Optimized carbon-coated LiFePO4 cathode material for lithium-ion batteries

Optimized carbon-coated LiFePO4 cathode material for lithium-ion batteries

Materials Chemistry and Physics 115 (2009) 245–250 Contents lists available at ScienceDirect Materials Chemistry and Physics journal homepage: www.e...

851KB Sizes 1 Downloads 122 Views

Materials Chemistry and Physics 115 (2009) 245–250

Contents lists available at ScienceDirect

Materials Chemistry and Physics journal homepage: www.elsevier.com/locate/matchemphys

Optimized carbon-coated LiFePO4 cathode material for lithium-ion batteries Y.Z. Dong, Y.M. Zhao ∗ , Y.H. Chen, Z.F. He, Q. Kuang School of Physics, South China University of Technology, Wushan Road, Tianhe, Guangzhou, Guangdong 510640, PR China

a r t i c l e

i n f o

Article history: Received 11 January 2008 Received in revised form 12 October 2008 Accepted 28 November 2008 Keywords: Powder diffraction Inorganic compounds Electrochemical properties

a b s t r a c t Lithium iron phosphate (LiFePO4 ) cathode material has been synthesized by a solid-state reaction. The XRD patterns of the samples show that the single-phase LiFePO4 compounds can be obtained in our experimental conditions. According to Popa theory, using the result from Rietveld refinement, the shape and the size of crystallite can be obtained. The result shows that the use of carbon gel in precursors do not change the structure of the crystal, but it can inhibit the particle growth and restrain the anisotropy growth of the grain at a lower temperature. At a higher temperature, carbon-coated LiFePO4 shows an anisotropy growth, i.e. growth rate along (1 0 0) crystal plane is more rapid than that of (1 1 1) crystal plane. In our experimental conditions, a spherical carbon-coated LiFePO4 can be synthesized successfully at 650 ◦ C. The electrochemical testing indicated that the spherical carbon-coated LiFePO4 had the excellent performance. Its initial specific capacities were 156.7 mAh g−1 under the rate of C/10. At the 50th cycle, the reversible specific capacities were found to approach 151.2 mAh g−1 (the ratio of 96.5% of initial capacity). Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved.

1. Introduction In recent years, lithium-ion batteries have been of great potential use for electric vehicles (EVs), hybrid electric vehicles (HEVs), dispersed energy storage systems, and other uses such as space batteries. The cost, safety, environmental friendliness, and long operational life of the cathode materials are of major concern for the application of large-scale lithium-ion batteries. Commercial lithium-ion batteries utilize generally cobalt-based oxides as the positive electrode [1,2], but its high cost and toxicity prohibit its use in a large-scale or biomedical applications. Manganese-based materials are attractive because of both their low cost and low toxicity. However, they suffer from important capacity fading during cycling, especially at high temperatures. As part of an intensive search for alternative materials, lithium transition metal phosphates (LiFePO4 ) have become of great interest as storage cathodes for rechargeable lithium-ion batteries because of its high energy density (around 170 mAh g−1 ), low raw material cost, environmental friendliness and safety. In addition, because of its structural similarity between the charged and discharged states, it shows good cycle stability. However, owing to their negligible electronic conductivity, the capacities are only partially accessible even at low rates, presenting a major drawback to practical implementation of the materials [3]. Fortunately, low electronic conductivity can be successfully overcome by some methods such as modifying the LiFePO4 par-

∗ Corresponding author. Fax: +86 20 87112844. E-mail address: [email protected] (Y.M. Zhao).

