Electrochimica Acta 52 (2007) 6778–6783

Preparation and characterization of nano-sized LiFePO4 by low heating

solid-state coordination method and microwave heating
Lei Wang, Yudai Huang, Rongrong Jiang, Dianzeng Jia ∗
Institute of Applied Chemistry, Xinjiang University, Urumqi 830046, Xinjiang, People’s Republic of China
Received 10 January 2007; received in revised form 29 April 2007; accepted 30 April 2007
Available online 6 May 2007

Abstract
The precursors of LiFePO4 were prepared by low heating solid-state coordination method using lithium acetate, ammonium dihydric phosphate,
ferrous oxalate and citric acid as raw materials. Olivine phase LiFePO4 as a cathode material for lithium-ion batteries was successfully synthesized
by microwave heating in a few minutes. X-ray diffraction (XRD) and transmission electron microscope (TEM) were used to characterize its
structure and morphology. Cyclic voltammetry (CV) and charge–discharge cycling performance were used to characterize its electrochemical
properties. The results showed that the grain size of the optimal sample was about 40–50 nm, and the as-prepared particles were homogeneous.
The nano-sized LiFePO4 obtained has a high electrochemical capacity (125 mAh g−1 ) and stable cycle ability.
© 2007 Elsevier Ltd. All rights reserved.

Keywords: LiFePO4 ; Low heating solid-state coordination method; Microwave heating; Lithium-ion battery

1. Introduction to prepare LiFePO4 , such as co-precipitation procedure [12],

sol–gel preparation [13] and hydrothermal procedure [14].
First proposed by Padhi et al. [1], LiFePO4 has become a very However, these solution methods are rather complicated and
attractive cathode material for lithium-ion batteries due to its time-consuming.
advantages of environmental compatibility, low cost, excellent Low heating solid-state coordination method was believed
thermal stability and good cycling property. In addition, olivine to be a simple and effective one to fabricate a number of
LiFePO4 has a high theoretical capacity of 170 mAh g−1 with a chemical compounds, such as cluster compounds, coordination
flat discharge potential of 3.4 V versus Li/Li+ [2–4]. compounds, and solid-coordination compounds [15,16]. Specif-
The main problem of LiFePO4 lies in its poor rate capability, ically LiMn2 O4 cathode material was prepared by this method
which is attributed to its low electronic conductivity and slow [17–21].
kinetics of lithium ion diffusion through the LiFePO4 –FePO4 Nowadays, microwave heating technique is widely applied
interfaces [5,6]. Two approaches have been attempted to over- in chemistry. Unlike the conventional heating where the mate-
come this problem. One is to enhance its electronic conductivity rials are heated from outer surface to interior and thus result
by coating an electron-conducting layer around the particles or in steep thermal gradients, microwave heats the material on
by doping with guest cations [7–9]. The other is to minimize molecular level, leading to uniform heating. Ganesh et al.
the particle size by modifying synthesis conditions [2,10,11]. [22] reported that the microwave assisted combustion synthe-
However, the former approach requires extreme conditions and sis retained the nanocrystalline structure even after sintering,
expensive equipment, and the preparation of nano-LiFePO4 which was extremely useful in the synthesis of shaped prod-
(diameter < 100 nm) in the latter is very demanding. ucts for commercial applications. Higuchi et al. [23] introduced
To ensure homogeneous reactant distribution during heating microwave heating to the preparation of the LiFePO4 cathode
and pure crystallization, several techniques have been applied material. The difficulty of LiFePO4 synthesis with microwave
heating is the iron oxidation state, which could be controlled
by heating with the inert gas flow or by coating raw materials
∗ Corresponding author. Tel.: +86 991 858 0032; fax: +86 991 858 0032. with activated carbon or carbonaceous substance [24]. However,
E-mail address: [email protected] (D. Jia). the former way is discommodious, while the latter unavoid-

