Microwave solid-state synthesis of spinel Li4Ti5O12 nanocrystallites as anode material for lithium-ion batteries

Microwave solid-state synthesis of spinel Li4Ti5O12 nanocrystallites as anode material for lithium-ion batteries

Available online at www.sciencedirect.com Solid State Ionics 178 (2007) 1590 – 1594 www.elsevier.com/locate/ssi Microwave solid-state synthesis of s...

478KB Sizes 0 Downloads 25 Views

Available online at www.sciencedirect.com

Solid State Ionics 178 (2007) 1590 – 1594 www.elsevier.com/locate/ssi

Microwave solid-state synthesis of spinel Li4Ti5O12 nanocrystallites as anode material for lithium-ion batteries Juan Li a,b,⁎, Yong-Li Jin a , Xiao-Gang Zhang c,⁎, Hui Yang a a

c

Institute of Applied Chemistry, Xinjiang University, Urumqi 830046, People's Republic of China b School of Science, Xi'an Jiaotong University, Xi'an 710049, People's Republic of China College of Material Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, People's Republic of China Received 22 April 2007; received in revised form 19 October 2007; accepted 19 October 2007

Abstract Spinel Li4Ti5O12 nanocrystallites as anode material for lithium-ion batteries were synthesized by microwave method with Li2CO3 and TiO2 as reactants. The morphology and microstructure of the samples were characterized by X-ray diffraction (XRD) analysis, transmission electron microscope (TEM), scanning electron microscopy (SEM) and Fourier transformed infrared (FT-IR). The electrochemical properties of the samples were studied by cyclic voltammogram (CV) and galvanostatic charge/discharge tests. It was shown that the purity and the crystallinity of the composite oxides were dependent on the irradiation power and time and the crystallites were regular sphere-like nanoparticles about 40–50 nm in size. The spinel Li4Ti5O12 anode has a stable voltage plateau, high discharge capacity and an excellent cyclibility. © 2007 Elsevier B.V. All rights reserved. Keywords: Spinel Li4Ti5O12; Microwave method; Nanocrystallite; Anode material; Lithium-ion batteries

1. Introduction Lithium-ion batteries have been rapidly popularized as a power source for various portable electronic devices, such as video cameras, laptop computers, cell phones and so on [1], because of their advantages of high energy density and light weight. LiCoO2/graphite is the most commonly used electrochemical couple in commercial lithium-ion batteries. But a disadvantage of it is the safety issue, during charge process, at full charge, the anode (LiC6) is highly reactive [2]. In order to resolve the safety limitation, the spinel Li4Ti5O12 has been studied as an alternative anode. It has a stable voltage plateau of approximately 1.5 V versus Li/Li+[3]. Though it is higher than that of carbon material, it can be coupled with high-voltage cathode materials such as LiMn2O4 [4], LiCo0.4Fe0.4Mn3.2O8 [5] or ⁎ Corresponding authors. Li is to be contacted at the Institute of Applied Chemistry, Xinjiang University, Urumqi 83004, People's Republic of China, Tel.: +86 991 8582887; Zhang, College of Material Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, People's Republic of China, Tel.: +86 991 8582887. E-mail address: [email protected] (J. Li). 0167-2738/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.ssi.2007.10.012

Li2CoMn3O8 [6] for an approximately 2.5 V lithium-ion battery with high safety and reversibility. The spinel Li4Ti5O12 is an attractive anode material for lithium-ion batteries. One of the most important properties of it is that its lattice parameters are almost unchanged when lithium ion inserts and extracts. Per formula unit Li4Ti5O12 can insert three lithium ions, so its theoretical capacity is 175 mAh/g [7]. Li+ ions occupy all the 8a sites and 1/6 of the 16d sites while the Ti4+ ions occupy the other 16d sites, and all the 32e sites are occupied with O atoms. The spinel structure of Li4Ti5O12 provides a three-dimensional interstitial 8a tetrahedral and 16c octahedral space for the transport of the Li+ ions. The framework of Li4Ti5O12 is very robust. Therefore, insertion and extraction of lithium are extremely reversible [8]. The spinel Li4Ti5O12 was usually synthesized by a solidstate reaction of stoichiometric amounts of TiO2 and Li2CO3 or LiOH·H2O, with heating at 800–1000 °C for 12–24 h [9–11]. As a novel method, the microwave method has advantages of considerable lower synthesis temperature and shorter reaction time than that of the traditional method. Due to the high penetration depths of the microwave, it can be uniformly and rapidly

