- Email: [email protected]
Electrochemical performance of Si-multiwall carbon nanotube nanocomposite anode synthesized by thermal plasma
Thin Solid Films 587 (2015) 14–19
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Electrochemical performance of Si-multiwall carbon nanotube nanocomposite anode synthesized by thermal plasma Ye-Seul Na, Hyeonseok Yoo, Tae-Hee Kim, Jinsub Choi, Wan In Lee, Sooseok Choi ⁎, Dong-Wha Park ⁎ Department of Chemistry and Chemical Engineering and Regional Innovation Center for Environmental Technology of Thermal Plasma, Inha University, 100 Inha-ro, Nam-gu, Incheon 402-751, Republic of Korea
a r t i c l e
i n f o
Available online 27 December 2014 Keywords: Thermal plasma Silicon Multiwall carbon nanotubes Composite Lithium-ion battery Discharge capacity Scanning electron microscopy
a b s t r a c t Lithium-ion (Li-ion) batteries are widely used in electric devices and vehicles. Silicon is a promising material for the anode of Li-ion battery due to high theoretical specific capacity. However, it shows large volume changes during charge–discharge cycles leading to the pulverization of electrode. In order to improve such disadvantage, a multiwall carbon nanotube (MWCNT) has been used with silicon as composite material. In this work, Si-MWCNT nanocomposite was prepared in thermal plasma by attaching silicon nanoparticles to MWCNT column. Electrochemical tests for raw materials and synthesized nanocomposites were carried out. The discharge capacities of silicon, MWCNT, synthesized nanocomposites collected from a reaction tube, and a chamber were 4000, 310, 200, and 1447 mAh/g, respectively. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Lithium-ion (Li-ion) battery, which is one of secondary batteries, has been developed as a promising rechargeable power storage system because of its excellent characteristics such as high energy density and low self-discharge [1,2]. As a result, Li-ion batteries are widely used in electric devices and vehicles devices [3]. Recently, many researches are attempted to increase the storage capacity and cycling life time of Li-ion battery by changing anode materials. Silicon is widely used as anode material due to its high theoretical specific capacity of 4200 mAh/g, which is corresponding to the Li22Si5. Silicon-based electrodes, however, have enormous volume expansion up to 300% during charging and discharging cycle leading to electrode pulverization and poor capacity retention [4,5]. In order to overcome such problem, carbon materials are incorporated into the anode as a form of Si-carbon composite. The Si-carbon composite has good properties of high electronic conductivity, low-voltage range, and uniform cycle performance [6]. Especially, carbon nanotubes (CNTs) are suitable for the electrode of Li-ion batteries because of the high specific capacity of ~470 mAh/g compared with other carbon materials such as the graphite anode of 372 mAh/g. In addition, CNTs have unique cylindrical structure, which controls the swell of silicon nanoparticles due to the high tensile strength CNT [7,8]. Furthermore, CNTs can diffuse lithium ions into stable sites leading to high lithium capacity and can support silicon particles to reduce the pulverization of electrode. According to
⁎ Corresponding authors. E-mail addresses: [email protected] (S. Choi), [email protected] (D.-W. Park).
http://dx.doi.org/10.1016/j.tsf.2014.12.038 0040-6090/© 2015 Elsevier B.V. All rights reserved.
the number of graphene sheets, CNTs are divided into a single-walled carbon nanotube (SWCNT) and a multiwalled carbon nanotube (MWCNT), which consists of a single rolled sheet and multiple rolled sheets, respectively. The MWCNT is suitable to transfer charges compared with SWCNT because it has large amount of graphitic tubes that hold a lot of defects of stable sites for silicon particles in various size [9]. Si-MWCNT nanocomposite has been synthesized by pyrolysis, chemical vapor deposition, and sol-gel methods, which require several process steps [10–12]. On the other hand, thermal plasma has been applied for synthesizing nanomaterials in a single process step by high temperature and rapid quenching [13]. In addition, thermal plasma method is environmentally clean process for synthesizing Si-MWCNT nanocomposite because it does not require pretreatment and additional chemicals except Si powder leading to a low contamination. Therefore, a low contamination in Si-MWCNT nanocomposite product is achievable by using thermal plasma synthesis method. Moreover, the size and the content of Si nanoparticles in Si-MWCNT nanocomposite are controllable by manipulating operation condition. In this work, Si-MWCNT nanocomposite was synthesized by a nontransferred direct current (DC) arc plasma system using a micro-sized silicon powder and a commercial MWCNT. This study was focused on to evaluate the electrochemical performance of synthesized Si-MWCNT nanocomposite. The chemical bonding change of product was confirmed by X-ray diffractometry with CuKα source (XRD, DMAX-2500, Rigaku Co.). Size distributions and morphologies of raw materials and products were observed by field emission scanning electron microscopy at 15 kV (FE-SEM, S-4300, Hitachi Co.). In addition, electrochemical performance tests for raw materials and synthesized Si-MWCNT nanocomposites were conducted to compare their capacities.
