Solvent-assisted solution combustion synthesis of the LiFePO₄/C cathode material for Li-ion storage

Summary

LiFePO₄, with a high specific capacity, thermal and chemical stability, and long-term cyclability, is a good cathode material for lithium-ion batteries. In this research, the main aim was to investigate the effect of solvent type—water, ethanol, methanol, a water-50 Vol. % methanol mixture, and a water-50 Vol. % ethanol mixture—on the electrochemical performance and morphology of LiFePO₄/C powders prepared by the solution combustion method. The physicochemical characteristics were investigated by X-ray diffraction, scanning electron microscopy, Raman spectroscopy, X-ray photoelectron spectroscopy, and various electrochemical techniques. The LiFePO₄/C powders prepared with a mixed water-50 Vol % ethanol solvent exhibited a discharge specific capacity of 105 mAh g⁻¹ at a current rate of 1C and a capacity retention of 80% after 3000 cycles at 10C.

Graphical abstract

Highlights

• •LiFePO₄/C was prepared by the solution combustion method using various solvents
• •Water-50 Vol. % ethanol led to finer microstructure and higher specific surface area
• •LiFePO₄/C exhibited a capacity retention of 80% following 3000 cycles at 10°C

Applied sciences; Energy storage

Introduction

Recently, energy demand has increased significantly due to rapid industrial development and changing lifestyles. In the meantime, the lack of sustainable fossil fuel resources necessitates finding alternative renewable energy resources.¹ However, the renewable sources have an intermittent nature, demanding high-performance energy storage devices.^1,2 Lithium-ion batteries (LIBs) are highly regarded as energy storage solutions because of their high energy density and relative safety.³ The storage performance, including safety, energy density, and rate capability of LIBs, mainly depends on their cathode, anode, and electrolyte as their key components.^2,4

Among various cathodic materials for lithium-ion batteries (LIBs), including olivine-, spinel-, and layered-structured materials, olivine-structured LiFePO₄ has emerged as a promising choice for cathode applications. This is due to its high theoretical capacity (170 mAh g⁻¹), satisfactory chemical and thermal stability, and relatively high cyclability.⁵ Nevertheless, the use of LiFePO₄ is hindered by its intrinsically low electronic and ionic conductivities.⁶ To address these limitations, several strategies have been employed, including particle downsizing,⁷ doping,⁸ morphology control,⁹ carbon coating,¹⁰ and compositing,¹¹ all aimed at enhancing the material’s conductivity through adjustments in the synthesis process.^12,13 In recent studies, the selection of phosphorus sources has been shown to significantly affect the morphology and electrochemical behavior of LiFePO₄-based materials.¹⁴ For instance, Song et al.,¹⁵ synthesized LiFePO₄ materials using adenosine triphosphate (ATP) as a phosphorus source. The optimized LiFePO₄ exhibited a uniform rod-like morphology and demonstrated excellent electrochemical performance. Specifically, the LiFePO₄/C composites with carbon coating delivered discharge capacities of 154.3 mAh g⁻¹ and 110.1 mAh g⁻¹ at 1 C and 10 C, respectively. Remarkably, after 100 cycles at 10 C, the composite retained a capacity of 99.8 mAh g⁻¹, corresponding to a capacity retention rate of 90.02%.

To date, several chemical synthesis methods have been developed to control the properties of LiFePO₄ cathode materials. These methods include solvothermal,¹⁶ sol-gel,¹⁷ co-precipitation,¹⁷ and solution combustion.¹⁸ Among these, solution combustion synthesis (SCS) stands out as a promising route due to its high yield, feasibility, chemical versatility, and capacity to produce powders with a high specific surface area.¹⁹ The SCS is based on the dissolution of metal precursors and organic fuels in a suitable solvent, which is then gelled and ignited by heating the solution. The solvent type is known to be one of the main components of the SCS because it has a significant effect on the solubility of metal precursor, gelation process, and combustion behavior.²⁰ For example, Sarmadi et al.²¹ reported that the solvent type plays a crucial role in the combustion synthesis of LiFePO₄ material. Using the bio-based fuel L-lysine with water and ethanol solvents, they showed that ethanol’s lower polarity and organic nature promote graphitization, reduce particle size (∼70 nm), and increase the specific surface area (56 m² g⁻¹), resulting in enhanced electronic conductivity and Li⁺ diffusion (36 mAh g⁻¹ at 10 C, 85% retention after 2000 cycles).