J. Li et al. / Solid State Ionics 178 (2007) 1590–1594

absorbed by reactants or heating medium, and thus the high temperature needed in a solid-state reaction can be obtained within several minutes [12]. Recently, microwave irradiation synthesis method has been developed to synthesize a number of inorganic materials (Fe3O4, SnO2, CeO2 and ZrO2) [12–16] and metals (Pt, Ru, Ag) [17–19] and so forth. Microwave preparation has also been applied to prepare electrode materials for lithium batteries. Nakayama et al. [20] prepared LiMn2O4, Yang et al. [12] synthesized LiV3O8 by the microwave solid-state method and Wang et al. [21] prepared SnO2-graphite by microwave-assisted synthesis. All the synthesized products show a good electrochemical performance. In this paper, the spinel Li4Ti5O12 was synthesized by a microwave solid-state method in a modified domestic microwave oven. X-ray diffraction (XRD) analysis, transmission electron microscope (TEM), scanning electron microscope (SEM), Fourier transformed infrared (FT-IR), cyclic voltammogram (CV) and galvanostatic charge/discharge tests were performed in order to characterize and investigate the properties of the anode materials. 2. Experimental 2.1. Synthesis of spinel Li4Ti5O12 A modified domestic microwave oven (Whirlpool, maximum 850 W) was applied for the microwave reaction system. The starting materials were Li2CO3 and TiO2 (anatase). These materials were weighed in stoichiometric ratio and thoroughly mixed using an agate mortar. 1 g of the mixed powder was placed in a 30 ml porcelain crucible with a lid, and placed in the center of secondary 50 ml porcelain crucible. Because the reactants of Li2CO3 and TiO2 could only absorb a small amount of microwaves at a relatively low temperature, the double crucible system was devised, which, using charcoal as a heating medium between two porcelain crucibles, allowed to achieve the high temperature needed in the solid-state reaction. Then the double crucible system was placed inside a cavity in Kaowool base. Finally, the microwave reaction system was placed in the centre

Fig. 1. X-ray diffraction patterns of the samples. A: 500 W 10 min B: 500 W 15 min C: 700 W 10 min D : 700 W 15 min (TiO2⁎ Li2TiOo3 Li2O▀).

1591

Fig. 2. FT-IR spectra of sample B (solid line) and the sample D (dot line).

of a rotating plate of a modified domestic microwave oven. Sample A and B were irradiated at 500 W for 10 and 15 min. with excess Li2CO3 8%, and sample C and D were irradiated at 700 W for 10 and 15 min. with a stoichiometric ratio of the staring material. 2.2. Characterization of the Li4Ti5O12 nanocrystallite The crystallographic textures of the samples were examined by X-ray diffraction (XRD) analysis on a Japan Rigaku D/Max 2400 with Cu Kα radiation. Fourier transformed infrared (FTIR) spectra were obtained on a Bruker EQUINOX 55. The morphology and microstructure of the sample D were observed using a transmission electron microscopy (TEM, Model Hitachi H-600, 200 kV) and scanning electron microscopy (SEM, Germany, Leo1430VP). 2.3. Electrochemical testing All electrochemical measurements were performed in simulated cells with metallic lithium used as the counter and reference electrode. The working electrode consisted of 80% active materials, 10% acetylene carbon, 10% poly (vinylidene fluoride) (PVDF) binder and a little 1-methyl-2-pyrrolidone (NMP) solvent to form uniform slurry. The slurry was homogeneously coated onto a copper foil and dried at 80 °C for 1 h and then at 120 °C for 4 h in vacuum oven. The electrode foils had a surface area of 1 cm2 and contained 2–4 mg of active material. A solution of 1 M LiPF6 dissolved in a mixture of ethylene carbonate (EC) and diethyl carbonate (DEC) (V:V = 1:1) was used as electrolyte and Celgard 2400 as separator. The cells were assembled in an argon-filled glove box, in which water and oxygen concentrations were kept under 2 ppm and 6 ppm, respectively. Cyclic voltammogram (CV) test was performed between 1 V and 2.5 V at a scan rate of 0.1 mv/s on a CHI600b electrochemical workstation system. Galvanostatic charge/discharge test was accomplished on Arbin BT 2042 battery test system, with a constant current density of 0.1 or 0.4 mA/cm2 and the voltage range from 1 V to 2.3 V.