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
2. Experiment 2.1. Synthesis of Si-MWCNT nanocomposite Fig. 1 shows the schematic diagram of the non-transferred DC thermal plasma system to prepare Si-MWCNT nanocomposite. The system consists of a power supply for generating thermal plasma, a DC plasma torch, two powder feeders, and a water-cooled reaction tube and a chamber. The powder feeder 1 (MF, Technoserve, Japan) was used to carry micro-sized silicon powder (99%, Aldrich, USA), which was injected into the plasma core at anode region, while the powder feeder 2 (PFV-100, TEKNA, Canada) was used to supply MWCNT (~ 95 wt. %, Hanwha Chemical, Korea), which was injected into the peripheral region of plasma flame at the entrance of the reaction tube. Silicon powder injected into the inside of the plasma torch was completely evaporated to form nanoparticles. On the other hand, MWCNT injected into the relatively low temperature region maintains its original structure well without chemical and physical changes. The experimental conditions are summarized in Table 1. Because Si-MWCNT nanocomposite was well synthesized at a relatively high power in our previous study, the input power was fixed at 12.4 kW in the present work. The reaction tube and the chamber were purged by argon gas at 20 L/min before plasma generation to avoid the oxidation of product. All experiments were operated at atmospheric pressure during system operation for 20 min. Flow rates of argon used for plasma forming gas, and Si carrier gas and MWCNT carrier gas were maintained at 15 L/min, 7 L/min, and 7 L/min, respectively. In such condition, feed rates of silicon powder and MWCNT were measured at 0.10 g/min and 0.25 g/min, respectively. After the synthesis experiment, products were collected from the reaction tube and the chamber. The chemical bonding of raw materials and products was investigated by XRD. In addition, the sizes and the morphology were confirmed by an FE-SEM. 2.2. Electrochemical characterization The electrochemical performances of silicon powder, MWCNT, and synthesized nanocomposites collected from the reaction tube and the
15
Table 1 Experimental condition for the synthesis of Si-MWCNT nanocomposites by thermal plasma process. Power (kW) Feed rate of feedstock (g/min)
Si MWCNT Mass ratio of feedstock (Si/MWCNT) Injection position Si MWCNT Plasma gas(L/min) Carrier gas (L/min)
12.4 0.1 0.25 0.4 Torch Reaction tube Ar 15 Ar 7 (Si and MWCNT, respectively)
chamber were analyzed after fabricating test samples with Swageloktype cells. Each sample was mixed with activated charcoal and polyvinylidene fluoride used as binder material at the mass ration of 3:1:1. The mixture was dissolved in an excessive N-methylpyrrolidone solution to form slurry. Then the slurry was coated on a Cu collector, which is punched to circular shape with 12 mm in diameter, and dried at 60 °C for 2 h. Lithium metal was cut to circular shape with 11 mm in diameter, and it used as a counterelectrode. In addition, a Cu emitter was punched to circular shape with 12 mm in diameter. For the preparation of electrolyte, 1 mol of LiPF6 was dissolved in the mixture of ethylene carbonate, dimethyl carbonate, and ethyl methyl carbonate at the same volume faction. The working electrode and the counterelectrode were separated by glass microfiber filters punched to 12 mm in diameter. The assembly process of test cells was carried in an inert Ar atmosphere using a glove box to prevent moisture and oxidation. The image of cell assembly is shown in Fig. 2. The charge–discharge cycling was performed in the voltage range from 0.01 to 1.5 V at the current density of 100 mA/g. 3. Results and discussion Fig. 3 exhibits XRD patterns of raw materials of the silicon powder and the commercial MWCNT. It is observed that silicon peaks at the 28.4°, 47.0°, 56.0°, 59.0°, 76.5°, and 88.0° corresponding cubic structure in Fig. 3(a), while carbon peaks at 26.5° and 42.3° indicate graphitic
Fig. 1. Schematic diagram of the non-transferred thermal plasma system to synthesize Si-MWCNT nanocomposite.
16
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
Fig. 2. The assembly diagram of test cell for the lithium-ion battery.
carbon corresponding hexagonal structure in Fig. 3(b). Morphologies of silicon and MWCNT used as raw materials are indicated in Fig. 4. Silicon raw material has irregular plate shape in the size range of 10–40 μm, and MWCNT has tangled structure, which has relatively uniform diameter of 30–40 nm. XRD patterns of synthesized Si-MWCNT nanocomposite are presented in Fig. 5. Synthesized products were collected from the reaction tube and the chamber, respectively. It is confirmed that silicon and MWCNT
peaks are measured at the same position compared with Fig. 3. Such results mean that each structure of silicon and MWCNT could be maintained well in thermal plasma. The intensity of MWCNT in Si-MWCNT nanocomposite is relatively lower than that of silicon at the 26.5° because a small amount of MWCNT was injected compared with silicon. In both Fig. 5(a) and (b), XRD peaks for silicon carbide with lower intensity were observed at 35.6°, 59.9°, and 71.7° corresponding to cubic structure. MWCNT was injected into the peripheral area of plasma jet where plasma temperature is relatively low to prevent the complete sublimation of MWCNT, while the outermost wall of MWCNT was slightly sublimated to provide sites for Si nanoparticles. Therefore, evaporated carbon reacts with Si vapors in plasma jet leading to the formation of silicon carbide. The intensity of silicon peaks in
Fig. 3. The XRD patterns of raw materials; (a) silicon power and (b) MWCNT.
Fig. 4. FE-SEM images of raw materials; (a) silicon powder and (b) MWCNT.
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
17
Fig. 6. The FE-SEM of synthesized Si-MWCNT nanocomposite collected from (a) reaction tube and (b) chamber.
Fig. 5. XRD patterns of synthesized Si-MWCNT nanocomposite collected from (a) the reaction tube and (b) the chamber.
