrates. Their decomposition products in the flame were characterized to determine their forms, morphologies and transformation routes. Subsequently, NCM111 cathode materials with three different lithium salts (LiNO₃-NCM111, Li₂CO₃-NCM111, and LiOH—NCM111) were prepared via a two-step procedure using FASP, and comprehensive characterized in structure, morphology, and electrochemical performance to clarify the effect of lithium salts.

  Through the experimental procedure that was shown schematic of FASP in Fig. S1 (Supporting information), NCM111 precursors were synthesized both with and without lithium. Figs. 1a–d present comparative SEM images, revealing a key morphological distinction: NCM111 precursors with lithium exhibit significant surface roughness due to abundant nanoparticle aggregates (Figs. 1a and b), whereas NCM111 precursors without lithium show relatively smooth surfaces with minimal nanoparticle enrichment (Figs. 1c and d). Given that LiNO₃ presence in the raw material solution is the sole variable, these surfaces aggregates likely represent lithium compounds. A similar phenomenon has been reported in related literature [24], though their specific form of lithium remained undetermined. XRD analysis (Fig. 1e) further elucidates this difference. Both precursors crystallize in rock-salt (analogous to NiO) and spinel (resembling NiCo₂O₄ and NiMn₂O₄) structures. Critically, the precursor with lithium displays distinct Li₂CO₃ characteristic peaks. This indicates that the lithium particle clusters on the surface of the NCM111 precursor with lithium may exist primarily as Li₂CO₃. This may be due to lithium being an alkali metal, which is more active and may react not only during thermal decomposition but also with other components in the flame. In contrast, Ni, Co, and Mn, as first-row transition metal elements, are more stable and remain as stable oxides after thermal decomposition, similar to the flame aerosol pyrolysis synthesis of SiO₂ [28], TiO₂ [29], and ZrO₂ [30] from single-metal element precursor solutions.

  Figure 1

  

  Figure 1.  SEM images: (a, b) NCM111 precursor with lithium, (c, d) NCM111 precursor without lithium. (e) XRD patterns of NCM111 precursor with lithium and without lithium.

  DownLoad: Full-Size Img PowerPoint

  Under the high temperatures of the flame, the hydrated raw metal salts Ni(NO₃)₂·6H₂O, Co(NO₃)₂·6H₂O, and Mn(NO₃)₂·4H₂O undergo thermal dehydration to form anhydrous nitrates, and the following possible decomposition reactions.

  $ M\left(\mathrm{NO}_3\right)_a=M \mathrm{O}_{a / 2}+a \mathrm{NO}(\mathrm{~g})+3 / 4 a \mathrm{O}_2(\mathrm{~g}) $

  $ M\left(\mathrm{NO}_3\right)_a=M \mathrm{O}_{a / 2}+a \mathrm{NO}_2(\mathrm{~g})+1 / 4 a \mathrm{O}_2(\mathrm{~g}) $

  $ M\left(\mathrm{NO}_3\right)_a=M \mathrm{O}_a+a \mathrm{NO}(\mathrm{~g})+1 / 2 a \mathrm{O}_2(\mathrm{~g}) $

  $ M\left(\mathrm{NO}_3\right)_a=M \mathrm{O}_a+a \mathrm{NO}_2(\mathrm{~g}) $

  M is Li, Ni, Co or Mn.

  The specific decomposition reactions and temperatures of the four metal nitrates were determined by calculating the Gibbs free energy changes (∆G) and performing thermogravimetric analysis (TGA). Fig. 2 compares experimental thermogravimetric curves (red) with theoretical ∆G calculations versus temperature (black), alongside the derivative thermogravimetric (DTG) curve (blue) used to obtain the onset and end of the thermal decomposition temperature. The decomposition reactions of the four nitrates are significantly different. The decomposition reactions of LiNO₃ (1), Ni(NO₃)₂ and Co(NO₃)₂ (2), and Mn(NO₃)₂ (3) followed the same trend as the thermogravimetric curves. As summarized in Table 1, theoretical decomposition temperatures consistently fell between experimental onset and end values. All nitrates have decomposition temperatures below the flame temperature of 1780 ℃, as measured using a type-B thermocouple, guaranteeing complete decomposed to their respective metal oxides. Particularly, the decomposition temperature of LiNO₃ (660 ℃) significantly exceeds that of Ni, Co, and Mn nitrates (below 300 ℃). This distinct thermal behavior necessitates separate consideration of transformation routes: Ni, Co, and Mn nitrates can be treated as a group, while LiNO₃ requires independent analysis in the flame.