1592

J. Li et al. / Solid State Ionics 178 (2007) 1590–1594

3. Results and discussions Fig. 1 shows the XRD patterns of the samples synthesized by the microwave method at various irradiated power (500 W, 700 W) and time (10 min, 15 min), respectively. The presence of anatase TiO2 (2θ = 25.3°) in A, B and C patterns and of Li2TiO3 (2θ = 19.9°) and Li2O (2θ = 33.5°) in B pattern can be recognized. The cause of the presence of these impurity phases in the final products could not easily be ascribed to one factor, rather a series of factors combines to give rise to these byproducts, including the molar ratio of staring materials, irradiating power and reaction time. We believed that the presence of a Li2O impurity phase was due to an excess of Li2CO3. The presence of Li2TiO3 and unreacted anatase TiO2 was caused by lower irradiation power and inadequate irradiation time. From powder X-ray diffraction of sample D, it can be seen that all the diffraction peaks corresponded to the spinel Li4Ti5O12 (PDF 49-2027) and can be indexed in a cubic phase (Fd3m space group). So the single phase spinel Li4Ti5O12 was obtained by microwave method and the optimum synthesis conditions was a stoichiometric ratio of the staring material Li: Ti = 4:5 and an irradiation time of 15 min at 700 W. The phase purity of the sample B and sample D were also examined by Fourier transformed infrared (FT-IR ) spectroscopy. The FT-IR spectra of sample B(solid line) and sample D(dot line) are shown in Fig. 2. There are two absorption bands whose center located at 488 cm− 1 and 649 cm− 1 in sample B and 462 cm− 1 and 650 cm− 1 in sample D between 400 cm− 1 and 1000 cm− 1 can clearly be observed. The two absorption bands corresponded to the symmetric and asymmetric stretching vibrations of the octahedral groups [MO6] lattice [22]. It confirmed the presence of the spinel Li4Ti5O12 in the sample B and sample D. But compared with sample D, the structural difference was mainly manifested by some other absorption peaks located at 1430 cm− 1 and 1576 cm− 1 and an absorption band centered in 3446 cm− 1 in the range of 1000 cm− 1 to 4000 cm− 1 of the impurity phases in sample B. Moreover, the lack of absorption peaks of the impurity phase in sample D indicated that the spinel Li4Ti5O12 was pure. The FT-IR results were in agreement with

XRD results. That is, when the samples were prepared by microwave method, the temperature obtained from microwave energy was not enough to form a pure spinel Li4Ti5O12 at lower irradiation power (500 W) and higher irradiation power (700 W) could form single phase spinel Li4Ti5O12. According to the results of XRD and FT-IR analysis of the samples, sample D was chosen to observe its microstucture by TEM and SEM. TEM and SEM images of sample D are shown in Figs. 3 and 4, respectively. They demonstrated identical results. It was apparent that the crystallites of the spinel Li4Ti5O12 synthesized by microwave irradiation at 700 W for 15 min were regular sphere-like nanoparticles. They had homogeneous distribution without serious agglomeration and the average size of the nanoparticles was about 40–50 nm. Fig. 5 shows the typical cyclic voltammetric curve of an electrode made from sample D. A pair of reversible redox peaks can be clearly seen, corresponding to a cathodic and anodic process. The cathodic peak located at around 1.5 V corresponds to the voltage plateau of the first cycle discharge process in terms of which Li intercalated into the spinel Li4Ti5O12. The anodic peak located at 1.65 V corresponds to the voltage plateau of the first cycle charge process in which Li deintercalated from the

Fig. 3. TEM image of sample D.

Fig. 5. Cyclic voltammetric curve of sample D.

Fig. 4. SEM image of sample D.

J. Li et al. / Solid State Ionics 178 (2007) 1590–1594

Fig. 6. Voltage curves of the cycle of sample B (current density was 0.1 mA/cm2, figures in curves (1,6,10,16,20) denote cycle number).