Fig. 5(a) is higher than that of Fig. 5(b). It is expected that silicon powder injected into the high plasma temperature region was completely evaporated. Silicon vapors in high vapor pressure were saturated quickly in the reaction tube forming silicon nanoparticles. Subsequently, silicon nanoparticles were attached on to the surface of MWCNT and some of silicon nanoparticles as deposited on the surface of the reaction tube without the attachment with MWCNT. Therefore, relatively large amount of silicon nanoparticles were found at the reaction tube rather than the chamber. Fig. 6 shows FE-SEM images of synthesized Si-MWCNT nanocomposites collected from the reaction tube and the chamber, respectively. Both of silicon nanoparticles and nanotubes are found regardless of collection position. It is observed that although the mean diameter of MWCNT was decreased from 27 to 22 nm by the thermal damage of outermost wall during thermal plasma treatment, MWCNT structure in synthesized nanocomposite was well maintained. On the other hand, the shape of silicon powder was changed from micro-sized irregular plates to nanosized spherical particles. The size distribution of silicon nanoparticles, which were heavily agglomerated was in the range from 30 to 60 nm. The size uniformity of silicon nanoparticles is important factor for using Si-MWCNT nanocomposite as an electrode material. A narrow-sized distribution of nanoparticles in thermal plasma synthesis can be achieved by optimizing operation and design conditions such as the kind and flow rate of working gas and the geometry of reaction tube [14]. Silicon nanoparticles that have size below 150 nm, however, do not cause the cracking of the electrode during lithiation
reaction [15]. The size of silicon nanoparticles prepared in this work was distributed from 30 to 60 nm. Therefore, a stable operation of a Li-ion battery is expected by using the Si-MWCNT nanocomposite prepared in the present work. In comparison between Fig. 6(a) and (b), silicon nanoparticles in Si-MWCNT nanocomposite are uniformly dispersed on MWCNT column in the case of the chamber. It is because silicon nanoparticles have enough time to be attached on the MWCNT surface in the case of the chamber while some of silicon nanoparticles are directly deposited on the reaction tube as discussed above. In addition, since the surface temperature of injected MWCNT is lower than plasma jet, saturated vapors are easily condensed on the MWCNT column, and then synthesized Si-MWCNT nanocomposites are flow down to the chamber region. Therefore, relatively large amount of agglomerated silicon nanoparticles were observed at the reaction tube, while silicon nanoparticles are uniformly dispersed on the MWCNT column in the Si-MWCNT nanocomposite collected at the chamber. Fig. 7 shows the first discharge curve of raw materials and synthesized Si-MWCNT nanocomposites at the fixed current rate of 100 mA/g. In Fig. 7(a), the silicon electrode shows the highest discharge capacity of 4000 mAh/g, which is similar with the theoretical capacity of 4200 mAh/g. Such result means that Si atom was formed by being removed from Li22Si5 alloy. A Si atom is delithiated during the discharge, and then 4.4 Li atoms are formed leading to high capacity. First discharge capacity, however, is rapidly decreased as voltage is increased to 0.1 V, and then capacity is slowly decreased between in the range of 0.1 V to 0.6 V. As a result, it is confirmed that the silicon electrode shows instability leading to rapid capacity loss. On the other hand, the MWCNT electrode in Fig. 7(b) shows the first discharge capacity of 310 mAh/g because a Li atom is formed by the delithiating of 6 carbon atoms from LiC6 alloy. Therefore, the MWCNT electrode shows low capacity than the silicon
18
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
Fig. 7. The first discharge curve of at a current rate of 100 mA/g with different anode materials of (a) pristine silicon, (b) pristine MWCNT and synthesized Si-MWCNT nanocomposites collected from (c) the reaction tube and (d) the chamber.
electrode. In the MWCNT electrode, first discharge capacity is slowly decreased with increasing the voltage to 0.6 V, and the capacity is rapidly decreased in a high voltage over 0.6 V. In comparison to the silicon electrode, the MWCNT electrode shows stability during discharge, though the electrode capacity of MWCNT is noticeably lower than that of silicon. The first discharge curves of synthesized Si-MWCNT nanocomposites are presented in Fig. 7(c) and (d). The first discharge capacity of Si-MWCNT nanocomposite collected from the reaction tube is 200 mAh/g. The first discharge capacity of Si-MWCNT from the reaction tube is lower capacity than that of pristine MWCNT, and it is sharply decreased with increasing the voltage. Such results are attributed an irregular dispersion and agglomeration of silicon nanoparticles. Therefore, lithium ions and charges in the anode are difficult to move to cathode leading to poor capacity. On the other hand, the first discharge capacity of Si-MWCNT nanocomposite collected from the chamber exhibits 1447 mAh/g. This measured capacity is lower than that of pristine silicon but higher than that of pristine MWCNT. The discharge capacity of nanocomposite collected from the chamber was higher than that from the reaction tube. The volume expansion of silicon nanoparticles in the discharge reaction is suppressed when they are dispersed uniformly on MWCNT, which acts as the supporter. Since a relatively uniform dispersion of silicon nanoparticles on MWCNT were observed in Si-MWCNT nanocomposite collected from the chamber, the discharge capacity of nanocomposite collected from the chamber is higher than that from the reaction tube. The capacity of the Si-MWCNT electrode fabricated products from the chamber is slowly decreased with increasing voltage up to 0.7 V, then the capacity is sharply decreased in the voltage range of 0.7–2.3 V. Such results of the Si-MWCNT electrode fabricated using product from the chamber are similar with those of the MWCNT electrode, which shows a high stability. Therefore, it is expected that the electrode fabricated using Si-MWCNT nanocomposite collected from the chamber can prevent the volume expansion of silicon nanoparticles by MWCNT. In the voltage of 2.3–2.5 V, it is observed that a distinguishable plateau in Fig. 7(d) because of the formation of solid electrolyte interphase (SEI), which is formed by the decomposition of organic electrolyte at the
electrode surface during first cycle generally. The SEI layer acts as an electrically insulating to prevent further decomposition of the electrolyte leading to suppress self-discharge [16,17]. Fig. 8 exhibits curves of first charge capacity with different electrodes of pristine silicon, pristine MWCNT, and Si-MWCNT nanocomposites. Although the range of charge is set up between 0.01 and 1.5 V, the initial voltage is shown at least 0.3 V in all electrodes. In addition, poor first charge capacities of silicon, MWCNT, and Si-MWCNT nanocomposites collected from reaction tube and chamber of 7.7, 7.9, 7.0, and 8.3 mAh/g are measured. Those values mean that the resistance of electrodes is pretty high because of electrode thickness and impurities. MWCNT used in the present work has a high bulk density of 0.10 g/cm3 leading to the loss of power density, poor cycling performance, and instability [18]. When electrode thickness is increased, charge transfer
Fig. 8. First charge curves with different anode materials of micro-sized silicon, MWCNT, and synthesized Si-MWCNT nanocomposites collected from the reaction tube and the chamber.
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
resistance is increased with the decrease of lithium diffusion [19–21]. Therefore, electrodes exhibit high initial voltage and poor charge capacity. Si-MWCNT nanocomposite collected from the chamber is the highest charge capacity among all electrodes. These results show that Si-MWCNT nanocomposite, which is silicon nanoparticles attached to MWCNT column, was well synthesized for anode material with higher capacity than the MWCNT electrode and more stable than the silicon electrode. Fig. 9 shows FE-SEM images of Si-MWCNT nanocomposite electrodes after discharge–charge cycle. In Fig. 9(a), Si-MWCNT nanocomposite collected from reaction tube, the morphology is similar with Fig. 6(a), which shows agglomerated silicon nanoparticles around MWCNT. On the other hand, Fig. 9(b) for Si-MWCNT electrode fabricated by products from the chamber shows modified structure from Fig. 6(b). Silicon nanoparticles and MWCNT are slightly agglomerated after a charge–discharge cycle because organic solution is mixed with samples during cell assembly. Despite cycling, however, it is observed that silicon nanoparticles maintained their spherical shape without volume expansion, and they are well dispersed around MWCNT column in the case of Si-MWCNT collected from the chamber.