ticles with carbon [4–8] and also by minimizing particle size [9,10]. These structures either adopted a sp3 character for carbon or they formed linear polymers with a sp2 character on the carbon atoms in the chain. This category of carbon seriously impacts the electrochemical performance of LiFePO4 . The carbon content and particle size seriously influences the tap density of LiFePO4 , which is an important physical property that must be considered. LiFePO4 can be synthesized by high-temperature solid-state reactions [3], hydrothermal procedures [11], or by sol–gel methods [12]. Although the olivine phase can be easily obtained under the hydrothermal conditions, the results of X-ray diffraction analysis, comparison of the lattice parameters of a = 10.381 Å, b = 6.013 Å, and c = 4.716 Å with those (a = 10.333 Å, b = 6.011 Å, and c = 4.696 Å) for ordered LiFePO4 , suggest that there are about 7% iron atoms in the lithium site [13]. These iron atoms can block diffusion of the lithium ions, and therefore, have the negative influence on the electrochemical performance. Thus, it is very important to ensure the ordering of the lithium and iron atoms in this material. Solid-state reaction synthesis of LiFePO4 is mainly performed using Li2 CO3 as Li-precursor [14–16] under the carbon-thermal reduction (CTR) process or using the mixed gas (95% Ar + 5% H2 ) flow as reducing agent, and the impurities are often found in the final products unless the sintering temperature are increased to 800 ◦ C. The high sintering temperature can cause the abrupt particle growth, which has the negative influence on the electrochemical performances, and thus, the synthesis route using other Li-precursor instead of Li2 CO3 should be developed. Recently, Ying and co-workers have studied particle morphology relationships in correlation to volume energy densities. The powders, which are composed of spherical particles, have higher

0254-0584/$ – see front matter. Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.matchemphys.2008.11.063

246

Y.Z. Dong et al. / Materials Chemistry and Physics 115 (2009) 245–250

densities than irregular particles [17–19]. The spherical particles have less contacting interface and better fluidity characteristics than the irregular particles. When the spherical particles vibrate, the small particles easily move and occupy the available vacancies. Therefore, spherical particles are easy to close pack [18]. In this paper, we report our results on the preparation and properties of spherical carbon-coated LiFePO4 compounds synthesized using LiF as lithium source. The solid-state reaction mechanism when using LiF as Li-precursor and a mixed gas (95% Ar + 5% H2 ) flow and carbon gel as reducing agents was analyzed based on the X-ray diffraction (XRD) analysis, where the carbon gel was added directly as carbon source. The preparation conditions on the properties of LiFePO4 as the cathodes materials were investigated and, especially, the influence of temperature on the size and the morphology of the grains was studied. 2. Experimental LiFePO4 samples were prepared by mixing stoichiometric amounts of (NH4 )2 HPO4 , FeC2 O4 ·2H2 O, and LiF as starting materials; the precursors were dispersed into acetone and then ball milled for 7 h in a planetary mill. The rotating speed was 250 rpm and the ball to charge weight ratio was 20:1. After evaporating acetone, the mixture was first decomposed at 350 ◦ C in a N2 atmosphere for 10 h to allow H2 O and NH3 to evolve. The reagents were then re-ground prior to heating in a sealed tube furnace. The samples were heated at a rate of 2 ◦ C min−1 to the temperatures ranging from 650 ◦ C to 800 ◦ C, respectively, under a stream of a mixture of 95% Ar + 5% H2 . The materials were held for 10 h at the upper temperature and slowly cooled down to room temperature prior to removal from the furnace. The gray powders (bare LiFePO4 ) were obtained. For the carbon-coated LiFePO4 , carbon gel (with the total amount of 10 wt% for the final product LiFePO4 ) was added to the mixture of the starting materials. After the same mechanochemical treatment, the carbon-coated LiFePO4 powders were synthesized with the same heat-treatment conditions as was adopted for bare LiFePO4 . The initial characterization of the material was carried out using powder X-ray diffraction, using a MAX 18A-HF diffractometer with rotating anode, which had an 18 kW X-ray generator and Cu K␣ radiation. A graphite monochromator was used for diffracted beams. All the diffraction intensity data for Rietveld refinement analysis were collected by a step scan mode with a scanning step of 0.02◦ and a sampling time of 2 s. Rietveld refinement was performed using the FullProf program to obtain the crystal structure parameters. Scan electron microscopy (SEM, LEO 1530VP, Germany) was carried out in order to examine the effect of sintering temperature on powder morphology. The carbon content was determined by atomic absorption spectrometry as 9.12 wt% (referred to LiFePO4 ). For electrochemical measurements, the LiFePO4 electrodes were prepared by mixing with a conductive carbon (acetylene black) and a binder (PVDF-HFP copolymer). The electrode constituents were mixed into slurry with acetone to achieve homogeneity. The resulting slurry was coated onto aluminum foil current collector using a doctor blade. After the acetone had evaporated, the resulting electrode composition was 85:10:5 of active material, carbon, and binder, respectively. The electrode was then dried at 50 ◦ C for 24 h and pressed with 5 MPa pressure, respectively. The electrodes fabricated were dried again at 80 ◦ C for 12 h in a vacuum and cut into 1 cm × 1 cm × 0.006 cm in size where about 5 mg of active materials was held on it. Two-electrode electrochemical cells were assembled in a Mikrouna glove box filled with high-purity argon, where the lithium metal foil was used as anode, Celgard® 2320 was used as separator, and 1 M LiPF6 in EC:DMC (1:1 vol.%) was used as an electrolyte. The electrochemical capacity measurements were performed in the voltage range between 2.5 V and 4.2 V, and the electrochemical capacities of samples were evaluated on the active materials.