1593

plateau at about 1.50 Vand charge plateau at about 1.60 V should correspond to the discharge and charge plateaus of spinel Li4Ti5O12, the plateaus at about 1.75 V and 1.95 V should correspond to the discharge and charge plateaus of anatase TiO2. Fig. 7 shows the discharge and charge curves at different cycles of sample D, which were also cycled between 1 V and 2.3 V at a constant current density was 0.1 mA/cm2. It can be seen from Fig. 7, at the initial discharge, the voltage of cell dropped quickly from open-circuit voltage (about 2.7 V) to 2.1 V, and then decrease as the discharge proceeds until it approaches to 1.50 V. The discharge and charge plateau is around 1.50 V and 1.60 V respectively, which were characteristic discharge and charge plateau of the spinel Li4Ti5O12. It also agrees with the data from cathodic and anodic specific peak potential in the cyclic voltammogram. The electrochemical reaction can be described as: xLi4 Ti5 012 þ 3xLiþ 3xe ⇌xLi7 Ti5 012;

spinel Li4 + xTi5O12. The difference between the cathodic and anodic peak potentials is about 150 mV, which is smaller than the previous results obtained by some authors [22, 23], but larger than the results obtained by other authors [24–26]. The difference is mainly attributed to the lithium-ion diffusivity rate varied in various spinels Li4Ti5O12 with different morphology and microstucture caused by different preparation methods. Moreover, the fact that there is no other redox peaks in the cyclic voltammogram indicates that the spinel Li4Ti5O12 is pure. Simulated test cells were fabricated from sample B, D with metallic lithium to carry out galvanostatic charge/discharge tests. Fig. 6 shows the discharge and charge curves at different cycles of sample B, which were cycled between 1 V and 2.3 V at a constant current density 0.1 mA/cm2. As can be seen in Fig. 6, both discharge and charge curves have two voltage plateaus: the discharge plateaus at about 1.75 V and 1.50 V and the charge plateaus at about 1.60 V and 1.95 V. Because this Li–Ti–O ternary phase prepared by microwave irradiation method at 500 W for 15 min with 8% excess Li2CO3 was a mixture of the spinel Li4Ti5O12, anatase TiO2, Li2TiO3 and Li2O, the discharge

The first discharge capacity was approximately 162 mAh/g. However, the discharge flat voltage increased from 1.50 V to 1.55 V after the second cycle and some irreversible capacity loss on the fist cycle has also been observed. It is primarily attributed to the fact that the surface passivation film grows and blocks the charge transfer reaction between anode and polymer electrolyte during the repeated cycling. Fig. 8 shows the discharge capacity versus cycle number of sample D, which were cycled between 1 Vand 2.3 Vat 0.1 mA/cm2 (a) and 0.4 mA/cm2 (b), respectively. The sample demonstrated a remarkably good electrochemical performance. From the chart, it can be seen that the first cycle discharge capacity achieves 162 mAh/g, the second cycle discharge capacity is 144 mAh/g (88% of initial discharge capacity) when the current density was 0.1 mA/cm2. After 20 cycles, compared to the second cycle, the discharge capacity kept at 94%. However, the discharge capacity notable reduced when the current was increased to 0.4 mA/cm2. But its cycle performance was also good. The first cycle discharge capacity is about 135 mAh/g. After 20 cycles, the discharge capacity kept at 90% (122 mAh/g). Li4Ti5O12 has a stable voltage

Fig. 7. Voltage curves of the cycle of sample D (current density was 0.1 mA/cm2, figures in curves (1,6,10,16,20) denote cycle number).

Fig. 8. The curve of discharge capacity versus cycle number of sample D: (a) 0.1 mA/cm2, (b) 0.4 mA/cm2.