4. Conclusions Si-MWCNT nanocomposite was synthesized by thermal plasma. The nanocomposite structure was silicon nanoparticles attached to MWCNT column, and well-distributed silicon nanoparticle was found in the chamber rather than the reaction tube. It is expected that silicon vapors are saturated and nanoparticles are produced in the reaction tube
Fig. 9. FE-SEM images of electrodes fabricated using Si-MWCNT nanocomposites collected from (a) the reaction tube and (b) the chamber after a charge–discharge cycle.
19
region. There, some of silicon nanoparticles are deposited on the surface of the reaction tube. In addition, silicon vapors could be condensed on MWCNT column directly because MWCNT is injected into the peripheral region of plasma jet. Therefore, well-synthesized Si-MWCNT nanocomposites flow to the chamber. The electrochemical test of raw materials and Si-MWCNT products collected from the reaction tube and the chamber was carried out. From the results, Si-MWCNT nanocomposite collected from the chamber shows relatively stable charge–discharge characteristics with an enhanced capacity compared with pristine MWCNT. Acknowledgments This work was supported by the World Class 300 Project (10043264, Development of the electrode materials for high efficiency (21%) and low-cost c-Si solar cell) funded by the Ministry of Trade, Industry and Energy (MOTIE, Korea). References [1] M.M. Thackeray, C. Wolverton, E.D. Isaacs, Electrical energy storage for transportation—approaching the limits of, and going beyond, lithium-ion batteries, Energy Environ. Sci. 5 (2012) 7851. [2] Y.P. Wu, C. Wan, C. Jiang, S.B. Fang, Lithium ion secondary batteries, Chemical Industry Press, Beijing, 2002. (references therein). [3] T. Kojima, T. Ishizu, T. Horiba, M. Yoshikawa, Development of lithium-ion battery for fuel cell hybrid electric vehicle application, J. Power Sources 189 (2009) 859. [4] J.R. Szczech, S. Jin, Nanostructured silicon for high capacity lithium battery anodes, Energy Environ. Sci. 4 (2011) 56. [5] L.Y. Beaulieu, K.W. Eberman, R.L. Turner, L.J. Krause, J.R. Dahn, Colossal reversible volume changes in lithium alloys, Electrochem. Solid-State Lett. 4 (2001) A137. [6] C. de las Casas, W. Li, A review of application of carbon nanotubes for lithium ion battery anode material, J. Power Sources 208 (2012) 74. [7] Y. NuLi, J. Yang, M. Jiang, Synthesis and characterization of Sb/CNT and Bi/CNT composites as anode materials for lithium-ion batteries, Mater. Lett. 62 (2008) 2092. [8] J.P. Salvetat, J.M. Bonard, N.H. Thomson, A.J. Kulik, L. Forró, W. Benoit, L. Zuppiroli, Mechanical properties of carbon nanotubes, Appl. Phys. 69 (1999) 255. [9] C.H. Mi, G.S. Cao, X.B. Zhao, A non-GIC mechanism of lithium storage in chemical etched MWNTs, J. Electroanal. Chem. 562 (2004) 217. [10] R.A. Afrea, T. Sogaa, T. Jimboa, M. Kumarb, Y. Andob, M. Sharon, Growth of vertically aligned carbon nanotubes on silicon and quartz substrate by spray pyrolysis of a natural precursor: Turpentine oil, Chem. Phys. Lett. 414 (2005) 6. [11] S.C. Wang, F. Yang, M. Silva, A. Zarow, Y. Wang, Z. Iqbal, Membrane-less and mediator-free enzymatic biofuel cell using carbon nanotube/porous silicon electrodes, Electrochem. Commun. 11 (2009) 34. [12] J. Bae, Fabrication of carbon microcapsules containing silicon nanoparticles–carbon nanotubes nanocomposite by sol-gel method for anode in lithium ion battery, J. Solid State Chem. 184 (2011) 1749. [13] M. Shigeta, A.B. Murphy, Thermal plasmas for nanofabrication, J. Phys. D. Appl. Phys. 44 (2011) 174025. [14] J.-G. Li, Y. Sakka, Controlled thermal plasma processing of ceramic nanopowders, in: S. Somiya (Ed.), Handbook of Advanced Ceramics, Academic Press, Japan, 2013, p. 979. [15] X.H. Liu, L. Zhong, S. Huang, S.X. Mao, T. Zhu, J.Y. Huang, Size-dependent fracture of silicon nanoparticles during lithiation, ACS Nano 6 (2012) 1522. [16] C.K. Chan, R. Ruffo, S.S. Hong, Y. Cui, Surface chemistry and morphology of the solid electrolyte interphase on silicon nanowire lithium-ion battery anodes, J. Power Sources 189 (2009) 1132. [17] K. Zaghib, AFM Study on SEI growth on graphite negative electrode at elevated temperatures, in: K. Striebel, K. Zaghib, D. Guyomard (Eds.),Anodes, U.S.A., 11–19, 2003, Lithium and Lithium-ion Batteries: Proceedings of the International Symposium 28, 2004. [18] H. Zheng, J. Li, X. Song, G. Liu, V.S. Battaglia, A comprehensive understanding of electrode thickness effects on the electrochemical performances of Li-ion battery cathodes, Electrochim. Acta 71 (2012) 258. [19] U. Tocoglu, O. Cevher, M.O. Guler, H. Akbulut, Core–shell tin-multi walled carbon nanotube composite anodes for lithium ion batteries, Int. J. Hydrogen Energy (2014). http://dx.doi.org/10.1016/j.ijhydene.2014.05.053. [20] Y. Ito, M. Kawakubo, M. Ueno, H. Okuma, Q. Si, T. Kobayashi, K. Hanai, N. Imanishi, A. Hirano, M.B. Phillipps, Y. Takeda, O. Yamamoto, Carbon anodes for solid polymer electrolyte lithium-ion batteries, J. Power Sources 214 (2012) 84. [21] Y. Hu, X. Li, D. Geng, M. Cai, R. Li, X. Sun, Influence of paper thickness on the electrochemical performances of graphene papers as an anode for lithium ion batteries, Electrochim. Acta 91 (2013) 227–233.
