1. Introduction
Rechargeable lithium-ion batteries (LIBs) have rapidly integrated into various facets of our daily lives over the past decades, particularly as portable energy storage solutions, owing to their high energy density and extended cycle life. Consequently, extensive research has been dedicated to LIBs to address the growing demand for enhanced lithium utilization and specific energy density [
1]. Since the cathode materials employed in LIBs significantly underpin their performance, a large amount of research has concentrated on advanced cathode materials [
2]. Since it was first introduced in 1997, olivine-structured lithium iron phosphate (LiFePO
4) has emerged as one of the most widely used cathode materials employed in rechargeable lithium-ion batteries [
3]. It has been extensively studied due to its affordability, abundant raw material supply, a high theoretical capacity of 170 mAh/g, superior cycle performance, and remarkable stability at elevated temperatures [
4]. However, it encounters the challenge of low intrinsic conductivity for both lithium ions and electrons, leading to its poor performance at high rates and low temperatures [
5]. To address this limitation, two strategies have been explored extensively in recent years. Firstly, particle nanosizing has been implemented to reduce the diffusion distance, enhancing the efficiency of ion and electron transfer. Additionally, surface-coated carbon has been widely recognized as an effective approach to extend the surface area available for the interfacial transfer of conductive particles and, thus, improve the low electronic conductivity of LiFePO
4 [
6,
7].
In recent years, carbon coated LiFePO
4 with the nanostructure (Nano-LiFePO
4/C) has been prepared using various physical and chemical synthesis routes. The physical solid-state method, which has developed over a few decades, stands out as the most prevalent technique for producing Nano-LiFePO
4/C [
8,
9,
10,
11]. However, the solid-state approach is associated with several shortcomings: (1) The preparation process includes a series of complex procedures, including high-energy milling, spray drying, calcination, and grinding. Not only does this result in costly equipment requirements, but it also leads to a huge waste of energy and time. (2) Given that the raw materials are a heterogeneous mixture of solids, it makes it difficult to achieve uniform composition [
12]. (3) The particle size is inherently related to the size of the initial raw materials, and in the solid-state method the final product often yields particles larger than 500 nm. These limitations significantly hamper its application in high-rate and low-temperature batteries. Chemical methods such as the hydrothermal/solvothermal method, the sol–gel method, the spray pyrolysis method, high mixing continuous rotating reactor technology (HMCRR), and the solution combustion method were notable for their high homogeneity and small particle size (~100 nm), which is difficult to achieve in the solid-state method [
11,
13,
14]. Among these solution synthesis routes, solution combustion synthesis (SCS) stands out as a novel and efficient approach for the preparation of nanoscale materials. In this method, a strongly exothermic chemical reaction between an oxidizer (generally metal nitrites) and organic fuels in an aqueous solution is utilized. After further calcination with a carbon source, such as glucose, ultrafine Nano-LiFePO
4/C nanoparticles were produced [
15]. Solution combustion synthesis (SCS) possesses several noteworthy characteristics: (1) The SCS reaction finishes in a short time and a portion of the heat arises from the self-propagating reaction, rendering it an energy-efficient and time-saving methodology. (2) The raw materials are intimately mixed in an aqueous solution at the molecular level, ensuring a high level of homogeneity in the resulting product. (3) SCS yields products characterized by a high specific surface area and a finely structured nanoarchitecture, achieved through the generation of large volumes of gases. These attributes significantly enhance the electrochemical performance of LiFePO
4/C [
16].
In previous studies, Nano-LiFePO
4/C has been synthesized with a single fuel such as sucrose, glycine, urea, L-Lysine, or glucose [
15,
17,
18,
19,
20]. However, the electrochemical performance of the obtained LiFePO
4/C was far from satisfactory due to the large particle size and particle agglomeration. Recently, a mixture of organic fuels with different functions has been verified to be useful in modifying the performance of LiFePO
4/C using SCS. Nano-LiFePO
4/C prepared with CTAB-based mixed fuels showed an ultrafine structure and a specific capacity of 137 mAh/g at 0.1 C [
21,
22]. In the mixed fuels, CTAB played the role of a cationic surfactant and performed as a soft template during the gelation process, promoting control over the final combusted products, while citric acid or glycine were used for gelation, leading to less agglomeration and high homogeneity. However, fuels such as CTAB and glycine are expensive materials, which makes it difficult to achieve large-scale fabrication [
23,
24].