In this work, the LiFePO₄/C cathode material was prepared by SCS using cetyltrimethyl ammonium bromide (CTAB) as the organic fuel and various solvents, including water, ethanol, methanol, and mixed water-50 Vol. % ethanol and mixed water-50 Vol. % methanol. Various characterization methods were used to investigate how the solvent type affects the morphology and electrochemical performance of LiFePO₄ powders.

Results and discussion

Structural and microstructural characterization

XRD patterns of the LFP-W, LFP-E, LFP-M, LFP-MW, and LFP-EW samples are presented in Figure 1. The position and intensity of the diffraction peaks closely match those of the reference pattern (PDF2 no. 96-110-1112) for the olivine-structured LiFePO₄ phase with space group Pnma. The sharpness of the peaks suggests the high crystallinity of the samples.²² The absence of irrelevant reflections indicates the high purity of the synthesized LFP/C powders. Table 1 presents the lattice parameter and unit cell volume calculated from the Rietveld refinement results (Figure S1 of the Supporting Information file). The well-crystalline LiFePO₄ material without crystal defects has a unit cell volume of 291.4 ų.²³ The lower unit cell volume of various solvents indicates the presence of crystal defects, such as antisite defects and vacancies, which control the electrochemical performance. The EW sample has the closest value to 291.4 ų, indicating its higher crystallinity.

Figure 1.

XRD patterns of the calcined samples at 750°C for 6 h

Table 1.

Lattice parameters and unit cell volume

Samples	a (Å)	b (Å)	c (Å)	Volume (Å3)
LFP-EW	10.333	6.011	4.690	291.337
LFP-MW	10.329	6.010	4.693	291.329
LFP-M	10.325	6.007	4.694	291.133
LFP-E	10.326	6.007	4.694	291.123
LFP-W	10.334	6.008	4.692	291.305

Figure 2A displays the Raman spectra for the LFP-EW sample, revealing four different peaks: G, D₁, D₃, and D₄, corresponding to graphite carbon, disorder phonon, amorphous carbon moieties, and C–C/C=C stretching vibrations of polyene-like structures, respectively.²⁴ The exact positions of the G and D1 peaks are shown in Table S1 of the Supporting Information file. The I_D1/I_G values of the LFP-W, LFP-E, LFP-M, LFP-MW, and LFP-EW samples are 0.93, 0.80, 1.31, 1.15, and 0.71, respectively (Figure 2B), suggesting that a mixed solvent of water and ethanol increases the degree of graphitization. On the other hand, methanol as the solvent hinders the formation of graphitized, well-ordered carbon, leading to a higher I_D1/I_G value. These results might be due to the different combustion behavior using the methanol solvent, which affects the functional groups of surface particles.²⁵

Figure 2.

Raman spectra

(A) The LFP-EW sample.

(B) The I_D1/I_G ratio.

The composition and chemical state of the LFP-EW sample were investigated using X-ray photoelectron spectroscopy (XPS). The survey spectrum (Figure 3A) displays characteristic signals for the elements lithium (Li), iron (Fe), phosphorus (P), oxygen (O), and carbon (C). In the high-resolution Fe 2p spectrum (Figure 3B), two distinct peaks are observed at binding energies (BEs) of 709.8 eV and 724.2 eV, corresponding to the Fe 2p_3/2 and Fe 2p_1/2 states, respectively. Additionally, two satellite peaks are observed at 715.6 and 729.6 eV, providing compelling evidence for the presence of Fe²⁺ in the LFP structure.²⁶ Furthermore, the observed difference in binding energy between the Fe2p and satellite peak is approximately 5.0 eV, which is below the 8 eV associated with Fe³⁺, suggesting the presence of Fe²⁺ at the sample’s surface.²⁷ According to Figure 3C, the high-resolution XPS spectrum of C 1s was deconvoluted into four components corresponding to C(sp²), C(sp³), C–O, and C=O bonds. The main peak centered at approximately 283.7 eV is attributed to sp²-hybridized carbon (graphitic carbon), while the smaller component at around 284.8 eV corresponds to sp³-hybridized carbon associated with disordered carbon or defect sites. The peaks located at ∼283.3 eV and ∼287.1 eV are assigned to C–O and C=O groups, respectively. The quantitative analysis based on the fitted peak areas shows that the C(sp²) fraction (47%) is significantly higher than the C(sp³) fraction (18%), confirming the predominance of graphitic carbon in the sample. The strong intensity and narrower FWHM of the sp² peak further indicate a higher degree of structural order, which is characteristic of graphitic domains. In contrast, the minor presence of sp³ and oxygenated carbons suggests limited surface disorder or oxidation.²⁸ The Li1s peak (Figure 3D) was deconvoluted into two peaks, corresponding to the Li–O bond (55.8 eV) and Fe²⁺ (57.3 eV), respectively.²⁹ There is a distinct peak at a BE of 132.7 eV in the P 2p spectrum (Figure 3E), indicating the P⁵⁺ state. According to the O 1s spectrum (Figure 3F), the main peak at 530.7 eV confirms the presence of O²⁻ in the crystal lattice. In addition, two shoulder peaks at 532.3 eV and 533.0 eV corresponding to C–O and C=O chemical bonds can be assigned to the carbon layer on the LFP particles. The XPS results indicate that the correct oxidation states of the elements in the LiFePO₄/C material were successfully determined using a mixed ethanol-water solvent via the solution combustion synthesis method.