0013-4686/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2007.04.104
 L. Wang et al. / Electrochimica Acta 52 (2007) 6778–6783 6779

ably induces carbon residual. In this work, we improved the the lid. Container 3 was inverted to cover all these and form
method and solved these problems by a self-assembly carbon an insulated room. Activated carbon, as microwave absorber,
seal reactor. heated the precursors and helped to make reductive atmosphere
In this study, low heating solid-state coordination method was by carbon oxidation reactions.
adopted to prepare the precursors of LiFePO4 and microwave Thermal analysis was performed by thermogravimetric
heating was used to synthesize LiFePO4 . By these two sim- analysis (TGA) and differential scanning calorimetry (DSC)
ple methods, nano-sized LiFePO4 with electronically conductive (TGA–DSC, STA 449C, NETZSCH, Germany) at a heating rate
materials was obtained. The self-assembly carbon seal reactor of 20 ◦ C min−1 . Powder X-ray diffraction (XRD, MXP18AHF,
helped to ensure the purity of LiFePO4 . The whole process, per- MAC, Japan) using Cu K␣ radiation (λ = 1.54056 Å) was
formed without any solvent, was environmentally benign. The adopted to identify the crystalline phase of the resulting materi-
obtained LiFePO4 samples exhibit homogeneous microstructure als. The grain size and morphology of LiFePO4 were observed
and good electrochemical properties. using an accelerating voltage of 100 kV transmission electron
microscope (TEM, H-600, Hitachi, Japan).
2. Experimental LiFePO4 was thoroughly mixed with 5% polyvinyllidene flu-
oride and 15% acetylene black in N-methylpyrrolidene. The
Four LiFePO4 samples were prepared by low heating solid- obtained slurry was then brushed on an aluminum foil substrate
state coordination reaction and microwave heating process. and dried in a vacuum oven at 120 ◦ C for 4 h. Lithium foil was
Stoichiometric NH4 H2 PO4 and CH3 COOLi were mixed in an used as anode and Celgard 2400 as the separator. The electrolyte
agate mortar and ground with a pestle for 0.5 h in order to obtain was a mixture of 1 M LiPF6 -ethylene carbonate (EC)/dimethyl
homogeneous mixture. FeC2 O4 ·2H2 O and suitable amount of carbonate (DMC) (1:1, v/v). The cells were assembled in an
citric acid (in three samples) were added and ground for another argon-filled glove box. Cyclic voltammetry (CV) (CHI660B
10 min. Then, the self-assembly carbon seal reactor (see Fig. 1) CHENHUA, China) curves were conducted from 2.5 to 4.1 V at
with precursors inside were put into a microwave oven (3.0 GHz, a scanning rate of 0.1 mV s−1 . The charge–discharge cycle was
850 W). After a few minutes of microwave irradiation, the performed on a battery test instrument (CT2001A, KINGNUO,
olivine phase LiFePO4 was obtained. China) at a voltage range of 2.5–4.1 V at 1/2C at room temper-
The four samples obtained in this process are hereafter named ature.
LF1 (sample obtained without citric acid added), LF2 (sample
with 1.25 mmol citric acid added), LF3 (sample with 2.5 mmol 3. Results and discussion
citric acid added) and LF4 (sample with 3.75 mmol citric acid
added). Ground with a pestle for 30 min in an agate mortar,
To make the self-assembly carbon seal reactor (shown in NH4 H2 PO4 and CH3 COOLi first changed to a soft and pulpy
Fig. 1), containers 1 and 2 (with precursors) were put inside the mixture (with strong acetic acid odor emission) and then turned
base. Activated carbon was added to seal up all slits and cover to white solid powder. XRD (shown in Fig. 2) shows that
NH4 H2 PO4 and CH3 COOLi experience the low heating solid-
state coordination reaction. Nearly all diffraction peaks of
CH3 COOLi and NH4 H2 PO4 vanished from the precursor and
some minor peaks of the new product appeared. The precursor,

Fig. 2. XRD patterns of the starting materials and precursor. (a) CH3 COOLi;
Fig. 1. The self-assembly carbon seal reactor: 1–3, containers; 4, lid; 5, (b) NH4 H2 PO4 ; (c) the precursor formed after grinding NH4 H2 PO4 and
microwave absorber. CH3 COOLi for 30 min.
6780 L. Wang et al. / Electrochimica Acta 52 (2007) 6778–6783