In this study, inexpensive and easily accessible urea–sorbitol mixed fuels have been developed to prepare Nano-LiFePO4/C using SCS, where sorbitol was used as a complexing agent for Fe3+ to avoid precipitation and decomposable urea served to prevent the agglomeration of particles. The effect of mixed fuels on the combustion behavior, microstructure, and electrochemical performance of Nano-LiFePO4/C was studied in detail.
3. Results and Discussion
Figure 2a shows the typical temperature profile of the mixed solution during the solution combustion process, which shows the characteristics of the volume combustion mode [
26]. The entire solution was initially heated uniformly with the evaporation of free and part of bond water resulting in a mild temperature change (Stage Ⅰ). Subsequently, the temperature increased at a higher rate following serious bumping (Stage Ⅱ). When the temperature reached ignition temperature (
Tig~135 °C), it increased sharply to the maximum temperature (
Tm = 459 °C) and was accompanied by the release of large amounts of gases (N
xO, CO
2, H
2O, and NH
3) and rapid volume expansion (Stage III), but no flame was observed. After the cooling stage (Stage Ⅳ), foamy as-combusted samples were obtained.
Figure 2b presents the TG/DSC curves of the mixture during solution combustion. The evaporation of water and the decomposition of urea and Fe (NO
3)
3 should be responsible for a constant loss in weight of about 60% and a series of endothermic peaks on the DTA curve before 240 °C. Accompanied with a strong exothermic reaction between the oxidizer and the organic fuels, the mixture was rapidly dried. At the same time, a sudden mass loss of around 20 wt% happened at around 246 °C, with large amounts of gas and thermal energy released. At the same time, the temperature of the mixture reached the maximum. After the combustion reaction, weight loss still continued owing to the slow decomposition of residual organic fuels until all fuels were consumed [
27]. Hence, the rapid change in temperature could be considered as an indication of the occurrence of the solution combustion reaction. Time–temperature curves for the SCS of the solution using varying amounts of urea and sorbitol were shown in
Figure 2c,d. It is evident that the fuel to oxidant ratio and the composition of the mixed fuels influence the maximum temperature as well as the combustion behavior. Notably, in the sample only containing urea as fuel (U
1.6S
0), no temperature peak was observed, indicating a combustion reaction had not occurred with single urea under the fuel-lean condition. According to combustion theory,
Tm was determined using the initial temperature (
T0), the heat of combustion (
Qc), and the heat dissipation (
Qd):
where
Cp is the heat capacity at a constant pressure. In the same reaction system, larger
Qc and smaller
Qd will lead to a higher
Tm [
15]. Except for U
1.6S
0, samples show similar time–temperature curves. Regarding U
1.6S
y, the
Tm increased with an increasing amount of sorbitol, suggesting a more vigorous reaction (larger
Qc) was happened. Meanwhile, the
Tm tended to decrease with the amount of urea as more heat was taken by the endothermic decomposition reaction of the urea (larger
Qd).
The XRD patterns of the powders before and after calcining were presented in
Figure 3a. Before calcining, the XRD was noisy due to its poor crystallinity and the powders were composed of phases such as Li
3Fe
2(PO
4)
3 (JCPDS Card No. 47-0107), Fe
2O
3 (JCPDS Card No. 85-0987), and P
2O
5 (JCPDS Card No. 23-1301). During the high-temperature calcining process in the N
2, reducible activated carbon was generated from the pyrolysis process of glucose. And Fe (III) phases were subsequently reduced and reacted to form LiFePO
4 with Fe (II) as per the following possible reactions. As a result, LiFePO
4/C with carbon coating was produced.
Figure 3b presents the Rietveld refinement results for the U
1.6S
0.8 sample. All reflections were successfully indexed based on the orthorhombic LiFePO
4 crystal structure within the Pnma space group, and no impurities were detected. Furthermore, a carbon phase was not found for its amorphous structure. The obtained lattice parameters (a = 10.331 Å, b = 6.008 Å, and c = 4.694 Å) are consistent with those reported in reference [
28].
Figure S1 illustrates the XRD patterns of Nano-LiFePO
4/C samples after being calcined at 750 °C for 6 h. All the diffraction peaks of the samples can be indexed according to orthorhombic crystal structure and no obvious impurity phases were observed, suggesting the successful preparation of single-phase LiFePO
4/C (JCPDS Card No. 81-1173). Carbon content testing was conducted using a carbon and sulfur combined tester, and the results are presented in
Table 1. The carbon in the Nano-LiFePO
4/C composite material originates from two sources: (1) 20 wt% glucose added to the as-combusted powders before calcination and (2) residual impurities containing carbon from the excess urea–sorbitol mixed fuels.