  Figure 2

  

  Figure 2.  Experimental TGA curves, DTG curves and theoretical ∆G calculations of the raw nitrate salts: (a) LiNO₃, (b) Ni(NO₃)₂·6H₂O, (c) Co(NO₃)₂·6H₂O, (d) Mn(NO₃)₂·4H₂O.

  DownLoad: Full-Size Img PowerPoint

  Table 1

  

  Table 1.  Theoretical (∆G < 0) and experimental decomposition temperatures of LiNO₃, Ni(NO₃)₂·6H₂O, Co(NO₃)₂·6H₂O, Mn(NO₃)₂·4H₂O.

  DownLoad: CSV

  Sample	Theoretical decomposition temperature (℃)	Experimental onset decomposition temperature (℃)	Experimental end decomposition temperature (℃)
  LiNO3	660.0	506.1	696.2
  Ni(NO3)2·6H2O	264.0	268.3	345.1
  Co(NO3)2·6H2O	275.0	223.9	300.8
  Mn(NO3)2·4H2O	180.0	186.8	255.8

  TGA confirms that Li, Ni, Co, and Mn nitrates decomposed completely in the flame, yielding their respective oxides as the primary decomposition products. This section delves into the subsequent transformation routes of these four elements, Li, Ni, Co, and Mn, commencing from their oxide states. The present study utilizes a non-premixed CH_4nullO₂ flame, producing predominantly CO₂ and H₂O as combustion products. The potential reactions involving Li₂O, NiO, CoO, and MnO₂ with CO₂ and H₂O are outlined as follows:

  $ M \mathrm{O}_a+a \mathrm{H}_2 \mathrm{O}(\mathrm{~g})=M(\mathrm{OH})_{2 a} $

  $ M \mathrm{O}_a+a \mathrm{CO}_2(\mathrm{~g})=M\left(\mathrm{CO}_3\right)_a $

  $ M(\mathrm{OH})_{2 a}+a \mathrm{CO}_2(\mathrm{~g})=M\left(\mathrm{CO}_3\right)_a+a \mathrm{H}_2 \mathrm{O}(\mathrm{~g}) $

  $ M \mathrm{O}_{2 a}+a \mathrm{H}_2 \mathrm{O}(\mathrm{~g})=M(\mathrm{OH})_{2 a}+1 / 2 a \mathrm{O}_2(\mathrm{~g}) $

  $ M \mathrm{O}_{2 a}+a \mathrm{CO}_2(\mathrm{~g})=M\left(\mathrm{CO}_3\right)_a+1 / 2 a \mathrm{O}_2(\mathrm{~g}) $

  M is Li, Ni, Co or Mn.

  Fig. 3a illustrates the hypothesized transformation route of the lithium element in the flame. As shown in reaction (1), LiNO₃ decomposes thermally into Li₂O and NO. Consequently, the possible reactions (5–9) involving Li₂O as a reactant were proposed. Fig. 3b presents the calculated ∆G for the above reactions from 0 ℃ to 2000 ℃, indicating thermodynamic feasibility (ΔG < 0) for all reactions below approximately 1050 ℃. CO₂ and H₂O, produced by the combustion of methane, can react with Li₂O, to form Li₂CO₃ and LiOH, respectively [39]. To validate the accuracy of the transformation route LiNO₃ in the flame, FASP and spray pyrolysis experiments were conducted using neat LiNO₃ solutions. Figs. 3c and d and Fig. S4a (Supporting information) present the characterization results obtained under the same conditions as the NCM111 precursors with and without lithium, with carrier gas (N₂), CH₄, and O₂ flow rates of 1.5, 0.25, and 1.00 L/min, respectively. Fig. 3c shows the XRD pattern where only the characteristic peak of Li₂CO₃ is present, indicating that Li₂O is completely converted to Li₂CO₃. The SEM images in Fig. 3d and Fig. S4a reveal that the Li₂CO₃ particle size is significantly smaller than that of NCM111 precursors, matching the size of nanoclusters observed on NCM111 precursors with lithium, suggesting these nanoclusters are Li₂CO₃. When the dilution gas (N₂ at 2.25 L/min) was introduced into the CH₄ line to lower the flame temperature, the XRD patterns of the yielded products (Fig. 3e) were dominated by Li₂CO₃ characteristic peaks, with LiNO₃ lower intensity characteristic peaks also present. This indicates that even at reduced temperatures, Li₂O can still be converted to Li₂CO₃. Minor residual LiNO₃ results from incomplete thermal decomposition at lower temperature. These experimental findings essentially confirm the transformation route of LiNO₃-Li₂O-Li₂CO₃, corresponding to reaction (6). So does LiOH, the intermediate product of the transformation route of LiNO₃ in the flame (Fig. 3a), exist? In the CH_4nullO₂ flame, where there is a large amount of CO₂, LiOH would react rapidly with CO₂, making it difficult to detect in the products of FASP. To address this challenge, we designed a spray pyrolysis experiment. By using N₂ as the carrier gas, we avoided contact with CO₂. We fed neat LiNO₃ aerosol into a tube furnace set at 680 ℃ to ensure the pyrolysis of LiNO₃ (Fig. S2 in Supporting information). Fig. 3f shows the XRD pattern of the product from this experiment. Prominent LiOH peaks and weak LiNO₃ peaks are visible, indicating that Li₂O reacts with H₂O to form LiOH. This confirms the transformation route of LiNO₃-Li₂O-LiOH, corresponding to reactions (5) and (7).