Figure 3.

X-ray photoelectron spectroscopy

(A) Full XPS of the LiFePO₄/C powders prepared by a mixed ethanol-water solvent.

(B) High-resolution spectra of Fe2p.

(C) C1s.

(D) Li1s.

(E) P2p.

(F) O1s.

Figure 4 exhibits the SEM microstructures of the various samples. The LFP powders are composed of quasi-spherical particles that are hardly distinguishable due to their considerable agglomeration. By the ethanol solvent, the particles are somewhat separated, while the LFP particles are tightly aggregated by the methanol solvent. By mixing water with ethanol or methanol, the porosity increases as agglomeration and particle size decrease. The microstructure of SCS products depends on the physicochemical properties of the precursor solution and the amount and type of organic fuels, thereby tuning the characteristics of the gelation process, the combustion reaction rate, the combustion products, and the combustion temperature.²⁰ The agglomerated microstructure can be attributed to the slow combustion reaction for liberating the gaseous products.³⁰ The ethanol solvent, with a lower boiling point (78.4°C) than that of the water solvent (100 °C), can promote gelation and disintegrate the dried gels, leading to less agglomeration following the combustion reaction.³¹ However, the methanol solvent has a higher surface tension (22.51 mN m⁻¹ at room temperature) than that of the ethanol solvent (21.82 mN m⁻¹ at room temperature),³² leading to further agglomeration.³³ Furthermore, burning the ethanol solvent liberates higher-gaseous products, benefiting from its porous microstructure.³⁴

Figure 4.

SEM images

(A and B) LFP-W.

(C and D) LFP-E.

(E and F) LFP-EW.

(G and H) LFP-M.

(I and J) LFP-MW samples.

Scale bar of (A, C, E, G, and I) 5 μm and (B, D, F, H, and J) 2 μm.

The microstructure of the SCS products is dependent on the solvent type through two steps. The first step is the gradual removal of solvent during the drying process, which is related to the surface tension and boiling point of the solvent.³⁵ As shown in Figure 5, when the extent of solvent removal reaches a specific limit, the surface tension of the solvent (or dispersing liquid), which exerts an inward force, dictates how the primary particles assemble. High-surface-tension solvents tend to bring primary particles into closer contact, forming denser agglomerates.³³ On the contrary, low-surface-tension solvents do not exert inward forces of as much magnitude, leading to looser agglomerates. The second factor is the combustion step, in which the residual solvent is removed during a very short period. The fast removal of solvent at very high temperatures is often assumed to lead to open and voluminous primary particle assemblies.³¹ The boiling point of the solvent plays a crucial role in the combustion step rather than its surface tension. More specifically, solvents of higher boiling points survive the drying step more effectively. The higher the amount of the remnant solvent, the greater the impact of the fast solvent removal on pore formation and surface area increase.

Figure 5.

A schematic of the effect of surface tension on agglomeration

In the LFP-W sample, the adverse effects of high surface tension and boiling point during the drying step outweigh any potential benefits in the combustion steps. On the other hand, the LFP-E and LFP-M samples may retain the high surface area during the drying step, whereas the ethanol and methanol solvents do not survive this step and therefore do not participate in pore and surface formation during the combustion method due to their lower boiling points. Mixing water with ethanol allows a compromise between preserving surface area in the drying step and forming pores in the combustion step. The former is the contribution of the low-surface-tension of ethanol solvent, and the latter is the contribution of the high-boiling-point of water.³⁶ The higher surface area values and less agglomerated texture of the LFP-EW sample synthesized with a water-50 Vol. % ethanol solvent are most likely due to the lower surface tension of ethanol and the higher boiling point of water (Table S2).³⁴