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Electrochemical performance of Si-multiwall carbon nanotube nanocomposite anode synthesized by thermal plasma Ye-Seul Na, Hyeonseok Yoo, Tae-Hee Kim, Jinsub Choi, Wan In Lee, Sooseok Choi ⁎, Dong-Wha Park ⁎ Department of Chemistry and Chemical Engineering and Regional Innovation Center for Environmental Technology of Thermal Plasma, Inha University, 100 Inha-ro, Nam-gu, Incheon 402-751, Republic of Korea
a r t i c l e
i n f o
Available online 27 December 2014 Keywords: Thermal plasma Silicon Multiwall carbon nanotubes Composite Lithium-ion battery Discharge capacity Scanning electron microscopy
a b s t r a c t Lithium-ion (Li-ion) batteries are widely used in electric devices and vehicles. Silicon is a promising material for the anode of Li-ion battery due to high theoretical specific capacity. However, it shows large volume changes during charge–discharge cycles leading to the pulverization of electrode. In order to improve such disadvantage, a multiwall carbon nanotube (MWCNT) has been used with silicon as composite material. In this work, Si-MWCNT nanocomposite was prepared in thermal plasma by attaching silicon nanoparticles to MWCNT column. Electrochemical tests for raw materials and synthesized nanocomposites were carried out. The discharge capacities of silicon, MWCNT, synthesized nanocomposites collected from a reaction tube, and a chamber were 4000, 310, 200, and 1447 mAh/g, respectively. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Lithium-ion (Li-ion) battery, which is one of secondary batteries, has been developed as a promising rechargeable power storage system because of its excellent characteristics such as high energy density and low self-discharge [1,2]. As a result, Li-ion batteries are widely used in electric devices and vehicles devices [3]. Recently, many researches are attempted to increase the storage capacity and cycling life time of Li-ion battery by changing anode materials. Silicon is widely used as anode material due to its high theoretical specific capacity of 4200 mAh/g, which is corresponding to the Li22Si5. Silicon-based electrodes, however, have enormous volume expansion up to 300% during charging and discharging cycle leading to electrode pulverization and poor capacity retention [4,5]. In order to overcome such problem, carbon materials are incorporated into the anode as a form of Si-carbon composite. The Si-carbon composite has good properties of high electronic conductivity, low-voltage range, and uniform cycle performance [6]. Especially, carbon nanotubes (CNTs) are suitable for the electrode of Li-ion batteries because of the high specific capacity of ~470 mAh/g compared with other carbon materials such as the graphite anode of 372 mAh/g. In addition, CNTs have unique cylindrical structure, which controls the swell of silicon nanoparticles due to the high tensile strength CNT [7,8]. Furthermore, CNTs can diffuse lithium ions into stable sites leading to high lithium capacity and can support silicon particles to reduce the pulverization of electrode. According to
⁎ Corresponding authors. E-mail addresses: [email protected] (S. Choi), [email protected] (D.-W. Park).
http://dx.doi.org/10.1016/j.tsf.2014.12.038 0040-6090/© 2015 Elsevier B.V. All rights reserved.
the number of graphene sheets, CNTs are divided into a single-walled carbon nanotube (SWCNT) and a multiwalled carbon nanotube (MWCNT), which consists of a single rolled sheet and multiple rolled sheets, respectively. The MWCNT is suitable to transfer charges compared with SWCNT because it has large amount of graphitic tubes that hold a lot of defects of stable sites for silicon particles in various size [9]. Si-MWCNT nanocomposite has been synthesized by pyrolysis, chemical vapor deposition, and sol-gel methods, which require several process steps [10–12]. On the other hand, thermal plasma has been applied for synthesizing nanomaterials in a single process step by high temperature and rapid quenching [13]. In addition, thermal plasma method is environmentally clean process for synthesizing Si-MWCNT nanocomposite because it does not require pretreatment and additional chemicals except Si powder leading to a low contamination. Therefore, a low contamination in Si-MWCNT nanocomposite product is achievable by using thermal plasma synthesis method. Moreover, the size and the content of Si nanoparticles in Si-MWCNT nanocomposite are controllable by manipulating operation condition. In this work, Si-MWCNT nanocomposite was synthesized by a nontransferred direct current (DC) arc plasma system using a micro-sized silicon powder and a commercial MWCNT. This study was focused on to evaluate the electrochemical performance of synthesized Si-MWCNT nanocomposite. The chemical bonding change of product was confirmed by X-ray diffractometry with CuKα source (XRD, DMAX-2500, Rigaku Co.). Size distributions and morphologies of raw materials and products were observed by field emission scanning electron microscopy at 15 kV (FE-SEM, S-4300, Hitachi Co.). In addition, electrochemical performance tests for raw materials and synthesized Si-MWCNT nanocomposites were conducted to compare their capacities.