3. Results and discussion Fig. 1(a) and (b) shows the XRD patterns of the bare LiFePO4 and the carbon-coated LiFePO4 synthesized by solid-state reaction at 650 ◦ C. The XRD pattern of the bare LiFePO4 (Fig. 1(a)) was indexed as an orthorhombic system (space group Pnma). Noticeable structural changes were not observed in the XRD pattern of the carbon-coated LiFePO4 (Fig. 1(b)), the cell parameters obtained from Rietveld refinement are a = 10.316 Å, b = 6.002 Å, c = 4.694 Å, volume = 290.637 Å3 for bare LiFePO4 basically consistent with the values a = 10.312 Å, b = 6.005 Å, c = 4.691 Å, volume = 290.483 Å3 for carbon-coated LiFePO4 . There is no evidence for crystalline carbon, nor is amorphous peak present. This is undoubtedly due to the small amount of carbon and the thinness of the layer on the lithium

Fig. 1. X-ray diffraction patterns of LiFePO4 sintered at 650 ◦ C: (a) bare LiFePO4 and (b) carbon-coated LiFePO4 .

iron phosphate. The amount of carbon was determined by atomic absorption spectrometry to be around 9.12%. That shows that the olivine structure was well maintained after using carbon gel and impurity phases such as Fe2 O3 or Fe2 P, which were reported by others [20,21], were not observed in carbon-coated LiFePO4 . But the diffraction intensity of the bare LiFePO4 is bigger than the carboncoated LiFePO4 . It means that the former has a big particle size. This can be seen from Fig. 2(a) and (b). Fig. 2(a) and (b) shows the shape and size of the grain for the bare LiFePO4 and carbon-coated LiFePO4 , respectively, sintered at 650 ◦ C resulted from the Rietveld refinement according to Popa theory [22]. Several parameters, such as cell parameter, texture, crystallite size, microstrain important to XRD patterns were considered in the refinement. According to the Popa theory, the shape and the size of the grain resulted from the refinement are the comprehensive effects of the truly X-ray diffraction of all the grains. From this figure, we can see that the size of the carbon-coated LiFePO4 grain was much smaller than that for the bare LiFePO4 . In addition, the grain shape of the bare LiFePO4 is an approximate octahedron but the grain shape of the carbon-coated LiFePO4 is spherical. It is basically consistent with the result of SEM (Fig. 3(a) and (b)). We can see in Fig. 3(b) that the grain size for carbon-coated LiFePO4 is much smaller than that for the bare LiFePO4 shown in Fig. 3(a). This indicated that the addition of carbon inhibited the particle growth during sintering process. A similar result was reported by Huang et al. [7] suggesting that the addition of fine carbon to the precursors reduced the particle size of the LiFePO4 . Prosini et al. [23] and Chen and Dahn [24] also showed that the particle size decreased as the amount of carbon increased. In addition, the most important is that addition of carbon to the precursors restrained the anisotropy growth of sample at a lower temperature and induced the formation of spherical grain. It is well known that LiFePO4 and FePO4 are poor electronic conductors because they each contain Fe cations with just one oxidation state (2+ or 3+ , respectively). It means that use of small sized and regular shaped LiFePO4 particle will be an efficient method for the improvement of the electronic conductors [7]. Fig. 4(a) and (b) shows the discharge profiles of the bare LiFePO4 and the carbon-coated LiFePO4 at different discharge rates, ranging from 0.1C to 1C. The cell was cycled between 2.5 V and 4.2 V. The carbon-coated LiFePO4 has an excellent flat voltage plateau around 3.4 V. The voltage was kept almost constant when discharge was changed. It can be seen that carbon-coated LiFePO4 exhibits excellent rate capacities with initial discharge capacities of 152 mAh g−1