Figure 6A illustrates the N₂ adsorption-desorption isotherms of the LFP-W, LFP-E, LFP-M, LFP-MW, and LFP-EW samples. The LFP/C powders with a mesoporous structure exhibit an H3 hysteresis loop in their IV-type adsorption-desorption isotherms (IUPAC classification).³⁷ There is a rapid rise in adsorption isotherms at high relative pressures, suggesting that micropore filling is followed by a plateau.³⁸ The desorption curve drops more gradually than the adsorption curve, providing further evidence of mesopores.³⁹ The LFP-EW sample demonstrates enhanced adsorption values, indicating a larger micropore volume, suggesting a more porous structure that potentially improves its electrochemical performance by increasing the accessibility of electroactive sites for lithiation/delithiation reactions.⁴⁰ The LFP-W, LFP-E, LFP-M, LFP-MW, and LFP-EW samples have the specific surface area values of 37.3, 60.5, 57.8, 73.2, and 87.5 m²/g, respectively. A mixed ethanol-water solvent with a proper balance of viscosity and surface tension yields a foamy gel and less-agglomerated combustion products, thereby substantially increasing the contact area between the electroactive material and the electrolyte, facilitating lithiation/delithiation reactions.^22,41 Pore size distributions calculated using the BJH method are presented in Figure 6B. The pore volumes are 0.104, 0.133, 0.139, 0.188, and 0.230 cm³ g⁻¹ for the LFP-W, LFP-E, LFP-M, LFP-MW, and LFP-EW samples, respectively. The higher pore volume of the LFP-EW sample confirms the effective role of the gaseous products liberated using a mixed water-50 Vol. % ethanol solvent.⁴²

Figure 6.

Textural properties

(A) N₂ isotherms.

(B) Pore size distribution plots.

Electrochemical performance

Figure 7A illustrates the specific capacity of the LFP-MW, LFP-W, LFP-E, and LFP-EW samples at various current rates. The LFP-EW sample demonstrates the highest specific capacity, measuring 130 mAh g⁻¹ at a current rate of 0.1C. In contrast, the LFP-MW sample shows the lowest specific capacity of 82 mAh g⁻¹ at the same current rate, along with the weakest rate capability. The lower ratio of graphitic carbon is likely the reason for the lower rate capability, as this more disordered carbon coating results in lower electronic conductivity. The higher discharge specific capacity and higher rate capability of the LFP-EW can be due to its higher specific surface area and porosity, higher crystallinity, smaller particle sizes, and more uniform microstructure.^20,43 The ethanol-water sample exhibits approximately 107 mAh g⁻¹ at 1 C, down from around 130 mAh g⁻¹ at 0.1 C, demonstrating good rate performance. In contrast, the methanol - water derived sample shows a sharp capacity decline (from about 83 mAh g⁻¹ at 0.1 C to around 25 mAh g⁻¹ at 1 C). This pronounced difference is attributed to the higher graphitic carbon content and improved electrical conductivity of the ethanol-water sample, which facilitates electron transport and mitigates polarization under high-rate conditions.⁴⁴ Figure 7B compares the charge/discharge curves at a current rate of 0.1C. The phase transition between the LiFePO₄ and FePO₄ results in a voltage plateau in the range of 3–3.4 V vs. Li⁺/Li.⁴⁵ The potential differences between charge and discharge plateaus are 229, 84, 60, and 52 mV for the LFP-MW, LFP-W, LFP-E, and LFP-EW samples, respectively. Therefore, the LFP-EW sample exhibits lower electrode polarization and facile de/intercalation of Li+ ions, owing to its higher crystallinity, electrical conductivity, and specific surface area.⁴⁶ Figure 7C illustrates the charge/discharge curves of the LFP-EW sample at various C rates. The voltage plateau shortens at higher current rates due to increased electrode polarization effects.⁴⁷ In contrast, the GCD curves maintain distinct voltage plateaus even at elevated current rates, attributed to the rapid kinetics of the de/lithiation process.⁴⁵ Figure 7D compares the long-term cycling performance at a current rate of 10 °C for 3000 cycles. The LFP-EW sample has the highest capacity retention of 81%, corresponding to a decrease of specific capacity from 60 to 48.6 mAh g⁻¹. The LPF-W and LFP-E samples exhibit lower average specific capacities of 35 and 18 mAh g⁻¹, respectively, with capacity retentions of 75.6 and 53% after 3000 cycles. The irreversible loss in specific capacity by cycling can be attributed to two reasons: (1) dissolution of Fe ions⁴⁸ and (2) the mechanical stresses induced by the structural changes during the de/intercalation processes.⁴⁹ Charge/discharge cycling facilitates the exposure of more new active particles, making them suitable for participating in the delithiation/lithiation process.^48,49 Moreover, the Coulombic efficiency (Figure S3 of the Supporting Information) is approximately constant (100%) throughout the cycling process for all samples, demonstrating the reversible insertion and extraction of lithium ions.⁵⁰ Table 2 compares the electrochemical performance of LiFePO₄/C powders prepared by the solution combustion method. It is worth noting that the design of the electrolyte is an effective way for promoting the electrochemical performance of the LFP/C material.

