The porous spherical Mn and Ti co-doped Li₂FeSiO₄/C cathodes material for lithium-ion batteries

Abstract

Li₂Fe_1-x-yMn_xTi_ySiO₄/C (x = 0.05~0.20, y = 0.02~0.08), a porous spherical cathode material for lithium-ion batteries, has been prepared by combining the sol-gel, spray drying, and microwave synthesis method, and co-doped Li₂FeSiO₄ with the Mn²⁺ and Ti⁴⁺on the iron site. The effects of Mn²⁺ and Ti⁴⁺ co-doping on the crystal structure, microstructure, and electrochemical performance Li₂FeSiO₄ were investigated by X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and electrochemical tests. The results showed that the modified sample (Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C), when x and y are 0.15 and 0.04, respectively, has the higher specific capacity and better cycle stability as well as lower impedance and better reversibility. Its electronic conductivity reached 6.85 × 10⁻⁴ S/cm, lithium-ion diffusion coefficient reached 1.65 × 10⁻¹³ cm²/s, and initial charge/discharge–specific capacity reaches 197.7/187.8 mAh/g and high capacity retention of 78% after 20 charge/discharge cycles at 2 C. According to the study, the Li₂FeSiO₄/C cathodes material of Mn and Ti co-substitution at the Fe site could partly make up the disadvantage of single Mn doping, and might provide an effective guide for the dopant incorporation to Li₂FeSiO₄ systems.

Introduction

Silicate polyanion compound materials of lithium electrodes Li₂MSiO₄ (M = Co, Fe, Ni, Mn, etc.) are considered as alternative cathode materials for the next-generation lithium-ion battery (LIB), owing to their competitive advantages such as high theoretical capacity exceeding 330 mAh/g, safety, highly stability, environment friendly, abundant natural resources, and so forth [1,2,3,4,5,6,7]. Among several polyanionic compounds, orthorhombic metasilicate crystal (Pnm2₁, Pnmb, and P2₁/n), Li₂FeSiO₄ has become a hot spot of orthosilicate cathode materials because of some disadvantages of other materials, such as low electrochemical activity for Li₂CoSiO₄ [8, 9] and rapidly decreased capacity for Li₂MnSiO₄ and Li₂NiSiO₄ during charging and discharging [10,11,12]. However, the actual power density of Li₂FeSiO₄ (LFS) is low due to its low electronic conductivity and poor lithium ion mobility [13,14,15]. In order to solve these issues, researchers have made a number of efforts to enhance the electrochemical performance of Li-ion batteries such as carbon coating, [16, 17] cationic doping, [4, 6, 7, 18,19,20,21] and nano-crystallization [22]. For example, synthesized Li₂FeSiO₄/C/GO with citric acid and GO as carbon sources by Huang [23] et al. exhibited the better electrochemical properties than Li₂FeSiO₄, whose initial discharge capacity was 155 mAh/g and capacity retention achieved 96.8% after 30 cycles at 5 C. Nevertheless, carbon coating only increases the ionic conductivity of Li₂FeSiO₄ particles high on the surface and between particles, but not within Li₂FeSiO₄ particles, which limit its practical application to a certain degree. In addition, Kokal [24] et al. found that the radius of Mn²⁺ is similar to that of Fe²⁺ ion, and appropriate amount of Mn²⁺ substituted for Fe²⁺ can form solid solution in Li₂FeSiO₄ crystal and achieving a higher storage capacity because Mn²⁺/Mn³⁺ can be oxidized to Mn³⁺/Mn⁴⁺ to enable Li₂Fe_1-xMn_xSiO₄ to achieve more than one Li⁺-ion per formula unit extraction/insertion at a wide potential window. However, this process was usually accompanied with serious capacity fade and drastic poor rate capability due to the structural instability of Li₂Fe_1-xMn_xSiO₄ caused by Jahn-Teller effect during the first few cycles and subsequent upon further cycling [25]. Fortunately, Zheng [26] et al. showed that the n-type doping introduced by Ti⁴⁺ can not only improve the conductivity of Li₂FeSiO₄, but also enhance the coupling between SiO₄ and FeO₄ tetrahedrons through the hybridization of d-orbital electrons, so as to enhance the stability of cathode materials.