  Figure 3

  

  Figure 3.  (a) Transformation route of LiNO₃ in the flame. (b) ΔG variation with temperature for reactions of Li₂O with CO₂ or H₂O. (c) XRD pattern of LiNO₃-N₂–0.00. (d) SEM image of LiNO₃-N₂–0.00. (e) XRD pattern of LiNO₃-N₂–2.25. (f) XRD pattern of LiNO₃-spray pyrolysis.

  DownLoad: Full-Size Img PowerPoint

  Fig. 4a illustrates the ΔG varies with temperature for NiO reactions with CO₂ and H₂O. For reaction (5) of NiO, ΔG > 0 across the temperature range, indicating thermodynamic infeasibility. While NiO reaction (6) exhibits ΔG < 0 below 100 ℃, the low reaction temperature relative to the Li₂O results in negligible reaction kinetics, preventing equilibrium attainment. A similar trend is observed for CoO (Fig. 4b), although reactions (5) and (6) exhibit ΔG < 0 below approximately 180 ℃, kinetic limitations preclude significant reaction progress. For MnO₂ (Fig. 4c), ΔG > 0 for reactions (8) and (9) confirms thermodynamic non-spontaneity. FASP experiments were conducted for the neat Ni, Co, and Mn nitrates with (N₂ = 2.25 L/min) and without (N₂ = 0.00 L/min) diluent gas. The XRD patterns of the products are displayed in Figs. 4d-f. Specifically, Ni(NO₃)₂ decomposes to NiO, which remains thermally stable in the flame without further transformation. Co(NO₃)₂ decomposes to cobalt(Ⅱ) oxide (CoO), which undergoes partial oxidation to form cobalt(Ⅲ) oxide (Co₃O₄) under the high-temperature flame conditions. Mn(NO₃)₂ initially decomposes to manganese(Ⅳ) oxide (MnO₂), followed by its reduction to manganese(Ⅲ) oxide (Mn₂O₃) and ultimately to manganese (Ⅱ, Ⅲ) oxide (Mn₃O₄). A summary of the transformation routes of the three nitrates (Ni, Co, and Mn) obtained in the experiment is shown in Fig. S3 (Supporting information). Critically, these experiments confirm that the thermal decomposition of the Ni, Co, and Mn nitrates leads to non-reactive oxides. This is corroborated by the XRD pattern of the NCM111 precursor without lithium (Fig. 1e), which shows only complex oxide phases without carbonate or hydroxide signatures. Figs. 4g–i and Figs. S4b-d (Supporting information) show the SEM images of Ni, Co, and Mn nitrate FASP products. Co(NO₃)₂ and Mn(NO₃)₂ products have smoother surfaces, which is consistent with the SEM results of the NCM111 precursor without lithium, supporting the conclusion that Co and Mn oxides dominate the morphology of NCM111 precursor without lithium.

  Figure 4

  

  Figure 4.  The ΔG variation with temperature for reactions of Ni, Co, and Mn oxides with CO₂ or H₂O: (a) NiO, (b) CoO, (c) MnO₂. XRD pattern of (d) Ni(NO₃)₂-N₂–0.00 and Ni(NO₃)₂-N₂–2.25, (e) Co(NO₃)₂-N₂–0.00 and Co(NO₃)₂-N₂–2.25, (f) Mn(NO₃)₂-N₂–0.00 and Mn(NO₃)₂-N₂–2.25. SEM image of (g) Ni(NO₃)₂-N₂–0.00, (h) Co(NO₃)₂-N₂–0.00, (i) Mn(NO₃)₂-N₂–0.00.