Figure 7.

Rate capability and cycling performance

(A) Rate capability of the LFP-W, LFP-E, and LFP-EW samples.

(B) The GCD curves of the LFP-W, LFP-E, and LFP-EW samples.

(C) The GCD curves of the LFP-EW sample.

(D) Cycling performance at 10C.

Table 2.

Electrochemical performance of the LiFePO₄/C powders prepared by the SCS method

Sample	Fuel type	Solvent	Specific capacity (mAhg−1)	Retention (%)	Reference
LiFePO4/C	L-Lysine	Ethanol	36 at 10C	85.35 at 5C 2000 cycles	Sarmadi et al.21
Mn-doped LiFe1-xMnxPO4	Glycine	water	167.5 at 0.1C	93.8 at 0.2C 100 cycles	Mulik et al.51
LiFePO4/C	CTAB and glycine	water	158 at 0.1C	95 at 0.2C 200 cycles	Duan et al.52
LiFePO4/C	Glycine and malonic acid	water	127 at C/3	–	Vujković et al.53
LiFePO4/C	CTAB and glycine	water	110 at 0.1C	98 at 1C 50 cycles	Karami et al.20
LiFePO4/C	CTAB	water	91 at 0.1C	–	Haghi and Masoudpanah54
LiFePO4/C	CTAB	water-50 Vol. % ethanol	105 at 1C	80 at 10C	This work

Figure 8 shows the cyclic voltammetry (CV) curves of the LFP-W, LFP-E, and LFP-EW materials at various scan rates. All samples have a pair of redox peaks, even at high scan rates. The redox peaks are due to the de/insertion of Li⁺ ions into the lattice of the LiFePO₄ phase, which are consistent with the charge/discharge voltage plateaus. By de/inserting the Li⁺ ions, a reversible phase transition occurs between LiFePO₄ and FePO₄ as follows³:

LiFePO₄→FePO₄+Li⁺+e⁻ (Equation 1)

Figure 8.

Analysis of chargestorage mechanisms

CV curves, the separated contributions of capacitive- and diffusion-controlled mechanisms at 0.5 mV s⁻¹, and contribution ratios vs. scan rate for (A–C) LFP-W, (D–F) LFP-E, and (G–I) LFP-EW samples.

Figure 8 illustrates that the redox peaks become gradually separated as the scan rate increases. Figure 8 illustrates that the LFP-EW sample exhibits the highest peak current and the smallest potential difference between the anodic and cathodic peaks, confirming significant electrochemical reversibility, enhanced kinetic characteristics, and reduced electrochemical polarization.⁵⁵ By plotting a linear relationship between peak currents (I_p) and the square root of scan rate (ν^1/2) (Figure S4 of the Supporting information file), the diffusion coefficient of lithium ions (D_Li) can be calculated according to the Randles-Sevcik equation⁵⁶:

$$I_p=2.69\times {10}^5{n}^{3/2}{C}_{0}A{D_{Li}}^{1/2}{v}^{1/2}$$ (Equation 2)

where n is the number of electrons transferred during the electrochemical reaction, A (cm²) is the electrode area, C₀ (mol cm⁻³) is the initial concentration of lithium ions, R (J K⁻¹·mol⁻¹) is the gas constant, and T (K) is the temperature. Figure S4 shows a linear relationship between Ip and ν^0.5, which is used to calculate the diffusion coefficient (D_Li) (Table 3). The LFP-EW sample has the highest D values among the two other samples, which is attributed to its smaller particle size and shorter effective transmission path, thereby accelerating Li-ion diffusion and improving the capacitive behavior.⁵⁷ Specifically, Ip and v are related by a power law, Ip = aν^b, in which the b value signifies the storage mechanism of Li⁺ ions in the LiFePO₄ powders. The b value, calculated from the linear slope of Ln(Ip) vs. Ln(ν) plot (Figure S5 of Supporting Information file), is summarized in Table 3 for the oxidation and reduction peaks. The similar b values of both oxidation and reduction peaks confirm that the diffusion-controlled processes are the predominant mechanism.⁵⁸ The b_oxidation and b_reduction values of the LFP-W sample are 0.53 and 0.58, respectively, suggesting that the storage of Li⁺ ions is constrained by ion diffusion inside the electrode matrix.⁵⁸ Furthermore, the b values are 0.62 and 0.64, respectively, for the oxidation and reduction peaks in the LFP-E sample, indicating that the redox reaction mechanism is controlled by the faradaic redox rate rather than by Li⁺ ion diffusion.³⁷ Nevertheless, the mixed ethanol-water solvent exhibits the b values of 0.70 and 0.72 (which fall within the typical range of 0.5–0.8 reported in most LFP studies^59,60,61 revealing a substantial influence of the faradaic redox reaction rate.³⁷ The role of various mechanisms can be calculated from the following equation.⁵⁶