In this article, to overcome drastic capacity fade and poor rate capability of Mn-doped Li₂FeSiO₄, we report a porous spherical cathode material, which are Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C synthesized by sol-gel, spray drying, and microwave synthesis method for lithium-ion batteries. And flat charge/discharge plateau accompanied by the extraction/insertion of more than one Li⁺-ions in a narrow potential window 1.5–4.8 V for Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C (polymorph with orthorhombic structure, P2₁/n space group) was obtained [27]. Therefore, the co-doped Li₂FeSiO₄ with the Mn²⁺ and Ti⁴⁺on the iron site can not only show an obvious improvement in specific capacity, but also have some positive effect on the rate and cycling performance [28]. In addition, X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), atomic force microscope (AFM), and electrochemical tests were used to investigate systematically the effects of Mn²⁺ and Ti⁴⁺ co-doping on the crystal structure, morphology, and electrochemical performance.

Experimental

Synthesis of Li₂Fe_1-x-yMn_xTi_ySiO₄/C composite

The synthesis of Li₂Fe_1-x-yMn_xTi_ySiO₄/C, which is porous spherical, was carried out by a novel method of sol-gel, spray drying, and microwave synthesis. The main product was 0.04 M Li₂Fe_1-x-yMn_xTi_ySiO₄/C. It was prepared as follow: Firstly, dispersing carbon source (starch) in de-ionized water to obtain an aqueous solution with a carbon content of 6 wt.% in a beaker, which 6 wt.% was the theoretical carbon content of Li₂Fe_1-x-yMn_xTi_ySiO₄/C (the actual coated carbon content of is 5. 47 wt.%). Secondly, taking iron source (FeNO₃·9H₂O) and manganese source (Mn (CH₃COO)₂·4H₂O) (the molar mass of Fe and Mn ratio is 1-x:x, x = 0.05, 0.10, 0.15, 0.20) blended by mechanical stirring, then added 0.08 M lithium source (CH₃COOLi) and 0.2 g of surfactant (PVP) with continually mechanical stirring, followed by adding 0.04 M silicon source (SiO₂-sol) and a specific amount of titanium source (nano-TiO₂) (the molar mass of Fe, Mn, and Ti ratio is 1-x-y:x:y, y = 0.02, 0.04, 0.06, 0.08), blended by sonication at room temperature for 20 min. Thirdly, the sol mixture was turned into a sphere morphology, by spray drying with the inlet temperature at 200  C and out let temperature at 100 °C. Then, the precursor was dried in vacuum at 110 °C for 8 h and pressed into piece with tablet press (30 MPa). Finally, the piece was placed in a tube furnace and heated at different heat treatment temperature and time in an argon atmosphere.

Physico-chemical characterization

The X-ray diffraction (XRD, Bruker/D2 PHASER) was used to characterize the phase composition of the materials in reflection mode with Cu-K_α1 radiation (λ = 1.5406 nm), in the 2θ range between 10° and 80° in steps of 0.02° with an integration time of 2 s per step. The morphology and structure were evaluated by a field emission scanning electron microscopy (FESEM, Quanta 450 FEG), a transmission electron microscopy (TEM, FEI Tecnai G2 F30), and an atomic force microscope (AFM, E-sweep). The actual carbon content was evaluated by an elemental analyzer (ELEMENTAR, Vario EL cube).