  DownLoad: Full-Size Img PowerPoint

  In summary, the transformation routes of the four nitrates can be categorized into two groups: Ni, Co, and Mn nitrates decompose at lower temperatures (<300 ℃), producing non-reactive composite oxides; While LiNO₃ decomposes at a significantly higher temperature (660 ℃), producing reactive Li₂O, eventually forming LiOH and Li₂CO₃ under the environment of CO₂ and H₂O, as well as the unconverted LiNO₃. Different lithium salts have distinct physicochemical properties, such as melting point, decomposition temperature, and chemical composition, critically governing NCM cathode material structure and electrochemical performance. The influence of lithium on the NCM material properties is evident not only in the post-annealing treatment but also during the FASP of precursor. Since the carbonate ions (CO₃^2-) in Li₂CO₃ and hydroxide ions (OH^-) in LiOH can precipitate with metal ions like Ni²⁺, obtaining a well-mixed precursor solution is difficult [40]. To isolate and elucidate the specific role of lithium, the present study prepared Ni, Co, and Mn oxide precursor, added lithium salts (LiNO₃, Li₂CO₃, LiOH·H₂O), and explored their impact on NCM cathode material properties during FASP.

  Figs. 5a–c and Figs. S5a–c (Supporting information) show that Li₂CO₃-NCM111 and LiOH—NCM111 maintain the spherical structure with significantly larger primary particles, whereas the LiNO₃-NCM111 tends to agglomerate, forming irregular particles with smaller primary particles. This morphological divergence persists after post-annealing in Fig. S6 (Supporting information), LiNO₃-NCM111 solidified, whereas Li₂CO₃-NCM111 and LiOH—NCM111 retains the powder states. The more uniform spherical morphology and larger primary particles of Li₂CO₃-NCM111 and LiOH—NCM111 suggest better electrochemical performance compared to LiNO₃-NCM111. The elemental distribution in Figs. 5d–f and Figs. S5d–f (Supporting information) for NCM111 cathode materials prepared with different lithium salts are also homogeneous, and the elemental ratios listed in Table S1 (Supporting information), Ni:Co:Mn ≈ 0.33:0.33:0.33, are consistent with the NCM111 precursor without lithium, indicating no significant elemental segregation during post-annealing treatment. Fig. 5g presents the XRD patterns of NCM111 cathode materials prepared with different lithium salts. All the samples adopt the α-NaFeO₂ layered structure (space group R-3m, PDF #06–0063), with sharp diffraction peaks indicating good crystallinity [12]. No extraneous phases were observed, demonstrating material purity. Clear (006)/(102) and (108)/(110) peak splitting confirms a well-ordered layered structure. This splitting is more distinct in Li₂CO₃-NCM111 and LiOH—NCM111 than in LiNO₃-NCM111. (003)/(104) peak intensity ratios (Table 2) assess Li/Ni cation disorder in layered structures. I₍₀₀₃₎/I₍₁₀₄₎ > 1.2 indicates low cation disorder and developed layered structure [41]. LiOH—NCM111 exhibits the highest ratio (1.356), signifying the lowest cation disorder degree. To evaluate precise crystal structure information, Rietveld refinement in GSAS-Ⅱ was performed, the results are shown in Fig. S7 (Supporting information) and Table 2 [42]. The c/a ratio typically characterizes triangular distortion. The differences in the values of a and c are small for all samples, leading to minimal variation in the c/a ratio, which indicates that the lithium salts do not significantly influence this parameter [43].

  Figure 5

  

  Figure 5.  SEM images of the NCM111 cathode materials of (a) LiNO₃-NCM111, (b) Li₂CO₃-NCM111, and (c) LiOH–NCM111. EDS mapping images of the NCM111 cathode materials of (d) LiNO₃-NCM111, (e) Li₂CO₃-NCM111, and (f) LiOH–NCM111. (g) XRD patterns of NCM111 prepared with different lithium salts.

  DownLoad: Full-Size Img PowerPoint

  Table 2

  

  Table 2.  Structural parameters obtained from the Rietveld refinement of XRD patterns.

  DownLoad: CSV

  Sample	a (Å)	c (Å)	c/a	I(003)/I(104)	Rp (%)	Rwp (%)
  LiNO3-NCM111	2.8631	14.2495	4.9769	1.290	5.64	3.46
  Li2CO3-NCM111	2.8624	14.2418	4.9754	1.321	4.84	3.14
  LiOH–NCM111	2.8621	14.2418	4.9760	1.356	4.78	3.16