i = k₁v + k₂v^0.5 (Equation 3)

where k₁ν is the contribution of the capacitive-controlled mechanism due to the double-layer charging/discharging, and k₂ν^0.5 is the diffusion-controlled mechanism.³⁷ Figures 8B–8E and 8H show the current response of capacitive- and diffusion-controlled processes at a scan rate of 0.5 mV s⁻¹. The contribution ratios vs. scan rate are shown in Figures 8C–8F and 8I. The capacitive contribution increases significantly from 84.6 to 93.9% at 0.1–0.8 mV s⁻¹ in the LFP-EW sample, due to its higher specific surface area.⁶²

Table 3.

Characteristic values derived from CV curves

Samples	Soxidation	DLi+	Sreduction	DLi+	boxidation	breduction
LFP-EW	0.128	11.7 × 10−11	−0.108	8.30 × 10−11	0.70	0.72
LFP-E	0.116	9.61 × 10−11	−0.097	6.70 × 10−11	0.62	0.64
LFP-W	0.042	1.26 × 10−11	−0.035	8.72 × 10−12	0.53	0.53

To determine the electrode kinetics, electrochemical impedance spectroscopy (EIS) is performed on the LFP-W, LFP-E, and LFP-EW samples. The Nyquist plots with the equivalent circuit are presented in Figure 9A. The charge transfer resistance (R_ct) results in a semicircle in the medium-frequency region. At low frequencies, Li+ ion diffusion in the porous electrode structure leads to an inclined linear relationship between -Z″ and Z', known as the Warburg resistance.⁶³ Table 4 shows that the R_ct value of the LFP-EW electrode (18.37 Ω) is significantly lower than those of the LFP-W (49.52 Ω) and LFP-E (19.65 Ω) electrodes, which is attributed to its higher electrical conductivity resulting from the increased crystallinity achieved with a mixed ethanol-water solvent. The lithium diffusion coefficient (D_Li+) can be calculated by the following equation⁶⁴:

$$D_{Li+}=\frac{R^2{T}^2}{2A^2{n}^4{F}^4{C}^2{\sigma }^2}$$ (Equation 4)

where σ, Warburg factor, is the linear slope of the Z′ vs. ω^−0.5 plot (Figure 9B). The Li⁺ diffusion coefficients are illustrated in Table 4. Significantly, the higher D_Li⁺ value of the LFP-EW sample is due to its proper structural and microstructural properties.

Figure 9.

Electrochemical impedance spectroscopy

Nyquist plots (A) and Zʹ vs. ω^−1/2 (B) for the LFP-W, LFP-E, and LFP-EW samples.

Table 4.

Resistance parameters and diffusion coefficient obtained by fitting the EIS spectra

Samples	Rs	Rct	DLi+
LFP-EW	3.11	18.37	4.00 × 10−9
LFP-E	8.77	19.65	3.59 × 10−9
LFP-W	10.24	49.52	8.58 × 10−10

The GITT curves of the LFP-EW sample as a function of time are typically exhibited in Figure 10A (Figure S6 of the Supporting Information file). As shown, all three samples exhibited symmetric behavior during charge and discharge cycles. Additionally, the LFP-EW sample demonstrated longer charging and discharging times due to slower voltage changes, reflecting the overpotential associated with the respective charge/discharge processes.⁶⁵ As shown in Figure 9A and Figure S6, the charge and discharge ion migration rates of LFP-EW are far higher than those of LFP-E and LFP-W, indicating the superior ion mobility of LFP-EW.⁶⁶ The D_Li⁺ value can be determined as follows:⁶⁷

$${\mathrm{D}}_{\text{Li}+}=\frac{4}{\mathrm{\pi }\mathrm{\tau }}{\left(\frac{{\mathrm{m}}_{\mathrm{b}}{\mathrm{V}}_{\mathrm{M}}}{{\mathrm{M}}_{\mathrm{B}}\mathrm{S}}\right)}^2{\left(\frac{\Delta {\mathrm{E}}_{\mathrm{s}}}{\Delta {\mathrm{E}}_{\mathrm{\tau }}}\right)}^2\left(\mathrm{\tau }\ll {\mathrm{L}}^2/{\mathrm{D}}_{\text{Li}}\right)$$ (Equation 5)