Y.Z. Dong et al. / Materials Chemistry and Physics 115 (2009) 245–250

247

Fig. 2. The shapes and sizes of the grain for sample: (a) bare LiFePO4 sintered at 650 ◦ C and (b–e) carbon-coated LiFePO4 sintered at 650 ◦ C, 700 ◦ C, 750 ◦ C and 800 ◦ C, respectively.

Fig. 3. The scanning electron microscopy (SEM) images of the (a) bare LiFePO4 and (b) carbon-coated LiFePO4 sintered at 650 ◦ C.

248

Y.Z. Dong et al. / Materials Chemistry and Physics 115 (2009) 245–250

Fig. 5. X-ray diffraction patterns of carbon-coated LiFePO4 sintered at different temperatures: (a) 650 ◦ C, (b) 700 ◦ C, (c) 750 ◦ C and (d) 800 ◦ C.

Fig. 4. The discharge profiles of the LiFePO4 sintered at 650 ◦ C: (a) bare LiFePO4 and (b) carbon-coated LiFePO4 .

(0.1C), 140 mAh g−1 (0.2C), 132 mAh g−1 (0.5C) and 123 mAh g−1 (1C). But flat voltage for bare LiFePO4 has a big change when different discharge was adopted and discharge faded more quickly. This effect can be attributed to the improvement of lithium-ion diffusion because of the use of carbon gel, the decrease of the size and regular shape for particles (see Fig. 2). In order to study the influence of shape on electrochemical property of carbon-coated LiFePO4 , different temperatures were adopted in synthesize process. Fig. 5(a)–(d) shows the X-ray diffraction patterns of carbon-coated LiFePO4 synthesized at 650 ◦ C, 700 ◦ C, 750 ◦ C and 800 ◦ C, respectively. For comparison, the bottom of the figure shows the data of peak positions and their relative intensities of a crystalline LiFePO4 sample from JCPDS card (No. 81-1173). As shown in Fig. 5, our results suggest that single-phase LiFePO4 compounds can be obtained in the sintering temperature ranging from 650 ◦ C to 800 ◦ C under our experimental conditions. The X-ray diffraction patterns of the LiFePO4 compounds were successfully indexed with orthorhombic lattice using the program Dicovl, and the space group of the LiFePO4 compounds was derived to be Pnma based on the reflection conditions. Typical refinement result of the carbon-coated LiFePO4 compound sintered at the temperature of 650 ◦ C is shown in Fig. 6. The weighted factors Rwp obtained from Rietveld refinement for the samples sintered at the temperatures from 650 ◦ C to 800 ◦ C were ranged from 11.90% to 9.81%. The corresponding grain shapes of these samples are presented in Fig. 2(b)–(e). From Fig. 2(b), we can observe that the shape of the grain in the sample by using carbon gel is spherical at 650 ◦ C. But the shapes have obvious change with the