  The initial charge/discharge curves of NCM111 cathode materials prepared with different lithium salts at 0.1 C rate (1 C = 160 mA/g) between 2.8 V and 4.3 V are shown in Fig. 6a. The initial discharge specific capacities and Coulombic efficiencies are 161.5 mAh/g and 90.78% for Li₂CO₃-NCM111, and 162.6 mAh/g and 90.13% for LiOH—NCM111. In contrast, LiNO₃-NCM111 exhibits significantly lower values of 158.4 mAh/g and 90.57%. The melting points of LiOH, Li₂CO₃, and LiNO₃ are 471, 618, and 264 ℃, respectively. The lower initial discharge capacity of LiNO₃-NCM111 is attributed to lithium loss during annealing, resulting from its low melting point. The rate performance of the NCM111 cathode materials at 0.1–1 C, is compared in Fig. 6b. At a current density of 1 C, the discharge specific capacities of Li₂CO₃-NCM111 and LiOH—NCM111 were 134.7 mAh/g and 130.1 mAh/g respectively, while that of LiNO₃-NCM111 was significantly lower at 95.6 mAh/g. Fig. 6c shows the cycling performance of NCM111 cathode materials at 0.33 C rate between 2.8 V and 4.3 V at 25 ℃. After 200 cycles, the discharge specific capacities of Li₂CO₃-NCM111 (125.0 mAh/g) and LiOH—NCM111 (129.8 mAh/g) are significantly higher than that of LiNO₃-NCM111 (118.7 mAh/g), and similar trends are observed in capacity retention. The Coulombic efficiencies of LiNO₃-NCM111, Li₂CO₃-NCM111, and LiOH—NCM111 are lower during the first 3 cycles, likely due to residual alkali on the sample surfaces, as these three samples have not undergone washing.

  Figure 6

  

  Figure 6.  Electrochemical performance of NCM111 cathode materials prepared with different lithium salts: (a) Initial charge/discharge curves (0.1 C, 1 C = 160 mA/g). (b) Rate performance from 0.1 C to 1 C. (c) Cycling performance at 0.33 C within a potential range of 2.8–4.3 V at 25 ℃.

  DownLoad: Full-Size Img PowerPoint

  Table 3 summarizes the electrochemical performances of NCM111 cathode materials, with LiOH—NCM111 exhibiting superior electrochemical performance. This results from its well-ordered layered structures, very low degree of Li/Ni disorder, uniform elemental distribution, stable spherical structures, and larger primary particles as detailed in Fig. 5 and Table 2. Fig. S8 (Supporting information) shows the dQ/dV curves for NCM111 cathode materials during cycling. All exhibit similar redox peak pairs, indicating identical charge/discharge mechanism attributed to Ni^2+/4+ and Co^3+/4+ reactions [44,45]. However, the voltage differences of redox peaks (H1-M) between LiOH—NCM111, Li₂CO₃-NCM111 and LiNO₃-NCM111 increases sequentially during cycling. The smaller the voltage difference of redox peaks, the smaller the polarization and the better the reversibility of the sample. Therefore, LiOH—NCM111 shows enhanced reversibility [46]. Finally, Table S2 (Supporting information) compares the electrochemical performance of LiOH—NCM111 with those from typical synthesis methods and commercial NCM111. Overall, the electrochemical performance of NCM111 cathode materials prepared by FASP by the present study is either equivalent to or superior to these of materials prepared by typical methods or from commercial materials.

  Table 3

  

  Table 3.  Summary of electrochemical performance of NCM111 cathode materials prepared with different lithium salts.

  DownLoad: CSV

  Sample	Initial discharge capacity (mAh/g)	Initial coulomb efficiency (%)	Discharge specific capacity after 200 cycles (mAh/g)	Capacity retention after 200 cycles (%)
  LiNO3-NCM111	158.4	90.57	118.7	81.5
  Li2CO3-NCM111	161.5	90.78	125.0	81.0
  LiOH–NCM111	162.6	90.13	129.8	85.6

  In the present study, the transformation routes of the Li, Ni, Co, and Mn metal elements in FASP and the effects of the lithium salts on NCM cathode material were investigated. For Li, Ni, Co, and Mn nitrates commonly used in FASP, thermodynamic calculations and thermogravimetric analyses were performed, showing that the decomposition temperature of LiNO₃ (660 ℃) was significantly higher than that of Ni, Co, and Mn nitrates (less than 300 ℃). Therefore, to simplify the analysis of the transformation routes of Li, Ni, Co, and Mn elements in the flame, Ni, Co, and Mn nitrates could be studied as a group. The results suggested that Ni, Co, and Mn nitrates were converted to non-reactive stable oxides such as NiO. while LiNO₃ was converted to the reactive Li₂O, eventually leading to form LiOH and Li₂CO₃ under the environment of CO₂ and H₂O in the flame. Then, to clarify the effects of the three lithium salts on the NCM cathode material, NCM111 cathode materials with three different lithium salts (LiNO₃-NCM111, Li₂CO₃-NCM111, and LiOH—NCM111) were prepared by FASP. Characterization results showed that the lithium salts significantly influenced the structure of NCM111 cathode materials; LiOH—NCM111 had the lowest degree of Li/Ni disorder; Li₂CO₃-NCM111 and LiOH—NCM111 maintained a stable spherical structure with larger primary particles, whereas LiNO₃-NCM111 generated irregular particles. Electrochemical tests indicated that electrochemical performance of LiOH—NCM111, such as initial discharge specific capacity, cycling performance, rate performance, and reversibility were overall good.