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
2. Experiment 2.1. Synthesis of Si-MWCNT nanocomposite Fig. 1 shows the schematic diagram of the non-transferred DC thermal plasma system to prepare Si-MWCNT nanocomposite. The system consists of a power supply for generating thermal plasma, a DC plasma torch, two powder feeders, and a water-cooled reaction tube and a chamber. The powder feeder 1 (MF, Technoserve, Japan) was used to carry micro-sized silicon powder (99%, Aldrich, USA), which was injected into the plasma core at anode region, while the powder feeder 2 (PFV-100, TEKNA, Canada) was used to supply MWCNT (~ 95 wt. %, Hanwha Chemical, Korea), which was injected into the peripheral region of plasma flame at the entrance of the reaction tube. Silicon powder injected into the inside of the plasma torch was completely evaporated to form nanoparticles. On the other hand, MWCNT injected into the relatively low temperature region maintains its original structure well without chemical and physical changes. The experimental conditions are summarized in Table 1. Because Si-MWCNT nanocomposite was well synthesized at a relatively high power in our previous study, the input power was fixed at 12.4 kW in the present work. The reaction tube and the chamber were purged by argon gas at 20 L/min before plasma generation to avoid the oxidation of product. All experiments were operated at atmospheric pressure during system operation for 20 min. Flow rates of argon used for plasma forming gas, and Si carrier gas and MWCNT carrier gas were maintained at 15 L/min, 7 L/min, and 7 L/min, respectively. In such condition, feed rates of silicon powder and MWCNT were measured at 0.10 g/min and 0.25 g/min, respectively. After the synthesis experiment, products were collected from the reaction tube and the chamber. The chemical bonding of raw materials and products was investigated by XRD. In addition, the sizes and the morphology were confirmed by an FE-SEM. 2.2. Electrochemical characterization The electrochemical performances of silicon powder, MWCNT, and synthesized nanocomposites collected from the reaction tube and the
15
Table 1 Experimental condition for the synthesis of Si-MWCNT nanocomposites by thermal plasma process. Power (kW) Feed rate of feedstock (g/min)
Si MWCNT Mass ratio of feedstock (Si/MWCNT) Injection position Si MWCNT Plasma gas(L/min) Carrier gas (L/min)
12.4 0.1 0.25 0.4 Torch Reaction tube Ar 15 Ar 7 (Si and MWCNT, respectively)
chamber were analyzed after fabricating test samples with Swageloktype cells. Each sample was mixed with activated charcoal and polyvinylidene fluoride used as binder material at the mass ration of 3:1:1. The mixture was dissolved in an excessive N-methylpyrrolidone solution to form slurry. Then the slurry was coated on a Cu collector, which is punched to circular shape with 12 mm in diameter, and dried at 60 °C for 2 h. Lithium metal was cut to circular shape with 11 mm in diameter, and it used as a counterelectrode. In addition, a Cu emitter was punched to circular shape with 12 mm in diameter. For the preparation of electrolyte, 1 mol of LiPF6 was dissolved in the mixture of ethylene carbonate, dimethyl carbonate, and ethyl methyl carbonate at the same volume faction. The working electrode and the counterelectrode were separated by glass microfiber filters punched to 12 mm in diameter. The assembly process of test cells was carried in an inert Ar atmosphere using a glove box to prevent moisture and oxidation. The image of cell assembly is shown in Fig. 2. The charge–discharge cycling was performed in the voltage range from 0.01 to 1.5 V at the current density of 100 mA/g. 3. Results and discussion Fig. 3 exhibits XRD patterns of raw materials of the silicon powder and the commercial MWCNT. It is observed that silicon peaks at the 28.4°, 47.0°, 56.0°, 59.0°, 76.5°, and 88.0° corresponding cubic structure in Fig. 3(a), while carbon peaks at 26.5° and 42.3° indicate graphitic
Fig. 1. Schematic diagram of the non-transferred thermal plasma system to synthesize Si-MWCNT nanocomposite.
16
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
Fig. 2. The assembly diagram of test cell for the lithium-ion battery.
carbon corresponding hexagonal structure in Fig. 3(b). Morphologies of silicon and MWCNT used as raw materials are indicated in Fig. 4. Silicon raw material has irregular plate shape in the size range of 10–40 μm, and MWCNT has tangled structure, which has relatively uniform diameter of 30–40 nm. XRD patterns of synthesized Si-MWCNT nanocomposite are presented in Fig. 5. Synthesized products were collected from the reaction tube and the chamber, respectively. It is confirmed that silicon and MWCNT
peaks are measured at the same position compared with Fig. 3. Such results mean that each structure of silicon and MWCNT could be maintained well in thermal plasma. The intensity of MWCNT in Si-MWCNT nanocomposite is relatively lower than that of silicon at the 26.5° because a small amount of MWCNT was injected compared with silicon. In both Fig. 5(a) and (b), XRD peaks for silicon carbide with lower intensity were observed at 35.6°, 59.9°, and 71.7° corresponding to cubic structure. MWCNT was injected into the peripheral area of plasma jet where plasma temperature is relatively low to prevent the complete sublimation of MWCNT, while the outermost wall of MWCNT was slightly sublimated to provide sites for Si nanoparticles. Therefore, evaporated carbon reacts with Si vapors in plasma jet leading to the formation of silicon carbide. The intensity of silicon peaks in
Fig. 3. The XRD patterns of raw materials; (a) silicon power and (b) MWCNT.
Fig. 4. FE-SEM images of raw materials; (a) silicon powder and (b) MWCNT.
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
17
Fig. 6. The FE-SEM of synthesized Si-MWCNT nanocomposite collected from (a) reaction tube and (b) chamber.
Fig. 5. XRD patterns of synthesized Si-MWCNT nanocomposite collected from (a) the reaction tube and (b) the chamber.
Fig. 5(a) is higher than that of Fig. 5(b). It is expected that silicon powder injected into the high plasma temperature region was completely evaporated. Silicon vapors in high vapor pressure were saturated quickly in the reaction tube forming silicon nanoparticles. Subsequently, silicon nanoparticles were attached on to the surface of MWCNT and some of silicon nanoparticles as deposited on the surface of the reaction tube without the attachment with MWCNT. Therefore, relatively large amount of silicon nanoparticles were found at the reaction tube rather than the chamber. Fig. 6 shows FE-SEM images of synthesized Si-MWCNT nanocomposites collected from the reaction tube and the chamber, respectively. Both of silicon nanoparticles and nanotubes are found regardless of collection position. It is observed that although the mean diameter of MWCNT was decreased from 27 to 22 nm by the thermal damage of outermost wall during thermal plasma treatment, MWCNT structure in synthesized nanocomposite was well maintained. On the other hand, the shape of silicon powder was changed from micro-sized irregular plates to nanosized spherical particles. The size distribution of silicon nanoparticles, which were heavily agglomerated was in the range from 30 to 60 nm. The size uniformity of silicon nanoparticles is important factor for using Si-MWCNT nanocomposite as an electrode material. A narrow-sized distribution of nanoparticles in thermal plasma synthesis can be achieved by optimizing operation and design conditions such as the kind and flow rate of working gas and the geometry of reaction tube [14]. Silicon nanoparticles that have size below 150 nm, however, do not cause the cracking of the electrode during lithiation
reaction [15]. The size of silicon nanoparticles prepared in this work was distributed from 30 to 60 nm. Therefore, a stable operation of a Li-ion battery is expected by using the Si-MWCNT nanocomposite prepared in the present work. In comparison between Fig. 6(a) and (b), silicon nanoparticles in Si-MWCNT nanocomposite are uniformly dispersed on MWCNT column in the case of the chamber. It is because silicon nanoparticles have enough time to be attached on the MWCNT surface in the case of the chamber while some of silicon nanoparticles are directly deposited on the reaction tube as discussed above. In addition, since the surface temperature of injected MWCNT is lower than plasma jet, saturated vapors are easily condensed on the MWCNT column, and then synthesized Si-MWCNT nanocomposites are flow down to the chamber region. Therefore, relatively large amount of agglomerated silicon nanoparticles were observed at the reaction tube, while silicon nanoparticles are uniformly dispersed on the MWCNT column in the Si-MWCNT nanocomposite collected at the chamber. Fig. 7 shows the first discharge curve of raw materials and synthesized Si-MWCNT nanocomposites at the fixed current rate of 100 mA/g. In Fig. 7(a), the silicon electrode shows the highest discharge capacity of 4000 mAh/g, which is similar with the theoretical capacity of 4200 mAh/g. Such result means that Si atom was formed by being removed from Li22Si5 alloy. A Si atom is delithiated during the discharge, and then 4.4 Li atoms are formed leading to high capacity. First discharge capacity, however, is rapidly decreased as voltage is increased to 0.1 V, and then capacity is slowly decreased between in the range of 0.1 V to 0.6 V. As a result, it is confirmed that the silicon electrode shows instability leading to rapid capacity loss. On the other hand, the MWCNT electrode in Fig. 7(b) shows the first discharge capacity of 310 mAh/g because a Li atom is formed by the delithiating of 6 carbon atoms from LiC6 alloy. Therefore, the MWCNT electrode shows low capacity than the silicon
18
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
Fig. 7. The first discharge curve of at a current rate of 100 mA/g with different anode materials of (a) pristine silicon, (b) pristine MWCNT and synthesized Si-MWCNT nanocomposites collected from (c) the reaction tube and (d) the chamber.