Fig. 3. TEM micrographs of the precursors of LiFePO4 : (a) the mixture of NH4 H2 PO4 and CH3 COOLi after being ground for 30 min; (b) the resulting mixture of
(a) added with FeC2 O4 ·2H2 O in the mole ratio of 1:1with Li and ground for another 10 min.

different from raw materials, was molecular mixture of reaction and 450 ◦ C (m = 18.2%) might be ascribed to the formation of
components. The TEM photograph in Fig. 3a shows the typi- LiFePO4 . The endothermic peak appeared at 410 ◦ C was lower
cal morphology of the powder. The sharp contrast between the than that obtained in other methods [12]. It could be predicated
dark edge and the pale center of each sphere indicated its hollow that hollow spheres in the precursors might reduce the synthesis
nature. The sample was composed of a large number of hollow temperature.
spheres with their diameters ranging from 10 to 80 nm. Fig. 3b In the X-ray diffraction patterns (as shown in Fig. 5), main
illustrated the morphology of the precursor of LF1. There were diffraction peaks of the four LiFePO4 samples were almost the
some amorphous materials encircled by hollow spheres with the same as those of pure orthorhombic olivine phase LiFePO4
same diameter range of 10–80 nm. (JCPDS card no. 40-1499). This proved that the self-assembly
Since the reaction process of the four precursors is similar, the carbon seal reactor provided a reliable protective atmosphere.
precursor of LF1 was chosen as the representative sample. The Although prepared with citric acid addition, LF2, LF3 and LF4
TG–DSC results of the precursor of LF1 in N2 (shown in Fig. 4) had similar diffraction patterns to that of LF1. It indicated that
shows that the weight loss of the precursor started at 100 ◦ C and the carbon generated from citric acid carbonization was amor-
continued till 450 ◦ C, where weight loss stopped. This indicated phous as there was no diffraction peaks for crystalline carbon
that the reaction can be achieved before 450 ◦ C. Two endother- (graphite) [25]. According to Fig. 5, LF4 owned low diffraction
mic peaks in Fig. 4 corresponded to the two steps of weight loss. peaks, which implied a decrease in crystallinity of the compos-
The weight loss between 130 and 350 ◦ C (m = 29.2%) was due ites. This showed that the addition of excessive citric acid might
to the evaporation of crystal water in the mixed precursor and suppress the formation of crystalline LiFePO4 .
the decomposition of salts. DSC curve shows that this process Fig. 6 shows TEM images of the four LiFePO4 samples. The
are endothermic (Tpeak = 201 ◦ C). The weight loss between 350 particle size of LF1 ranges from 200 to 800 nm, while the particle

Fig. 4. TG–DSC curves of the precursor phase of LF1 recorded over the tem-
perature range from ambient to 900 ◦ C at a heating rate of 20 ◦ C min−1 under
nitrogen atmosphere. Fig. 5. XRD patterns of LiFePO4 samples: (a) LF1; (b) LF2; (c) LF3; (d) LF4.
 L. Wang et al. / Electrochimica Acta 52 (2007) 6778–6783 6781

Fig. 6. TEM images of the LiFePO4 samples: (a) LF1; (b) LF2; (c) LF3; (d) LF4.