Electrochemical characterization

Electrochemical characterization was done to test the electrochemical performance of Li₂Fe_1-x-yMn_xTi_ySiO₄/C cathodes material, which was performed by CR2032 cion cell. The cathode electrode was prepared by blending 80 wt.% active material (Li₂Fe_1-x-yMn_xTi_ySiO₄/C), 10 wt.% conductive additives (Super-P), and 10 wt.% waterborne adhesive (LB132) with magnetic stirring to form a slurry. Then, we coated slurry on aluminum foil current collector via doctor blade processing, followed drying under vacuum at 110 °C for 12 h. After that, the electrode film was cut into circular disks, and the average loading of active material is 3 mg/cm². Afterward, all the cells were assembled in an argon-filled glove box (Super1220/990/750) with lithium metal as the counter electrode and a polypropylene microporous film (Celgaed2400) in the middle as separator. The electrolyte was a solution of 1 M LiPF₆, which dispersed in the mixture of DMC:EMC:EC (the ratio of 1:1:1 by volume). The cells were tested between 1.5 and 4.8 V (vs. Li⁺/Li) at room temperature by CT2001A Land battery testing system [29]. The cyclic voltammetry (CV) tests were carried out at room temperature by CH1604E electrochemical workstation with a scanning rate of 0.1 mV/s in the potential window of 1.5 V to 4.8 V. Electrochemical impedance spectroscopy (EIS) was measured at room temperature using CH1604E electrochemical workstation in the frequency range from 0.1 Hz to 100 kHz with an AC voltage of 5 mV.

Results and discussion

Effect of Mn²⁺ doping on Li₂Fe_1-xMn_xSiO₄/C performance

XRD patterns of Li₂Fe_1-xMn_xSiO₄/C (x = 0.00, 0.05, 0.10, 0.15, 0.20) samples are showed in Fig. 1. It can be seen that Li₂Fe_1-xMnxSiO₄/C doped with different Mn²⁺ amounts exhibit no impurity phase (PDF#01-077-4374), diffraction peaks were all relatively sharply, and the crystallinity was high. Above all, different Mn²⁺ doping dosage did not change the crystal structure of Li₂FeSiO₄/C, and Li₂Fe_1-xMn_xSiO₄/C still belonged to P2₁/n structure [30, 31]. Doping the dopant ions to replace the position of original ions will form a solid solution and change the unit cell parameters, and thus cause the shift of diffraction peak. When 2θ is 24.3°, this diffraction peak which matches the (111) crystal face of Li₂FeSiO₄ shows obvious shift to the left when doped Mn²⁺ replaced the position of Fe²⁺ in Li₂FeSiO₄ lattice. With increasing the mixing amount of Mn²⁺, more the shift could be found from the XRD patterns. Moreover, the lattice parameters were obtain by Cell-Refinement function of the Jade5.0 software during the analysis of the XRD results and are listed in Table 1. From Table 1, with the doped dosage of Mn²⁺change, the lattice parameters vary significantly. It can be observed a constantly increasement of the unit cell volume with the increased amount of Mn²⁺. As a result, the diffusion rate and coefficient of lithium ions were boosted by the increase of crystal cell volume, which can expand the diffusion channel of lithium ions and reduce the resistance of lithium ion diffusion.

Fig. 1

XRD patterns of Li₂Fe_1-xMn_xSiO₄/C (x = 0.00, 0.05, 0.10, 0.15, 0.20)

Table 1 Lattice parameters of Li₂Fe_1-xMn_xSiO₄/C samples

The SEM has been adopted to detect the morphology of the Li₂Fe_1-xMn_xSiO₄/C composites, and images of the Li₂Fe_1-xMn_xSiO₄/C composites with the amount of doped Mn²⁺ range from x = 0.05 to x = 0.20 are shown in Fig. 2. As shown in Fig. 2, all the samples present similar morphology of porous spherical particles and show no obvious agglomeration. The particle size distribution of different samples has also been detected and listed in Table 2. It can be seen from Table 2 that particle size distribution (D₁₀, D₅₀, and D₉₀) and tap density of the as-synthesized Li₂Fe_1-xMn_xSiO₄/C exhibit little wave range with the amount of Mn²⁺ doped vary. The result proved that the amount of Mn²⁺ doping has no effect on the morphology, particle size distribution, and tap density of the synthesized Li₂Fe_1-xMn_xSiO₄/C.