  CRediT authorship contribution statement

  Junlei Wang: Writing – original draft, Resources, Investigation, Funding acquisition, Data curation. Wen Du: Software, Methodology, Data curation. Jinyu Li: Investigation, Data curation. Qianlong Wang: Validation, Supervision, Resources, Funding acquisition. Shilong Li: Writing – review & editing, Supervision, Project administration. Kun Wang: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was financially supported by the Tianjin Research Innovation Project for Postgraduate Students (No. 2022BKY057) and the National Natural Science Foundation of China (No. 52176124).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111665.

  
  
• 1. [1]

     M.A. Kebede, Curr. Opin. Electrochem. 39 (2023) 101261. doi: 10.1016/j.coelec.2023.101261

  2. [2]

     H. Xie, H. Peng, D. Jiang, et al., Chem. Eng. J. 470 (2023) 144051. doi: 10.1016/j.cej.2023.144051

  3. [3]

     J. Huang, Y. Wang, W. Ling, et al., J. Energy Storage 106 (2025) 114926. doi: 10.1016/j.est.2024.114926

  4. [4]

     J. Dai, Z. He, X. Li, et al., Chin. Chem. Lett. 36 (2025) 110063. doi: 10.1016/j.cclet.2024.110063

  5. [5]

     G. Yang, K. Pan, Z. Yan, et al., Chem. Eng. J. 452 (2023) 139405. doi: 10.1016/j.cej.2022.139405

  6. [6]

     H. Tang, D. Liu, J. Huang, et al., Chin. Chem. Lett. 36 (2025) 109987. doi: 10.1016/j.cclet.2024.109987

  7. [7]

     T.C. Liu, L. Yu, J.J. Liu, et al., Nat. Energy 6 (2021) 277–286. doi: 10.1038/s41560-021-00776-y

  8. [8]

     H. Zhao, W.Y.A. Lam, L. Sheng, et al., Adv. Energy Mater. 12 (2022) 2103894. doi: 10.1002/aenm.202103894

  9. [9]

     W. Liu, P. Oh, X. Liu, et al., Angew. Chem. Int. Ed. 54 (2015) 4440–4457. doi: 10.1002/anie.201409262

  10. [10]

      J. Kim, H. Lee, H. Cha, et al., Adv. Energy Mater. 8 (2018) 1702028. doi: 10.1002/aenm.201702028

  11. [11]

      C.H. Jo, N. Voronina, S.T. Myung, Energy Storage Mater. 51 (2022) 568–587. doi: 10.1016/j.ensm.2022.06.024

  12. [12]

      J. Wang, X. Lu, Y. Zhang, et al., J. Energy Chem. 65 (2022) 681–687. doi: 10.1016/j.jechem.2021.06.017

  13. [13]

      W. Du, J.L. Wang, Y.F. Xu, et al., J. Cleaner Prod. 472 (2024) 143499. doi: 10.1016/j.jclepro.2024.143499

  14. [14]

      H. Zheng, Y. Zhang, Z. Xu, et al., J. Energy Inst. 117 (2024) 101804. doi: 10.1016/j.joei.2024.101804

  15. [15]

      L. Lou, X. Liu, Y. Wang, et al., Chin. Chem. Lett. 36 (2025) 109726. doi: 10.1016/j.cclet.2024.109726

  16. [16]

      C. Chen, W. Du, Y. Xu, et al., Prog. Nat. Sci. Mater. 35 (2025) 814–821. doi: 10.1016/j.pnsc.2025.06.001

  17. [17]

      J. Wang, S. Li, W. Du, et al., Powder Technol. 464 (2025) 121233. doi: 10.1016/j.powtec.2025.121233

  18. [18]

      G.Y. Zang, J.A. Zhang, S.Q. Xu, et al., Energy 218 (2021) 119504. doi: 10.1016/j.energy.2020.119504

  19. [19]

      H. Chen, H.K. Mulmudi, A. Tricoli, Chin. Chem. Lett. 31 (2020) 601–604. doi: 10.1016/j.cclet.2019.05.016

  20. [20]