size of LF2, LF3 and LF4 is smaller with the diameter of 80–150, sion length of lithium ions in LiFePO4 and facilitated extraction
40–50 and 10–20 nm, respectively. This showed that the particle and insertion of lithium ions.
size of LiFePO4 was closely related to the amount of citric acid. The initial charge–discharge curves of the four LiFePO4 sam-
The formation of homogeneous nano-sized particles of LiFePO4 ples (shown in Fig. 8) illustrated the galvanostatic discharge
(with citric acid added) might be ascribed to the following three profiles of the samples at 1/2C. The flat voltage plateaus at
factors: (1) the precursors with hollow spheres obtained by low the range of 3.3–3.5V indicated the two-phase nature of the
heating solid-state coordination method might assist in form-
ing the initial nano-particles; (2) microwave assisted synthesis
helped to maintain the nano-crystalline structure even after sin-
tering [22]; (3) in the heating process, the expansion of added
citric acid reduced the surface tension of the precursors and
prevented the particles from growing up. All these factors may
interact with each other and lead to nano-particles, but no con-
clusive evidence has been achieved. Further work in this field is
being carried out.
Cyclic voltammaetry measurements were performed on
LiFePO4 samples at a scanning rate of 0.1 mV s−1 , voltage range
2.5–4.1 V. The CV profiles of LF1, LF2, LF3 and LF4 in the first
cycle (shown in Fig. 7) showed that anodic and cathodic peak
intensities of LF1 and LF2 were much lower than those of LF3
and LF4. Broadened peaks of LF1 and LF2 showed their slower
kinetics, while sharp peaks of LF3 and LF4 highlighted their
excellent kinetics [25,26]. The CV profiles showed that lithium
ions and electrons were quite active in redox reactions, which Fig. 7. The CV results of the LiFePO4 samples: (a) LF1; (b) LF2; (c) LF3; (d)
could be attributed to the smaller particles that reduced the diffu- LF4. Scanning rate: 0.1 mV s−1 . Voltage range: 2.5–4.1 V.
6782 L. Wang et al. / Electrochimica Acta 52 (2007) 6778–6783

particles with excellent crystallinity and carbon doping, which

reduce the diffusion length of the ions in the sample and enhance
the electronic conductivity.

4. Conclusion

LiFePO4 was prepared by microwave heating the precursors,

which were synthesized by low heating solid-state coordination
method. Compared with the complicated conventional furnace
heating procedures [26,28,29], this method presents a more
preferable and faster way of nano-LiFePO4 synthesis. In the
heating process, suitable citric acid addition led to the formation
of homogeneous nano-particles with the diameter of 40–50 nm
and good crystalline phase, which improved the cycling perfor-
mance of LiFePO4 . Charge–discharge testing indicated that the
Fig. 8. The initial charge–discharge curves of the four LiFePO4 samples: LF1,
sample with 2.5 mmol citric acid addition possessed satisfactory
LF2, LF3 and LF4. electrochemical capacity of 123 mAh g−1 at the 50th cycle. The
excellent cycle life and stability as well as the fast and efficient
lithium extraction and insertion reactions between LiFePO4 and method make the process feasible commercially. The whole syn-
FePO4 . Compared with LF1, LF2 and LF4, LF3 presented a thesis process, performed without any solvent, is economic and
more definite discharge plateau with a steep potential decline environmentally benign.
in the end. Though with smallest particles, LF4 showed poor
capacity, which might be due to the inferior crystallinity of the Acknowledgements
sample.
The cycling behaviors of different LiFePO4 samples at C/2 This work was partially supported by the National Nature Sci-
rate are described in Fig. 9. The discharge capacity of LF1 ence Foundations of China (Grant No. 20666005 and 20661003)
continuously decreased with further cycling. In contrast, the and the “ChunHui” plan of Ministry of Education of China
capacity of LF2, LF3 and LF4 slightly increased in the initial (Grant No. Z2004-2-65031).
few cycles. Similar phenomena were also reported previously
[13,27]. The slight difference in the capacity might be attributed References
to the difference in electrical conductivity and particle size.
The capacity of LF1 decreased much faster with 50 mAh g−1 [1] A.K. Padhi, K.S. Nanjundaswamy, J.B. Goodenough, J. Electrochem. Soc.
144 (1997) 1609.
at the first cycle and 30 mAh g−1 at the 50th. On the contrary, [2] A. Yamada, S.C. Chung, K. Hinokuma, J. Electrochem. Soc. 148 (2001)
samples LF2, LF3 and LF4 prepared with citric acid addition A224.
maintained good capacity without obvious capacity fading. The [3] S. Okada, S. Sawa, M. Egashira, J.I. Yamaki, M. Tabushi, J. Power Sources
capacity of LF3, especially, was 112 mAh g−1 at the first cycle, 97–98 (2001) 430.
125 mAh g−1 at the 10th and 123 mAh g−1 at the 50th cycle. [4] D.D. MacNeil, Z. Lu, Z. Chen, J.R. Dahn, J. Power Sources 108 (2002) 8.
[5] A.S. Andersson, J.O. Thomas, J. Power Sources 97–98 (2001) 498.
It is obvious that LF3 prepared with suitable amount of citric [6] K. Zaghib, K. Striebel, A. Guefi, J. Shim, M. Armand, M. Gauthier, Elec-
acid exhibits better electrochemical properties due to its nano- trochim. Acta 50 (2004) 263.
[7] N. Ravet, Y. Chouinard, J.F. Magnan, S. Besner, M. Gauthier, M. Armand,
J. Power Sources 97–98 (2001) 503.
[8] F. Croce, A. D’Epifanio, J. Hassoun, et al., Electrochem. Solid State Lett.
5 (3) (2002) A47.
[9] K. Konstantinov, S. Bewlay, G.X. Wang, et al., Electrochim. Acta 50 (2004)
421.
[10] P.P. Prosini, M. Carewska, S. Scaccia, P. Wisnicwski, M. Pasquali, Elec-
trochim. Acta 48 (2003) 4205.
[11] H.S. Kim, B.W. Cho, W. Cho, J. Power Sources 132 (2004) 235.
[12] S. Franger, F.L. Cras, C. Bourbon, H. Rouault, J. Power Sources 252 (2003)
119–121.
[13] M.M. Doeff, Y. Hu, F. McLarnon, R. Kostecki, Electrochem. Sol. State
Lett. 6 (2003) 207.
[14] S. Yang, P.Y. Zavalij, M.S. Whittingham, Electrochem. Commun. 3 (2001)
505.
[15] R.Y. Wang, D.Z. Jia, L. Zhang, L. Liu, Z.P. Guo, B.Q. Li, J.X. Wang, Adv.
Funct. Mater. 16 (5) (2006) 687.
[16] D.Z. Jia, J.Q. Yu, X. Xia, Chin. Sci. Bull. 43 (1998) 571.
[17] X.R. Ye, D.Z. Jia, J.Q. Yu, X.Q. Xin, Z. Xue, Adv. Mater. 11 (11) (1999)
941.
Fig. 9. Cycling performances of LF1, LF2, LF3 and LF4 samples at 1/2C. [18] Y.D. Huang, J. Li, D.Z. Jia, Chin. J. Inorg. Chem. 7 (2004) 837.
 L. Wang et al. / Electrochimica Acta 52 (2007) 6778–6783 6783