Fig. 2

SEM images of Li₂Fe_1-xMn_xSiO₄/C samples: ax = 0.05, bx = 0.10, cx = 0.15, dx = 0.20

Table 2 Particle size distribution of Li₂Fe_1-xMn_xSiO₄/C samples

Figure 3 shows the typical charge/discharge curves of the cathode with the Li₂Fe_1-xMn_xSiO₄/C (x = 0.05, 0.10, 0.15, 0.20) at 1/16 C. It figures that Li₂Fe_1-xMn_xSiO₄/C synthesized by different dosage of doped Mn²⁺ (x = 0.05, 0.10, 0.15, and 0.20) reveals the initial charge/discharge capacity of 160.4/145.5 mAh/g, 171.1/152.2 mAh/g, 194.8/171.7 mAh/g, and 177.6/128.1 mAh/g, respectively. And the corresponding coulomb efficiency values were 90.7%, 89.0%, 88.1%, and 72.1%, respectively. With the increase of Mn²⁺ content, the charge/discharge capacity of Li₂Fe_1-xMn_xSiO₄/C increased initially and then decreased. When x > 0.05 (x = 0.10, 0.15, and 0.20), the Li₂Fe_1-xMn_xSiO₄/C reveals higher charge capacity than 166 mAh/g. During the charging process, Mn²⁺ is oxidized to Mn⁴⁺ and the incorporation of Mn²⁺ into Li₂FeSiO₄/C can incorporate more than one Li⁺ [27]. When x = 0.20, Li₂Fe_1-xMn_xSiO₄/C owns the lowest specific capacity and coulomb efficiency. After Mn²⁺/Mn³⁺ is oxidized to Mn³⁺/Mn⁴⁺, it is not reduced to Mn²⁺/Mn³⁺ and combined with O²⁻ to generate MnO₄ tetrahedron to restore to the original vacancy [30].

Fig. 3

Charge/discharge curves of Li₂Fe_1-xMn_xSiO₄/C (x = 0.05, 0.10, 0.15, 0.20)

In Fig. 4, the curves show the cycling performance of Li₂Fe_1-xMn_xSiO₄/C cathode at 0.1 C. The Li₂Fe_1-xMn_xSiO₄/C synthesized by different Mn²⁺/Mn³⁺ quantity (x = 0.05, 0.10, 0.15, and 0.20) delivers a specific capacity of 131.8 mAh/g, 136.1 mAh/g, 147.8 mAh/g, and 93.8 mAh/g, respectively. And the related capacity retention ratio was 93.7%, 92.6%, 89.8%, and 76.2%, respectively. During the cycling process, Li₂Fe_1-xMn_xSiO₄/C synthesized with different amounts of Mn²⁺/Mn³⁺ exhibit different degrees of capacity decay, and Li₂Fe_0.8Mn_0.2SiO₄/C decays the most seriously. Since Li₂Fe_0.8Mn_0.2SiO₄/C has the highest content of Mn²⁺/Mn³⁺, it has a high chemical activity to be oxidized into Mn³⁺/Mn⁴⁺, and it may react with electrolyte but not with O²⁻ to form tetrahedron, which results in the material maintaining a certain structure. Therefore, the Li⁺ fails to be re-embedded into Li₂Fe_0.8Mn_0.2SiO₄/C after extracted. Capacity decreases as the number of circle increases, and the number of Li⁺ that can insert into Li₂Fe_0.8Mn_0.2SiO₄/C has decreased [32].

Fig. 4

Cycling performance of Li₂Fe_1-xMn_xSiO₄/C (x = 0.05, 0.10, 0.15, 0.20) samples

Effect of Ti⁴⁺ doping on Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C performance

XRD patterns of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C (y = 0.02, 0.04, 0.06, 0.08) are shown in Fig. 5. According to the XRD result, Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C doped by different Ti⁴⁺ quantities all belong to P2₁/n structure (PDF#01-077-4374) [29, 30]. Their diffraction peaks are sharp and with high crystallinity. The impurities of Fe, Mn, Ti, and Ti-compounds did not appear in the XRD patterns, indicating that different Ti⁴⁺ quantities have been doped into the lattice of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C. With the increase of Ti⁴⁺ doping, the diffraction peak of crystal plane (111) shifted significantly to the left, because the addition of Ti⁴⁺ to Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C lattice changes the cell parameters. As shown in Table 3, the crystal lattice volume of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C increased by raising the Ti⁴⁺ dosage.