increase in temperature. This is due to the fact that the rate along (1 0 0) crystal plane is more rapid than that of (1 1 1) crystal plane at the higher temperature. So the synthesizing temperature plays a key role in controlling the shape of the grain in our experimental conditions. In Fig. 7(a)–(d), we present the differential capacity plots for the carbon-coated LiFePO4 , synthesized at different temperatures, covering the extended voltage range of 2.5–4.2 V versus Li. A 10mV step size was used during this preliminary material evaluation experiment. All samples have only one pair of peaks indicative of a typical two-phase reaction between LiFePO4 and FePO4 . The anodic peak, which is indicative of lithium-ion extraction from LiFePO4 , occurs at ∼3.5 V versus Li/Li+ . The cathodic peak corresponding to the anodic peak occurs at ∼3.48 V and signifies a Fe2+ /Fe3+ redox potential. The small potential difference of about 0.02 V between the anodic and cathodic peaks for the LiFePO4 sintered at 650 ◦ C demonstrates good reversibility on charge–discharge cycling. But an increased potential difference between the anodic and cathodic peaks was found with the increase of sintering temperature. This may be influenced by the particle size and shape distribution. This suggests that the reversible performance of charge/discharge becomes worse. In addition, the other electrode has broad peaks in the differential capacity curves. In contrast, the spherical LiFePO4 electrode demonstrates sharp redox peaks. This

Fig. 6. The Rietveld refinement result for carbon-coated LiFePO4 sintered at 650 ◦ C.

Y.Z. Dong et al. / Materials Chemistry and Physics 115 (2009) 245–250

249

Fig. 7. Differential capacity plots of carbon-coated LiFePO4 sintered at different temperatures: (a) 650 ◦ C, (b) 700 ◦ C, (c) 750 ◦ C and (d) 800 ◦ C.

can be attributed to an improvement in the kinetics of the lithium intercalation/deintercalaction at the electrode/electrolyte interface and/or the rate of lithium diffusion in the film with the increasing specific surface area. The cycle performances of the carbon-coated LiFePO4 sintered at the temperatures ranging from 650 ◦ C to 850 ◦ C are shown in Fig. 8. As shown in Fig. 8, the carbon-coated LiFePO4 samples exhibit better capacity retention after 50 cycles, where a relative sustained cycling behavior is observed for the 4.2 V voltage cut-offs with the initial specific capacities of 156.7 mAh g−1 for the samples sintered at a temperature of 650 ◦ C. It can be seen from Fig. 8 that LiFePO4

compounds prepared at a relative lower temperature of 650 ◦ C show the highest initial specific capacities than those for the samples prepared at other temperatures used in our experimental process. At the 50th cycle, the discharge capacity between 2.5 V and 4.2 V is 92.6 mAh g−1 (74.2% of initial capacity), 151.2 mAh g−1 (96.5% of initial capacity), 138 mAh g−1 (92.3% of initial capacity), 122.2 mAh g−1 (86.1% of initial capacity), 110.5 mAh g−1 (82.6% of initial capacity) and 48.5 mAh g−1 (71.2% of initial capacity) for the samples sintered at the temperatures of 600 ◦ C, 650 ◦ C, 700 ◦ C, 750 ◦ C, 800 ◦ C, and 850 ◦ C, respectively. The high capacity and excellent cycle performance delivered by the carbon-coated LiFePO4 electrode should be attributed to its high electronic conductivity. Since individual LiFePO4 particles are connected with a carbon network, the active LiFePO4 material can be fully utilized for lithium extraction and insertion reactions. 4. Conclusion

Fig. 8. The cycle performance of the carbon-coated LiFePO4 sintered at different temperatures: (a) 650 ◦ C, (b) 700 ◦ C, (c) 750 ◦ C and (d) 800 ◦ C.

LiFePO4 compounds have been synthesized successfully by solid-state reaction using LiF as lithium source in our experiment. The X-ray diffraction pattern and Rietveld refinement results show that in our product there is no impurity phase of Fe2 O3 or Fe2 P. According to Popa theory, using the resulted from Rietveld refinement, the shape and the size of crystallite can be obtained. The result shows that the use of carbon gel in precursors do not change the structure of the crystal, the unit cell parameters (a = 10.316 Å, b = 6.002 Å, c = 4.694 Å and volume = 290.637 Å3 ) for bare LiFePO4 are basically consistent with the values (a = 10.312 Å, b = 6.005 Å, c = 4.691 Å and volume = 290.483 Å3 ) for carbon-coated LiFePO4 , but it can inhibit the particle growth and restrain the anisotropy growth