[19] Y.D. Huang, J. Li, D.Z. Jia, J. Colloid Interface Sci. 286 (2005) 263. [24] K.S. Park, J.T. Son, H.T. Chung, S.J. Kim, C.H. Lee, H.G. Kim, Elec-
[20] X.C. Tang, Y.P. Yang, L.Q. Li, D.Z. Jia, et al., Chin. J. Nonferrous Met. 14 trochem. Commun. 5 (2003) 839.
(2004) 871. [25] S.S. Zhang, J.L. Allen, K. Xu, et al., J. Power Sources 147 (2005)
[21] X.C. Tang, X.W. Song, P.Z. Shen, D.Z. Jia, Electrochim. Acta 50 (2005) 234.
5581. [26] H. Liu, C. Li, H.P. Zhang, et al., J. Power Sources 159 (2006) 717.
[22] I. Ganesh, R. Johnson, G.V.N. Rao, Y.R. Mahajan, S.S. Madavendra, B.M. [27] H.C. Shin, W. Cho, H. Jang, Electrochim. Acta 52 (4) (2006) 1472.
Reddy, Ceram. Int. 31 (2005) 67. [28] J. Lee, A.S. Teja, Mater. Lett. 60 (2006) 2105.
[23] M. Higuchi, K. Katayama, Y. Azuma, M. Yukawa, M. Suhara, J. Power [29] M.A.E. Sanchez, G.E.S. Brito, M.C.A. Fantini, et al., Solid State Ionics
Sources 119–121 (2003) 258. 177 (2006) 497.