Fig. 5

XRD patterns of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C (y = 0.02, 0.04, 0.06, 0.08)

Table 3 Lattice parameters of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C samples

Figure 6 shows the SEM diagram of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C (y = 0.02, 0.04, 0.06, 0.08). According to the images, the morphology of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C shows no change with different amounts of Ti⁴⁺. It can be seen from Table 4 that theTi⁴⁺ doping has no influence on the particle size distribution and tap density of the synthesized Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C.

Fig. 6

SEM images of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C samples: ay = 0.02, by = 0.04, cy = 0.06, dy = 0.08

Table 4 Particle size distribution of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C samples

In Fig. 7, the curves show that the Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C (y = 0.02, 0.04, 0.06, 0.08) charge/discharge at 1/16 C. It can be seen that the initial charge/discharge capacity of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C with different Ti⁴⁺ content (y = 0.02, 0.04, 0.06, and 0.08) was 191.3/176.7 mAh/g, 197.7/187.8 mAh/g, 194.7/177.7 mAh/g, and 186.8/155.4 mAh/g, respectively. And the coulomb efficiency was 92.4%, 95.0%, 91.3%, and 83.2%, respectively. With the increase of Ti⁴⁺ content, the charge-discharge capacity of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C increased firstly and then decreased. When y = 0.04, a discharge specific capacity of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C exhibits 187.8 mAh/g, which is equal to 1.13 Li⁺-ion de-intercalate from the crystal lattice. In addition, the coulomb efficiency reaches to 95.0%, suggesting that the incorporation of Ti⁴⁺ into the tetrahedron may make the Mn²⁺/Mn³⁺ and the MnO₄ tetrahedron generated by O²⁻ return to the original vacancy [27].

Fig. 7

Charge/discharge curves of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C sample: ay = 0.02, by = 0.04, cy = 0.06, dy = 0.08

In Fig. 8, the curves show the cycle performance of the Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C at 0.1 C. It can be seen that the Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C synthesized by different mixing Mn²⁺ quantity y = 0.02, 0.04, 0.06, and 0.08 delivers a specific capacity of 168 mAh/g, 177.3 mAh/g, 159.2 mAh/g, and 138.1 mAh/g, respectively. The capacity retention ratio was 97. 8%, 98.3%, 96.0%, and 92.7%, respectively. The cycling stability of Li₂Fe_0.85-yMn_0.15 Ti_ySiO₄/C improved firstly and then decreased with the increase of Ti⁴⁺ dosage. Therefore, the addition of appropriate amount of Ti⁴⁺ into Li₂Fe_0.85-yMn_0.15 Ti_ySiO₄/C not only stabilizes the crystal structure, but also enhances cyclic performance of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C. And furthermore, the synergistic doped Ti⁴⁺ and Mn²⁺ improves the specific capacity of the material.

Fig. 8

Cycling performance of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C (y = 0.02, 0.04, 0.06, 0.08)

Characterization of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C

The electrochemical performance of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C is the most outstanding, when the Fe, Mn, and Ti mole ratio of Li₂Fe_1-x-yMn_xTi_ySiO₄/C is 1-x-y:x:y = 0.81:0.15:0.04. Figure 9 shows the SEM diagram of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C. The figure shows that thickness of carbon coating layer on Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄ is about 15 nm, indicating Mn²⁺ and Ti⁴⁺ doped Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C after no change of starch of pyrolytic carbon Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄ coating effect.

Fig. 9

Morphology characterization of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C sample. a TEM images of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C sample, b carbon-coated layer, and c HRTEM images

Figure 10 shows the rate and cycling performance of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode. The long-term cycling performance of the Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode has been performed at 0.5 C for 50 circles. The specific capacity is 161.8 mAh/g after 50 cycles at 0.5 C times. And the rate performance of the batteries has been tested by a gradual raising current density from 0.2 C to 2 C. A high specific capacity can be observed from the rate performance test. The Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode displays specific capacity of 172.4 mAh/g, 162.8 mAh/g, 147.1 mAh/g, and 126.0 mAh/g at 0.2 C, 0.5 C, 1 C, and 2 C, respectively. Moreover, when we lowered the current density to 0.2 C and cycled for 10 times, the Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode still shows a specific capacity of 172.5 mAh/g. Even when the current density went back to 0.5 C, the specific capacity could return to 161.8 mAh/g. This outstanding rate performance indicates the high electrical conductivity and the fast Li⁺-ion conductivity of the Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C composite. The electrical conductivity characterization of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C exhibits a result of 6.85 × 10⁻⁴ S/cm, thus three times higher than the Li₂FeSiO₄ coated by starch carbon. And the long-term cycling performance of the Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode has been performed at 0.5 C for 50 circles. The specific capacity is 161.8 mAh/g after 50 cycles at 0.5 C times. The improved long-term cycling performance indicates that co-doped appropriate Mn²⁺ and Ti⁴⁺ into Li₂Fe_1-x-yMn_xTi_ySiO₄/C could keep the crystal structure of the material from collapsing. Although pure Mn²⁺doped can enhance the specific capacity, the cycling performance of the batteries will be worse. Overall, the Mn²⁺ and Ti⁴⁺co-doped, which is a novel method, demonstrates a feasible way to achieve a stable structure and good cyclic performance of this kind Li₂FeSiO₄ material.

Fig. 10

Rate and cycling performance of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode

Figure 11 is CV curves for the Li₂FeSiO₄/C and Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode at initial three circles. The scanning rate was 0.1 mV/s in the potential window of 1.5–4.8 V. During the negative scanning process, the reduction peaks exhibits lithium implantation behavior concentrated at about 2.4 V, with good repeatability. During the scanning circle from second and third, the peak shifts to the left and increases the area. This is the structural rearrangement to a more stable structure after the first cycle, which means the amount of lithium embedded increases. In the forward scanning process, the oxidation peak Li₂FeSiO₄/C, which is manifested as lithium removal behavior, has a “transition” weak peak at 3.6 V in addition to the main peak type of 3.3 V, which can be attributed to the oxidation of Fe²⁺ to Fe³⁺ [15, 33,34,35]. The Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C obtained an overlapped oxidation peak curve at about 3.3 V, thus implying the stable structure during cycling. Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C with high potential of oxidation peak is a slowly weakening trend, suggesting that Mn²⁺ and Ti⁴⁺ doped are conducive to oxidation reaction of Fe³⁺/Fe⁴⁺, Mn³⁺/Mn⁴⁺, and Ti³⁺/Ti⁴⁺ [14, 36,37,38,39]. Thus, we speculated that the Mn²⁺- and Ti⁴⁺-doped Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C have increased structural stability during de/intercalation.

Fig. 11

The CV profiles of a Li₂FeSiO₄/C and b Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode

As shown in Fig. 12, the electrochemical impedance spectra (EIS) have been conducted to illustrate the electrochemical performance of Li₂FeSiO₄/C and Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode after cycled 50 times at 0.5 C. The charge transfer resistance Rct in the high-frequency areas of Li₂FeSiO₄/C and Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode is 130 and 35 Ω, respectively. The slope of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C low-frequency area is significantly greater than that of Li₂FeSiO₄/C. This indicated that Mn²⁺ and Ti⁴⁺ co-doped increased the conductivity and significantly reduced the electrochemical impedance caused by the charge transfer on the surface of the electrode. According to formula (1) and (2), Li⁺ diffusion rate of the Li₂FeSiO₄/C and Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode could be calculated, [40, 41] where D is the diffusion coefficient of lithium ion, R is the gas constant (8.314 J/K mol⁻¹), T is the room temperature in our experiment (298 K), A is the surface area of the electrode (1.13 cm²), n is the number of the electrons per molecule attending the electronic transfer reaction which is 1, F is the Faraday constant (96,500 C/mol), C is the concentration of lithium ion in electrolyte (0.039 mol/cm³ for Li₂FeSiO₄) [38]. R_D is the ohmic resistance, R_L is the charge transfer resistance, ω is the angular frequency in the low-frequency region, and σ is the Warburg factor, which can be obtained from the slopes of linear fitting of Z″ versus ω^-1/2 plots in Fig. 12.