250

Y.Z. Dong et al. / Materials Chemistry and Physics 115 (2009) 245–250

of the grain at a lower temperature (650 ◦ C in our experimental conditions). At a higher temperature, carbon-coated LiFePO4 shows an anisotropy growth, i.e. growth rate along (1 0 0) crystal plane is more rapid than that of (1 1 1) crystal plane. In our experimental conditions, a spherical carbon-coated LiFePO4 can be synthesized successfully at 650 ◦ C. The electrochemical testing indicated that the spherical carbon-coated LiFePO4 had the excellent electrochemical performance. In differential capacity plot, the anodic peak occurs at ∼3.5 V versus Li/Li+ and the cathodic peak occurs at ∼3.48 V. The small potential difference of around 0.02 V between the anodic and cathodic peaks for the LiFePO4 sintered at 650 ◦ C demonstrates good reversibility on charge/discharge cycling. Its initial specific capacities are 156.7 mAh g−1 under the rate of C/10. At the 50th cycle, the reversible specific capacities are found to approach 151.2 mAh g−1 (the ratio of 96.5% of initial capacity). Acknowledgements This work was funded by NSFC Grant (No. 50772039) supported through NSFC Committee of China and Science and Technology Foundation and supported through the Science and Technology Bureau of Guangdong Government (No. 07118058).

References [1] M.M. Thackeray, J.O. Thomas, M.S. Whittingham, MRS Bull. 25 (2000) 39. [2] T. Kerr, J. Gaubicher, L.F. Nazar, Electrochem. Solid State Lett. 3 (2000) 460. [3] A.K. Padhi, K.S. Najundaswamy, J.B. Goodenough, J. Electrochem. Soc. 144 (1997) 1188. [4] G.X. Wang, L. Yang, Y. Chen, Electrochim. Acta 50 (2005) 4649. [5] R. Dominko, M. Bele, M. Gaberscek, J. Electrochem. Soc. 152 (2005) A607. [6] N. Ravet, J.B. Goodenough, S. Besner, Presented at the 196th Meeting of the Electrochemical Society, Honolulu, HI, October, 1999. [7] H. Huang, S.C. Yin, L.F. Nazar, Electrochem. Solid State Lett. 4 (2001) A170. [8] G.X. Wang, L. Yang, S.L. Bewlay, J. Power Sources 146 (2005) 521. [9] S. Franger, F.L. Cras, C. Bourbon, J. Power Sources 119 (2003) 252. [10] P.P. Prosini, M. Carwska, S. Scaccia, J. Electrochem. Soc. 149 (2002) A886. [11] S. Yang, P.Y. Zavalij, M.S. Whittingham, Electrochem. Commun. 3 (2001) 505. ˜ [12] Y. Hu, M.M. Doeff, R. Kostecki, R. Finones, J. Electrochem. Soc. 151 (2004) A1279. [13] S. Yang, Y. Song, K. Ngala, P.Y. Zavalij, M.S. Whittingham, J. Power Sources 119 (2003) 239. [14] A. Yamada, S.C. Chung, K. Hinokuma, J. Electrochem. Soc. 148 (2001) A224. [15] G. Li, H. Azuma, M. Tohda, J. Electrochem. Soc. 149 (2002) A743. [16] A.D. Spong, G. Vitins, J.R. Owen, J. Electrochem. Soc. 152 (2005) A2376. [17] J.R. Ying, C.Y. Jiang, C.R. Wan, J. Power Sources 129 (2004) 264. [18] J.R. Ying, C.R. Wan, C.Y. Jiang, J. Power Sources 99 (2001) 78. [19] X. He, J. Li, Y. Cai, J.R. Ying, J. Power Sources 150 (2005) 216. [20] S.Y. Chung, J.T. Bloking, Y.M. Chiang, Nat. Mater. 1 (2002) 123. [21] S.J. Kwon, C.W. Kim, W.T. Jeong, K.S. Lee, J. Power Sources 137 (2004) 93. [22] N.C. Popa, J. Appl. Crystallogr. 31 (1998) 176. [23] P.P. Prosini, D. Zane, M. Pasquali, Electrochim. Acta 46 (2001) 3517. [24] Z. Chen, J.R. Dahn, J. Electrochem. Soc. 149 (2002) A1184.