where τ, V_M, M_B, S, and m_b are the current duration time, molar volume, molecular weight of active material, the electrode’s surface area, and active material’s mass, respectively. The steady voltage charge induced by the voltage pulse is ΔE_s, and the voltage fluctuation during constant current charge/discharge is ΔE_τ,³⁷ shown in Figure S7 of the supporting information file. Figure 10B shows the trends in the diffusion coefficient, charge, and discharge capacity. All the samples showed a similar trend during the charging and discharging cycles. The LFP-EW sample displayed the highest diffusion coefficient with an average value of 2.25 × 10⁻¹⁰ cm² s⁻¹ across the whole charging and discharging cycles, while the other two samples demonstrated an approximately similar, but lower average value of 5.38 × 10⁻¹¹ and 6.12 × 10⁻¹¹ cm² s⁻¹ for the LFP-W and LFP-E samples, respectively, which agrees with the EIS and CV results. The cycling stability and Li-ion diffusion coefficient reported in this study are compared with those from earlier studies in Table 5, demonstrating the superiority of these materials.

Figure 10.

Electrode kinetics

(A) GITT curve of the LFP-EW sample as a function of time.

(B) Dependence of the D_Li⁺ value on charge and discharge capacity.

Table 5.

Comparison of the electrochemical performances of this work with other reported LiFePO₄ compounds

Cathodic material	Synthesis method	Capacity retention	DLi (cm2 s−1)	Reference
LiFePO4/C	SCS	81% in 10C for 3000 Cycles	11.7 × 10−11	This work
LiFePO4/C	Coated with electrophoretic deposition	89% in 5C for 300 Cycles	9.6 × 10−11	Park68
LiFePO4/C	–	69% in 1C for 200 Cycles	5.96 × 10−11	Li et al.69
LiFePO4/C	calcination/pyrolysis	75% in 1C for 50 Cycles (25 °C)	2.59 × 10−14	Hsieh et al.70
LiFePO4/C	calcination/pyrolysis	81% in 1C for 50 Cycles (40 °C)	2.42 × 10−13	Hsieh et al.70
LiFePO4/C	SCS	–	4.25 × 10−11	Karami et al.20

Finally, the well-crystalline LiFePO₄/C powders were successfully prepared by the solution combustion method using water, methanol, ethanol, water-methanol, and water-ethanol solvents. The mixed water-ethanol solvent led to smaller particles, a higher specific surface area, and a higher graphitic carbon domain. These microstructural features improved the electrochemical performance of the LiFePO₄/C powders, including a high specific capacity of 130 mAh g⁻¹ at a current rate of 0.1C, a high rate capability of 93% by increasing the current rate up to 10C, and a superior capacity retention of 81% for 3000 charge/discharge cycles at a current rate of 10C. Furthermore, the Li diffusion coefficient (derived from CV results) increased from 1.26 × 10⁻¹¹ to 11.7 × 10⁻¹¹ cm² s⁻¹ when the solvent was changed from water to a mixed water-ethanol solution, as characterized by CV, EIS, and GITT.

Limitations of the study

This study has two main limitations. First, access to high-resolution TEM imaging was limited, such data could directly visualize LiFePO₄ and carbon interfaces and atomic-scale defects, further validating the proposed mechanisms. Second, the equipement for assembling the full cell is limited.

Resource availability

Lead contact

Further information and requests for resources, materials, and data should be directed to and will be fulfilled by the Lead Contact: Dr. Seyyed Morteza Masoudpanah (Email: masoodpanah@iust.ac.ir).

Materials availability

This study did not generate new unique materials, cell lines, or biological resources. All synthesized LFP/C samples are available from the lead contact upon reasonable request.

Data and code availability

• •All data reported in this article will be shared by the lead contact upon reasonable request.
• •This article does not report any original code.
• •No publicly archived datasets were generated during this study.

Acknowledgments

Authors are grateful to Iran University of Science and Technology for supporting this work.

Author contributions

A.G. conceptualization, investigation, formal analysis, visualization, writing – original draft, and writing – review and editing, S.M.M. conceptualization, methodology, writing – review and editing, and supervision, M.T.J. visualization, formal analysis, visualization, and writing – original draft, and H.N. formal analysis.