Figure S2 illustrates the thermogravimetric curve of urea–sorbitol mixed fuels in an air atmosphere. The mixed fuel initiates decomposition at approximately 200 °C, with 10.6% and 4.1% residue remaining at 350 °C and 500 °C, respectively. In our study, an excess of fuels (
φ > 1) was employed. Despite subjecting the materials to a further high-temperature treatment at 500 °C, there was still organic content from excess fuel residues present in the LiFePO
4/C samples. When
y = 0.8, U
0.8S
0.8, U
1.2S
0.8, and U
1.6S
0.8 exhibited a relatively low carbon content, each with less than 3 wt%, indicating that the urea content had no noticeable impact on the carbon content as the excess urea pyrolyzed at high temperatures. Conversely, with
x = 4, the samples displayed an increased carbon content, corresponding to the rising residual content resulting from the increased usage of mixed fuels.
The SEM images of sample U
1.6S
0.8 before and after calcination are shown in
Figure 4a,b, respectively. Before calcining, a highly porous structure could be observed and the precursor exhibited well-distributed spherical particles with a mean particle size of less than 100 nm. The release of numerous gases and the rapid reaction rate during the combustion process impede particle growth, resulting in the nanoparticles and the formation of pores within the resulting powder. After calcining at 750 °C, an obvious increase in particle size was observed due to particle sintering and growth at high temperatures. SEM images of Nano-LiFePO
4/C samples with different fuel amounts are presented in
Figure 5. As the combustion reaction had not occurred, the sample prepared only using urea showed a bulky and dense microstructure. For other powders prepared with different content of fuels, they showed a porous and foamy structure as a result of the large amounts of gases released during the combustion reaction. The maximum temperature (
Tm) and mean particle size of the samples are summarized in
Table 2. With the increasing amount of sorbitol, the reaction between the nitrate and the fuels released more heat, resulting in a higher
Tm. The particle distribution curves are shown in
Figure S3. Consequently, the particle size increased with the amount of sorbitol for particle growth and sintering at higher temperatures. As a contrast, particle size as well as T
m decreased with the amount of urea because a higher level of gas release effectively reduced the maximum temperature and aggregation.
Figure 4c,d present the HRTEM images of U
1.6S
0.8. Aggregation with particles of a size less than 150 nm could be observed. Nano-LiFePO
4/C particles were covered by continuous carbon layers, which could improve the surface stability and enhance the conductivity between particles [
29].
The N
2 adsorption–desorption isotherms and pore size distribution plots of samples prepared using mixed fuels are presented in
Figure 6. The isotherm curves show H3 hysteresis on IV type isotherms with foamy microstructure [
30] and the samples possess a mesopore nature distributed in the range of 2–50 nm. Samples prepared using SCS showed a high specific surface area and a highly porous structure. It is clearly demonstrated that the surface area and pore volume varied with the amount of urea/sorbitol. As the content of sorbitol increased, both the specific surface area and pore volume decreased due to the higher
Tm and larger particle size. On the other hand, the specific surface area and pore volume tended to increase with the content of urea, which was attributed to the higher amount of gas released and the smaller particle size. As a result, sample U
1.6S
0.8 showed the highest specific surface area (42.01 m
2/g) and pore volume (0.146 cm
3/g) due to its lowest
Tm and smallest particle size.