1594

J. Li et al. / Solid State Ionics 178 (2007) 1590–1594

plateau, high discharge capacity and excellent cyclibility, owing to the spherical nanocrystalline morphology that shortens Li+ diffusion pathway lengths. 4. Conclusions In this work, the pure spinel Li4Ti5O12 was successfully synthesized by the microwave irradiation method. The optimum synthesis condition was a stoichiometric ratio of the staring material Li:Ti = 4:5 and an irradiation time of 15 min at 700 W. The product were regular near spherical nanoparticles. Electrochemical testing manifests that the sample Li4Ti5O12 has a stable voltage plateau, high discharge capacity and excellent cyclibility. The initial discharge capacity achieved 162 mAh/g at 0.1 mA/cm2 and 144 mAh/g at 0.4 mA/cm2. So, the microwave synthesis is fast, clean, energy efficient method. It most remarkably occurs at temperatures much lower than in conventional processes and the nanostructured Li4Ti5O12 obtained by this method are promising anode materials for lithium-ion batteries. Acknowledgements This project was supported by the National Science Foundation of China (Nos. 20663006 and 20403014) and Scientific Research Program of the Higher Education Institution of Xinjiang (XJEDU2006S206). We also thank Initial Fund of Xinjiang University (No.070196) for partial support of this work. References [1] C.M. Shen, X.G. Zhang, Y.K. Zhou, Li, Mater. Chem. Phys. 78 (2002) 437. [2] I. Belharouak, K. Amine, Electrochem. Commun. 5 (2003) 435. [3] T. Ohzuku, A. Ueda, N. Yamamoto, J. Electrochem. Soc. 142 (1995) 1431.

[4] P.P. Prosini, R. Mancini, L. Petrucci, V. Contini, P. Villano, Solid State Ionics 144 (2001) 185. [5] S. Panero, D. Satolli, M. Salomon, B. Scosati, Electrochem. Commun. 2 (2000) 810. [6] R.I. Eglitis, G. Borstel, Phys. Stat. Sol. (a) R13 (2005) 202. [7] E.M. Sorensen, S.J. Barry, H. Jung, H.K. Jung, J.R. Rondinelli, Chem. Mater. 18 (2006) 482. [8] D. Li, C. Ouyang, J. Shu, J. Jiang, Z.X. Wang, L.Q. Chen, Phys. Stat.Sol. (b) (2006) 1. [9] X.L. Yao, S. Xie, C.H. Chen, et al., Elctrochimica. Acta. 50 (2005) 4076. [10] K. Nakahara, R. Nakajima, T. Matsushima, H. Majima, J. Power Sources 117 (2003) 131. [11] K. Zaghib, M. Simoneau, M. Armand, M. Gauthier, J. Power Sources 81–82 (1999) 300. [12] G. Yang, G. Wang, W. Hou, J. Phys. Chem. B. 109 (2005) 11186. [13] J.J. Zhu, J.M. Zhu, X.H. Liao, J.L. Fang, M.G. Zhou, Chen, Mater. Lett. 53 (2002) 12. [14] P.Y. Hong, T.T. Pan, H.Z. Li, J. Magn. Magn. Mater. 303 (2006) 60. [15] X.H. Liao, J.M. Zhu, J.J. Zhu, J.Z. Xu, H.Y. Chen, Chem. Commun. (2001) 937. [16] J.H. Liang, Z.X. Deng, X. Jiang, F. Li, Y. Li, Inorg. Chem. 41 (2002) 3602. [17] W.Y. Yu, W.X. Tu, H.F. Liu, Langmuir 15 (1999) 6. [18] W.X. Yu, H.Y. Liu, Chem. Mater. 12 (2000) 564. [19] W.X. Tu, H.F. Liu, J. Mater. Chem. 10 (2000) 2207. [20] M. Nakayama, K. Watanabe, H. Ikuta, Y. Uchimoto, M. Wakihara, Solid State Ionics 164 (2003) 352. [21] Y. Wang, J.Y. Lee, J. Power Sources 144 (2005) 220. [22] H.J. Hao, Q.Y. Lan, Z.H. Xu, X.Q. Liu, X.Y. Ji, Solid State Ionics 176 (2005) 1201. [23] J.R. Li, Z.L. Tang, Z.T. Zhang, Electrochem. Commun. 7 (2005) 894. [24] G.X. Wang, D.H. Brandhurst, S.X. Dou, H.X. Liu, J. Power Sources 83 (1999) 159. [25] H.W. Lu, W. Zeng, Y.S. Li, Z.W. Fu, J. Power Sources 164 (2007) 874. [26] S. Bach, J.P. Pereira-Ramos, N. Baffier, J. Power Sources 81–82 (1999) 273.