electrode. In the MWCNT electrode, first discharge capacity is slowly decreased with increasing the voltage to 0.6 V, and the capacity is rapidly decreased in a high voltage over 0.6 V. In comparison to the silicon electrode, the MWCNT electrode shows stability during discharge, though the electrode capacity of MWCNT is noticeably lower than that of silicon. The first discharge curves of synthesized Si-MWCNT nanocomposites are presented in Fig. 7(c) and (d). The first discharge capacity of Si-MWCNT nanocomposite collected from the reaction tube is 200 mAh/g. The first discharge capacity of Si-MWCNT from the reaction tube is lower capacity than that of pristine MWCNT, and it is sharply decreased with increasing the voltage. Such results are attributed an irregular dispersion and agglomeration of silicon nanoparticles. Therefore, lithium ions and charges in the anode are difficult to move to cathode leading to poor capacity. On the other hand, the first discharge capacity of Si-MWCNT nanocomposite collected from the chamber exhibits 1447 mAh/g. This measured capacity is lower than that of pristine silicon but higher than that of pristine MWCNT. The discharge capacity of nanocomposite collected from the chamber was higher than that from the reaction tube. The volume expansion of silicon nanoparticles in the discharge reaction is suppressed when they are dispersed uniformly on MWCNT, which acts as the supporter. Since a relatively uniform dispersion of silicon nanoparticles on MWCNT were observed in Si-MWCNT nanocomposite collected from the chamber, the discharge capacity of nanocomposite collected from the chamber is higher than that from the reaction tube. The capacity of the Si-MWCNT electrode fabricated products from the chamber is slowly decreased with increasing voltage up to 0.7 V, then the capacity is sharply decreased in the voltage range of 0.7–2.3 V. Such results of the Si-MWCNT electrode fabricated using product from the chamber are similar with those of the MWCNT electrode, which shows a high stability. Therefore, it is expected that the electrode fabricated using Si-MWCNT nanocomposite collected from the chamber can prevent the volume expansion of silicon nanoparticles by MWCNT. In the voltage of 2.3–2.5 V, it is observed that a distinguishable plateau in Fig. 7(d) because of the formation of solid electrolyte interphase (SEI), which is formed by the decomposition of organic electrolyte at the
electrode surface during first cycle generally. The SEI layer acts as an electrically insulating to prevent further decomposition of the electrolyte leading to suppress self-discharge [16,17]. Fig. 8 exhibits curves of first charge capacity with different electrodes of pristine silicon, pristine MWCNT, and Si-MWCNT nanocomposites. Although the range of charge is set up between 0.01 and 1.5 V, the initial voltage is shown at least 0.3 V in all electrodes. In addition, poor first charge capacities of silicon, MWCNT, and Si-MWCNT nanocomposites collected from reaction tube and chamber of 7.7, 7.9, 7.0, and 8.3 mAh/g are measured. Those values mean that the resistance of electrodes is pretty high because of electrode thickness and impurities. MWCNT used in the present work has a high bulk density of 0.10 g/cm3 leading to the loss of power density, poor cycling performance, and instability [18]. When electrode thickness is increased, charge transfer
Fig. 8. First charge curves with different anode materials of micro-sized silicon, MWCNT, and synthesized Si-MWCNT nanocomposites collected from the reaction tube and the chamber.
Y.-S. Na et al. / Thin Solid Films 587 (2015) 14–19
resistance is increased with the decrease of lithium diffusion [19–21]. Therefore, electrodes exhibit high initial voltage and poor charge capacity. Si-MWCNT nanocomposite collected from the chamber is the highest charge capacity among all electrodes. These results show that Si-MWCNT nanocomposite, which is silicon nanoparticles attached to MWCNT column, was well synthesized for anode material with higher capacity than the MWCNT electrode and more stable than the silicon electrode. Fig. 9 shows FE-SEM images of Si-MWCNT nanocomposite electrodes after discharge–charge cycle. In Fig. 9(a), Si-MWCNT nanocomposite collected from reaction tube, the morphology is similar with Fig. 6(a), which shows agglomerated silicon nanoparticles around MWCNT. On the other hand, Fig. 9(b) for Si-MWCNT electrode fabricated by products from the chamber shows modified structure from Fig. 6(b). Silicon nanoparticles and MWCNT are slightly agglomerated after a charge–discharge cycle because organic solution is mixed with samples during cell assembly. Despite cycling, however, it is observed that silicon nanoparticles maintained their spherical shape without volume expansion, and they are well dispersed around MWCNT column in the case of Si-MWCNT collected from the chamber.