Declaration of interests

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 article.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Lithium nitrate	Merck Co	7790-69-4
Iron (III) nitrate nonahydrate	Merck Co	7782-61-8
Ammonium dihydrogen phosphate	Merck Co	7722-76-1
Cetyltrimethylammonium bromide	Sigma-Aldrich	57-09-0
Methanol	Merck Co	67-56-1
Ethanol	Merck Co	64-17-5

Method details

All the precursors, such as LiNO₃, Fe(NO₃)₃.9H₂O, NH₄H₂PO₄, cetyltrimethyl ammonium bromide (C₁₉H₄₂BrN), methanol, and ethanol, were purchased from Merck Co and Sigma-Aldrich.

Synthesis procedure

Stoichiometric amounts of the LiNO₃, Fe(NO₃)₃.9H₂O, and NH₄H₂PO₄ precursors were dissolved in 30 mL of solvent, including water, ethanol, methanol, mixed water-50 Vol. % ethanol, and mixed water-50 Vol. % methanol. Then, the appropriate amount of CTAB as organic fuel was added to the precursor solution according to the following reaction²⁰:

LiNO₃ + Fe(NO₃)₃ + NH₄H₂PO₄ + (27φ/118)C₁₉H₄₂BrN + (9/2)((3φ/2)-1)O₂ → LiFePO₄ + (513φ/118)CO₂ + ((567φ/118)+3)H₂O + (2.5+(27φ/236))N₂ + (27φ/236)Br₂ (Equation 6)

The fuel-to-oxidant ratios (φ) show the required amounts of external oxygen. The φ value was set as 5. The precursor solution was thickened on a hot plate under magnetic stirring at 75 °C and then completely dried overnight at 75 °C in an oven. The dried gel was burnt on a preheated hot plate for 5 minutes. The spongy combusted powders were manually crushed and homogenously mixed with 20 wt. % sucrose as a carbon source. Finally, the LiFePO₄ phase was well crystallized by calcination at 750 °C for 6 hours. The calcination atmosphere was Ar-5%H₂ (V:V). The obtained powders were coded according to the solvent type: water (LFP-W), ethanol (LFP-E), methanol (LFP-M), water-50 Vol. % methanol (LFP-MW), and water-50 Vol. % ethanol (LFP-EW).

Materials characterization

The crystal structure, lattice parameters, and unit cell volume were extracted from X-ray diffraction (XRD) patterns obtained on the D8 ADVANCE instrument (Bruker, Japan) equipped with CuKα irradiation (λ=1.54060 Å). The powders were analyzed in the diffraction range of 2θ=10-80°. The morphology was examined by scanning electron microscopy (Vega II, TESCAN, Czech Republic). The textural properties were evaluated from the N₂ ad/desorption isotherms recorded on the BELSORP Mini II instrument (BEL, Japan). The LiFePO₄/C powders were degassed at 250 °C for 5 hours. The graphite degree of the carbon layer was examined by Raman spectroscopy (XploRA PLUS, HORIBA) using a laser excitation of 532 nm. To determine the valence state of the Fe element, the X-ray photoelectron spectroscopy (XPS) was obtained on AXIS Supra+ system (Kratos Analytical, Japan).

Electrochemical measurements

Firstly, a slurry was prepared by vigorously stirring a mixture of LFP/C powders, carbon black, and polyvinylidene fluoride (70:15:15 wt. %) in the N-Methyl-2-pyrrolidone solvent. The homogenized slurry was coated onto an Al foil current collector. The coated foils were dried completely at 75 °C in an oven overnight. After drying, the foil was punched into circular electrodes with a diameter of 14 mm. For practical relevance, the specific capacity was calculated based on a mass loading of 2 mg cm^-2 on the Al current collector.⁷¹ The 2016-type coin cells were assembled in an Ar-filled glovebox using LFP/C working electrode, Celgard 2500 as separator, lithium foil as counter electrode, and 1 M solution of LiPF₆ in dimethyl carbonate, ethylene carbonate (1:1 V:V) as electrolyte.

The assembled cells were charged/discharged in the voltage range of 2-4 V vs. Li⁺/Li at various current rates of 0.1C, 0.2C, 0.5C, 1C, 2C, 5C, 10C, and 20C on a BTS-5V 10 mA battery tester (NEWARE, China). The cyclic voltammetry (CV) was recorded in the potential range of 2.0-4.0 V vs. Li/Li⁺ at various scan rates of 0.1, 0.2, 0.3, 0.5, and 0.8 mV s^-1 on a Radstat200 workstation (Kianshardanesh, Iran). The electrochemical impedance spectra (EIS) were taken at the frequency range of 0.01 Hz to 50 kHz with an alternating voltage amplitude of 10 mV. To examine the kinetic characteristics, the galvanostatic intermittent titration (GITT) technique was employed by a 30-minute charging process at 0.1C and a 50-minute duration as a resting period.

Published: December 17, 2025