The initial charge–discharge curves of samples at a rate of 0.1 C s are compared in
Figure 7a,d with the voltage range of 2.0 to 3.8V. All profiles displayed a similar smooth and monotonous charge–discharge voltage ramp with a plateau of around 3.4 V owing to the phase transition between LiFePO
4 and FePO
4. Sample U
1.6S
0 delivered the lowest discharge capacities for its bulky microstructure. For sample U
1.6S
y, they showed discharge capacities of 153.5, 132.2, and 122.7 mAh/g, respectively, which decreased with the increasing content of sorbitol. The deterioration of capacity was attributed to the increasing particle size resulting from increasing the
Tm [
12]. For sample U
xS
0.8, samples showed discharge capacities of 128.9, 131.2, 138.7, 144.9, and 153.5 mAh/g, respectively. The discharge capacities of the samples increased with the content of urea due to the reduction in particle size caused by decreased the T
m. The rate performance of samples is displayed in
Figure 7b,e. Additionally, as shown in
Figure 7c,f, except for U
1.6S
0, samples prepared using SCS with urea–sorbitol mixed fuels exhibited a very stable discharge capacity with more than 93% capacity retention after 200 cycles at a relative high rate of 1 C. As summarized in
Table 3, sample U
1.6S
0.8 with the least sorbitol and the most urea showed the highest capacity, which was 153.5, 149.9, 147.6, and 142.9 mAh/g at 0.1, 0.3, 0.5, and 1 C, respectively. This enhanced performance can be attributed to several key factors: 1. An increased electrolyte contact area: a larger particle surface area facilitates more extensive contact between the Nano-LiFePO
4/C and the electrolyte, leading to an improved electrode–electrolyte interface, which promotes an efficient charge transfer [
31]. 2. Enhanced lithium utilization: smaller particle sizes allow a higher proportion of the LiFePO
4 material to actively participate in electrochemical reactions. This greater lithium utilization contributes to increased energy storage capacity and enhanced overall performance [
32]. 3. Improved cycling stability: smaller particles are less susceptible to mechanical stress during the insertion and extraction of lithium ions. This enhanced mechanical durability results in improved cycling stability and a longer cycle life for the battery [
33]. The morphological changes in the Nano-LiFePO
4/C electrode were examined through SEM before and after undergoing 1 C cycling. As depicted in
Figure S4, prior to cycling, the electrode exhibited a compact and relatively flat surface. However, after 220 cycles of cycling at 1 C, the electrode’s surface became irregular and displayed discernible cracks. These changes were attributed to electrolyte erosion and the expansion/contraction of particles during the Li
+ intercalation and de-intercalation processes.
The resistances of the U
1.6S
y and the U
xS
0.8 electrodes before cycling were assessed and the resulting electrochemical impedance spectra (EIS) are displayed as Nyquist plots in
Figure 8 in the frequency range from 0.1 Hz to 100 kHz. All EIS profiles present a semicircular pattern in the high-frequency region and an inclined line in the low-frequency region. In accordance with the equivalent circuit model, the semicircle in the high-frequency region is related to the charge transfer resistance (R
ct). In the case of U
1.6S
y, the U
1.6S
0 sample showed the highest R
ct (242.4 Ω) due to its bulky microstructure. As y increased from 0.8 to 2.4, R
ct decreased from 139.3 Ω to 79.35 Ω, which is attributed to the reduction in particle size. For U
xS
0.8, R
ct exhibited a monotonic decrease as the quantity of urea used increased, which was primarily driven by the particle size reduction. The variations in R
ct offer a comprehensive explanation for the distinctions in electrochemical performance, indicating the favorable impact of a reduced particle size and a lower
Tm on the kinetics of Li
+ ion charge transfer. The electronic conductivity of the samples was exhibited in
Table S1. When
y = 0.8, the conductivity of the powders increased with the usage of urea due to their decreasing particle size. U
1.6S
1.6 and U
1.6S
2.4 exhibited lower resistivity, which was attributed to their higher carbon content. The diffusion coefficient of Li
+ ions was calculated using the Warburg factor (
σ) as shown below [
34]:
where
R is the gas constant,
T is the absolute temperature,
A is the surface area of the electrode,
n is the number of electrons,
F is Faraday’s constant,
C is the concentration of the lithium ions, and
σ is the Warburg factor. The calculation of D
Li+ was listed in
Table S2. D
Li+ tended to increase with the decrease in particle size and U
1.6S
0.8 showed the highest D
Li+. This result further illustrates that the particle size has a direct relationship with the Li
+ transport. A small particle size could facilitate the transport of Li
+, leading to the improvement of electrochemical performance.
To assess the performance of Nano-LiFePO
4/C synthesized using a combination of urea and sorbitol as mixed fuel, a comparative analysis of the electrochemical performance of sample U
1.6S
0.8 at various discharge rates was conducted. The rate performance of U
1.6S
0.8 at 0.2, 1.5, 1, 3, 5, and 10 C was shown in
Figure S5. We compared this sample with SCS samples prepared using alternative fuel in the existing literature. As displayed in
Figure 9, U
1.6S
0.8 exhibits the highest discharge capacity when subjected to discharge rates ranging from 0.2 C to 10 C, surpassing the performance of samples prepared with L-Lysine and CTAB-based fuels. Notably, U
1.6S
0.8 displays the least significant decrease in specific capacity as the discharge rate escalates from 0.2 C to 10 C, which is attributed to its diminutive particle size and high specific surface area. In comparison to commercial LiFePO
4/C synthesized through a solid-state method, samples produced via SCS with urea–sorbitol mixed fuels demonstrate a competitive electrochemical performance, showing strong potential for commercial applications in the future.