$$ D=\frac{R^2{T}^2}{2{A}^2{n}^4{F}^4{C}^2{\sigma}^2} $$

(1)

$$ {Z}_{\mathrm{re}}={R}_{\mathrm{D}}+{R}_{\mathrm{L}}+\sigma {\omega}^{-1/2} $$

(2)

Fig. 12

Electrochemical impedance spectra of a Li₂FeSiO₄/C and b Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C cathode

The obtained σ value of Li₂FeSiO₄/C and Li₂Fe_0.80Mn_0.15Ti_0.05SiO₄/C is 31.5 Ω s^-1/2 and 29.7 Ω s^-1/2, respectively. Finally, the Li₂Fe_0.80Mn_0.15Ti_0.05SiO₄/C has fast Li⁺ diffusion rate of 1.65 × 10⁻¹³ cm²/s and Li₂FeSiO₄/C of 1.47 × 10⁻¹³ cm²/s, which is consistent with the outstanding rate performance of Li₂Fe_0.80Mn_0.15Ti_0.05SiO₄/C. As confirmed by the diffusion coefficient, the Li₂Fe_0.80Mn_0.15Ti_0.05SiO₄/C cathode showed higher lithium ion conductivity. It could be attributed to that the co-doped Mn²⁺ and Ti⁴⁺ increase the cell volume and extending the Li⁺ diffusion path in and out of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C particles [6, 26, 42].

To study the effect of structural stability of Li₂FeSiO₄/C and Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C composite electrode during cycling, SEM and AFM were used to observe the surface morphology of the electrode before and after cycling (Fig.S1 and S2). It can be seen that the electrode surface becomes smooth after charge/discharge testing, and the morphology has no obvious change. The Li₂FeSiO₄/C and Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C can still maintain spherical structure after cycling [43].

We also have compared our results with some recent reported heteroatom-doped Li₂FeSiO₄/C materials as shown in Table S1. The result demonstrates that the Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C with porous spherical structure prepared by combining the sol-gel, spray drying, and microwave synthesis method shows a higher initial discharge specific capacity and faster lithium ion diffusion rate.

Conclusions

Porous spherical Li₂Fe_1-x-yMn_xTi_ySiO₄/C (x = 0.05~0.20, y = 0.02~0.08) composite has been successfully prepared via a novel method of sol-gel, spray drying, and microwave synthesis for cathode material. This method carried out the bulk phase modified by co-doping Mn²⁺ and Ti⁴⁺ on the iron site and a thick layer carbon coating at the same time. Co-doping Ti⁴⁺ into Li₂Fe_0.85Mn_0.15SiO₄/C can make Mn²⁺ and O²⁻ form tetrahedron again and return to the original vacancy to avoid the collapse of the crystal structure, and thus improves the stability of the material. In addition, Ti⁴⁺ can work with Mn²⁺ to increase the specific capacity of Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C. When mole ratio of Fe and Ti in Li₂Fe_0.85-yMn_0.15Ti_ySiO₄/C is 0.85-y:y (y = 0.04), Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C has a higher specific capacity and better cycle stability. The initial discharge specific capacity was 187.8 mAh/g, which is equal to 1.13 Li⁺ from Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C embedded in lattice. After cycling for 10 times at 0.2 C, the specific capacity was 172.5 mAh/g, and 161.8 mAh/g after cycled for 50 times at 0.5 C. In addition, by co-doping Mn²⁺ and Ti⁴⁺, the electric conductivity of Li₂Fe_0.81Mn_0.15Ti_0.04SiO₄/C increased to 6.85 × 10⁻⁴ S/cm. Most importantly, co-doped appropriate Mn²⁺ and Ti⁴⁺ into Li₂Fe_1-x-yMn_xTi_ySiO₄/C could keep the stability of crystal structure from collapsing, and point out a feasible way to achieve a stable structure and good cyclic performance of Li₂FeSiO₄ material.