4. Conclusions Si-MWCNT nanocomposite was synthesized by thermal plasma. The nanocomposite structure was silicon nanoparticles attached to MWCNT column, and well-distributed silicon nanoparticle was found in the chamber rather than the reaction tube. It is expected that silicon vapors are saturated and nanoparticles are produced in the reaction tube
Fig. 9. FE-SEM images of electrodes fabricated using Si-MWCNT nanocomposites collected from (a) the reaction tube and (b) the chamber after a charge–discharge cycle.
19
region. There, some of silicon nanoparticles are deposited on the surface of the reaction tube. In addition, silicon vapors could be condensed on MWCNT column directly because MWCNT is injected into the peripheral region of plasma jet. Therefore, well-synthesized Si-MWCNT nanocomposites flow to the chamber. The electrochemical test of raw materials and Si-MWCNT products collected from the reaction tube and the chamber was carried out. From the results, Si-MWCNT nanocomposite collected from the chamber shows relatively stable charge–discharge characteristics with an enhanced capacity compared with pristine MWCNT. Acknowledgments This work was supported by the World Class 300 Project (10043264, Development of the electrode materials for high efficiency (21%) and low-cost c-Si solar cell) funded by the Ministry of Trade, Industry and Energy (MOTIE, Korea). References [1] M.M. Thackeray, C. Wolverton, E.D. Isaacs, Electrical energy storage for transportation—approaching the limits of, and going beyond, lithium-ion batteries, Energy Environ. Sci. 5 (2012) 7851. [2] Y.P. Wu, C. Wan, C. Jiang, S.B. Fang, Lithium ion secondary batteries, Chemical Industry Press, Beijing, 2002. (references therein). [3] T. Kojima, T. Ishizu, T. Horiba, M. Yoshikawa, Development of lithium-ion battery for fuel cell hybrid electric vehicle application, J. Power Sources 189 (2009) 859. [4] J.R. Szczech, S. Jin, Nanostructured silicon for high capacity lithium battery anodes, Energy Environ. Sci. 4 (2011) 56. [5] L.Y. Beaulieu, K.W. Eberman, R.L. Turner, L.J. Krause, J.R. Dahn, Colossal reversible volume changes in lithium alloys, Electrochem. Solid-State Lett. 4 (2001) A137. [6] C. de las Casas, W. Li, A review of application of carbon nanotubes for lithium ion battery anode material, J. Power Sources 208 (2012) 74. [7] Y. NuLi, J. Yang, M. Jiang, Synthesis and characterization of Sb/CNT and Bi/CNT composites as anode materials for lithium-ion batteries, Mater. Lett. 62 (2008) 2092. [8] J.P. Salvetat, J.M. Bonard, N.H. Thomson, A.J. Kulik, L. Forró, W. Benoit, L. Zuppiroli, Mechanical properties of carbon nanotubes, Appl. Phys. 69 (1999) 255. [9] C.H. Mi, G.S. Cao, X.B. Zhao, A non-GIC mechanism of lithium storage in chemical etched MWNTs, J. Electroanal. Chem. 562 (2004) 217. [10] R.A. Afrea, T. Sogaa, T. Jimboa, M. Kumarb, Y. Andob, M. Sharon, Growth of vertically aligned carbon nanotubes on silicon and quartz substrate by spray pyrolysis of a natural precursor: Turpentine oil, Chem. Phys. Lett. 414 (2005) 6. [11] S.C. Wang, F. Yang, M. Silva, A. Zarow, Y. Wang, Z. Iqbal, Membrane-less and mediator-free enzymatic biofuel cell using carbon nanotube/porous silicon electrodes, Electrochem. Commun. 11 (2009) 34. [12] J. Bae, Fabrication of carbon microcapsules containing silicon nanoparticles–carbon nanotubes nanocomposite by sol-gel method for anode in lithium ion battery, J. Solid State Chem. 184 (2011) 1749. [13] M. Shigeta, A.B. Murphy, Thermal plasmas for nanofabrication, J. Phys. D. Appl. Phys. 44 (2011) 174025. [14] J.-G. Li, Y. Sakka, Controlled thermal plasma processing of ceramic nanopowders, in: S. Somiya (Ed.), Handbook of Advanced Ceramics, Academic Press, Japan, 2013, p. 979. [15] X.H. Liu, L. Zhong, S. Huang, S.X. Mao, T. Zhu, J.Y. Huang, Size-dependent fracture of silicon nanoparticles during lithiation, ACS Nano 6 (2012) 1522. [16] C.K. Chan, R. Ruffo, S.S. Hong, Y. Cui, Surface chemistry and morphology of the solid electrolyte interphase on silicon nanowire lithium-ion battery anodes, J. Power Sources 189 (2009) 1132. [17] K. Zaghib, AFM Study on SEI growth on graphite negative electrode at elevated temperatures, in: K. Striebel, K. Zaghib, D. Guyomard (Eds.),Anodes, U.S.A., 11–19, 2003, Lithium and Lithium-ion Batteries: Proceedings of the International Symposium 28, 2004. [18] H. Zheng, J. Li, X. Song, G. Liu, V.S. Battaglia, A comprehensive understanding of electrode thickness effects on the electrochemical performances of Li-ion battery cathodes, Electrochim. Acta 71 (2012) 258. [19] U. Tocoglu, O. Cevher, M.O. Guler, H. Akbulut, Core–shell tin-multi walled carbon nanotube composite anodes for lithium ion batteries, Int. J. Hydrogen Energy (2014). http://dx.doi.org/10.1016/j.ijhydene.2014.05.053. [20] Y. Ito, M. Kawakubo, M. Ueno, H. Okuma, Q. Si, T. Kobayashi, K. Hanai, N. Imanishi, A. Hirano, M.B. Phillipps, Y. Takeda, O. Yamamoto, Carbon anodes for solid polymer electrolyte lithium-ion batteries, J. Power Sources 214 (2012) 84. [21] Y. Hu, X. Li, D. Geng, M. Cai, R. Li, X. Sun, Influence of paper thickness on the electrochemical performances of graphene papers as an anode for lithium ion batteries, Electrochim. Acta 91 (2013) 227–233.