      J.L. Wang, J.Y. Li, S.L. Li, et al., Prog. Nat. Sci. Mater. 35 (2025) 485–491. doi: 10.1145/3760269.3760346

  21. [21]

      T. Ogihara, T. Kodera, K. Myoujin, et al., Mater. Sci. Eng. B: Adv. 161 (2009) 109–114. doi: 10.1016/j.mseb.2009.01.013

  22. [22]

      C. Abram, J.N. Shan, X.F. Yang, et al., ACS Appl. Energy Mater. 2 (2019) 1319–1329. doi: 10.1021/acsaem.8b01892

  23. [23]

      C. Yan, X.F. Yang, H. Zhao, et al., Proc. Combust. Inst. 38 (2021) 6623–6630. doi: 10.1016/j.proci.2020.06.332

  24. [24]

      J.N. Zhang, V.L. Muldoon, S.L. Deng, J. Power Sources 528 (2022) 231244. doi: 10.1016/j.jpowsour.2022.231244

  25. [25]

      J.N. Zhang, V.L. Muldoon, S.L. Deng, Proc. Combust. Inst. 39 (2023) 1165–1173. doi: 10.1016/j.proci.2022.10.014

  26. [26]

      J.E. Madero, J. Li, K.Y. Shen, et al., Appl. Energy Combust. Sci. 5 (2021) 100020.

  27. [27]

      J.H. Kim, J.H. Yi, Y.N. Ko, et al., Mater. Chem. Phys. 134 (2012) 254–259. doi: 10.1016/j.matchemphys.2012.02.060

  28. [28]

      H.K. Kammler, L. Mädler, S.E. Pratsinis, Chem. Eng. Technol. 24 (2001) 583–596. doi: 10.1002/1521-4125(200106)24:6<583::AID-CEAT583>3.0.CO;2-H

  29. [29]

      D. Hou, Q. Mao, Y. Ren, et al., Carbon 201 (2023) 189–199. doi: 10.1016/j.carbon.2022.09.002

  30. [30]

      W.Y. Teoh, R. Amal, L. Madler, Nanoscale 2 (2010) 1324–1347. doi: 10.1039/c0nr00017e

  31. [31]

      S. Tsantilis, H.K. Kammler, S.E. Pratsinis, Chem. Eng. Sci. 57 (2002) 2139–2156. doi: 10.1016/S0009-2509(02)00107-0

  32. [32]

      K. Okuyama, R. Ushio, Y. Kousaka, et al., AlChE J. 36 (2004) 409–419.

  33. [33]

      D. Hou, G. Wang, J. Gao, et al., Nanoscale 15 (2023) 5877–5890. doi: 10.1039/d3nr00060e

  34. [34]

      M. Bhat, S. Luo, J. Zhang, et al., Chem. Eng. J. 479 (2024) 147417. doi: 10.1016/j.cej.2023.147417

  35. [35]

      Y. Liang, K. Ku, Y. Lin, et al., ACS Appl. Mater. 13 (2021) 26915–26923. doi: 10.1021/acsami.1c03930

  36. [36]

      S. Yang, X. Wang, X. Yang, et al., J. Solid State Electrochem. 16 (2011) 1229–1237.

  37. [37]

      J.Z. Kong, F. Zhou, C.B. Wang, et al., J. Alloys Compd. 554 (2013) 221–226. doi: 10.1016/j.jallcom.2012.11.090

  38. [38]

      J. Zhu, J.C. Zheng, G.L. Cao, et al., J. Power Sources 464 (2020) 228207. doi: 10.1016/j.jpowsour.2020.228207

  39. [39]

      K. Matsumoto, R. Kuzuo, K. Takeya, et al., J. Power Sources 81-82 (1999) 558–561. doi: 10.1016/S0378-7753(99)00216-5

  40. [40]

      M. Ihalainen, M. Kortelainen, A. Mesceriakovas, et al., Adv. Powder Technol. 34 (2023) 104187. doi: 10.1016/j.apt.2023.104187

  41. [41]

      M. Jiang, P. Wang, Q. Chen, et al., Chin. Chem. Lett. 36 (2025) 110040. doi: 10.1016/j.cclet.2024.110040

  42. [42]

      B.H. Toby, R.B. Von Dreele, J. Appl. Crystallogr. 46 (2013) 544–549. doi: 10.1107/S0021889813003531

  43. [43]

      Z.D. Huang, X.M. Liu, S.W. Oh, et al., J. Mater. Chem. 21 (2011) 10777–10784. doi: 10.1039/c1jm00059d

  44. [44]

      Y. Wang, J. Roller, R. Maric, ACS Omega 3 (2018) 3966–3973. doi: 10.1021/acsomega.8b00380

  45. [45]

      K. Fröhlich, E. Legotin, F. Bärhold, et al., J. Solid State Electrochem. 21 (2017) 3403–3410. doi: 10.1007/s10008-017-3644-x

  46. [46]

      H. Sclar, D. Kovacheva, E. Zhecheva, et al., J. Electrochem. Soc. 156 (2009) A938–A948. doi: 10.1149/1.3212850

• 

  Figure 1  SEM images: (a, b) NCM111 precursor with lithium, (c, d) NCM111 precursor without lithium. (e) XRD patterns of NCM111 precursor with lithium and without lithium.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Experimental TGA curves, DTG curves and theoretical ∆G calculations of the raw nitrate salts: (a) LiNO₃, (b) Ni(NO₃)₂·6H₂O, (c) Co(NO₃)₂·6H₂O, (d) Mn(NO₃)₂·4H₂O.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Transformation route of LiNO₃ in the flame. (b) ΔG variation with temperature for reactions of Li₂O with CO₂ or H₂O. (c) XRD pattern of LiNO₃-N₂–0.00. (d) SEM image of LiNO₃-N₂–0.00. (e) XRD pattern of LiNO₃-N₂–2.25. (f) XRD pattern of LiNO₃-spray pyrolysis.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  The ΔG variation with temperature for reactions of Ni, Co, and Mn oxides with CO₂ or H₂O: (a) NiO, (b) CoO, (c) MnO₂. XRD pattern of (d) Ni(NO₃)₂-N₂–0.00 and Ni(NO₃)₂-N₂–2.25, (e) Co(NO₃)₂-N₂–0.00 and Co(NO₃)₂-N₂–2.25, (f) Mn(NO₃)₂-N₂–0.00 and Mn(NO₃)₂-N₂–2.25. SEM image of (g) Ni(NO₃)₂-N₂–0.00, (h) Co(NO₃)₂-N₂–0.00, (i) Mn(NO₃)₂-N₂–0.00.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  SEM images of the NCM111 cathode materials of (a) LiNO₃-NCM111, (b) Li₂CO₃-NCM111, and (c) LiOH–NCM111. EDS mapping images of the NCM111 cathode materials of (d) LiNO₃-NCM111, (e) Li₂CO₃-NCM111, and (f) LiOH–NCM111. (g) XRD patterns of NCM111 prepared with different lithium salts.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Electrochemical performance of NCM111 cathode materials prepared with different lithium salts: (a) Initial charge/discharge curves (0.1 C, 1 C = 160 mA/g). (b) Rate performance from 0.1 C to 1 C. (c) Cycling performance at 0.33 C within a potential range of 2.8–4.3 V at 25 ℃.

  下载: 全尺寸图片 幻灯片

  Table 1.  Theoretical (∆G < 0) and experimental decomposition temperatures of LiNO₃, Ni(NO₃)₂·6H₂O, Co(NO₃)₂·6H₂O, Mn(NO₃)₂·4H₂O.

  Sample	Theoretical decomposition temperature (℃)	Experimental onset decomposition temperature (℃)	Experimental end decomposition temperature (℃)
  LiNO3	660.0	506.1	696.2
  Ni(NO3)2·6H2O	264.0	268.3	345.1
  Co(NO3)2·6H2O	275.0	223.9	300.8
  Mn(NO3)2·4H2O	180.0	186.8	255.8

  下载: 导出CSV

  Table 2.  Structural parameters obtained from the Rietveld refinement of XRD patterns.

  Sample	a (Å)	c (Å)	c/a	I(003)/I(104)	Rp (%)	Rwp (%)
  LiNO3-NCM111	2.8631	14.2495	4.9769	1.290	5.64	3.46
  Li2CO3-NCM111	2.8624	14.2418	4.9754	1.321	4.84	3.14
  LiOH–NCM111	2.8621	14.2418	4.9760	1.356	4.78	3.16

  下载: 导出CSV

  Table 3.  Summary of electrochemical performance of NCM111 cathode materials prepared with different lithium salts.

  Sample	Initial discharge capacity (mAh/g)	Initial coulomb efficiency (%)	Discharge specific capacity after 200 cycles (mAh/g)	Capacity retention after 200 cycles (%)
  LiNO3-NCM111	158.4	90.57	118.7	81.5
  Li2CO3-NCM111	161.5	90.78	125.0	81.0
  LiOH–NCM111	162.6	90.13	129.8	85.6

  下载: 导出CSV
•