Extraction Properties of Nanocarbon Materials - The

Sep 29, 2009 - ... the increase of the length of lithium reversible insertion/extraction route. ... Handan Yildirim , Alper Kinaci , Zhi-Jian Zhao , M...
1 downloads 0 Views 4MB Size
J. Phys. Chem. C 2009, 113, 18431–18435

18431

Lithium Insertion/Extraction Properties of Nanocarbon Materials Jiaxin Li, Chuxin Wu, and Lunhui Guan* State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, YangQiao West Road 155 #, Fuzhou, Fujian 350002, People’s Republic of China ReceiVed: July 1, 2009; ReVised Manuscript ReceiVed: September 9, 2009

Five kinds of nanocarbon materials (NCMs)ssingle-walled carbon nanotubes (SWNTs), C60@SWNTs (C60peapod), multiwalled carbon nanotubes (MWNTs), graphite nanoflakes, and graphite nanoparticles containing some MWNTsswere systematically investigated as anode materials for Li ion batteries via scanning electron microscopy, transmission electron microscopy, Raman spectroscopy, and a variety of electrochemical testing techniques. Galvanostatic charge-discharge indicated that lithium storage capacity, Coulombic efficiency, and cyclability strongly depended on both the subtle structures and the subsequent treatments of NCMs. It was proved by cyclic voltammetry that lithium oxidation potentials increase with the increase of the length of lithium reversible insertion/extraction route. Raman spectroscopy revealed a mechanism of lithium insertion/ extraction in SWNT and C60-peapod electrodes, suggesting that lithium can enter the inner space of the tubes, the trigonal interstitial channels of the bundles, and the intercalation site between adjacent tubes, and that the dominant reversible sites for lithium storage are the trigonal interstitial channels. 1. Introduction Lithium second batteries (LIBs) are key components in the use of clean energy storage, hybrid electric vehicles, portable electronic devices, etc. Graphite, which has a theoretical maximum capacity of 372 mA h g-1 associated with its maximum LiC6, is commercially used as an anode material for LIBs. Recently, graphite-related materials, especially nanocarbon materials (NCMs), have received much attention as lithium insertion host materials in LIBs basically for their chemical and thermal tolerant, unique structure and electrical conductivity.1-3 For instance, multiwalled carbon nanotubes (MWNTs) have an interplanar spacing of 0.34 nm for the insertion and extraction of lithium. MWNTs prepared by several different approaches have exhibited reversible capacities of 80-540 mA h g-1 and very high irreversible capacities.4-7 Single-walled carbon nanotubes (SWNTs) form nanorope bundles with close-packed 2D triangular lattices. The rope crystallites offer an all-carbon host lattice for lithium storage. Li capacities in the purified SWNTs were reported to be 170-600 mA h g-1, which can be further improved up to 1000 mA h g-1 by either ball milling or chemical etching.8-12 Especially, C60@SWNT (C60-peapod) was proven to be a new promising anode material of the LIBs due to its sophisticated tube internal structure.13 Recently, theoretic calculations and experiments were extensively carried out for clarifying the lithium storage mechanism. Theoretical investigations mainly focused on the intercalation of Li with NCMs and the energy barriers for entry and diffusion of Li inside the NCMs.14-17 First-principles study predicted that Li ions can intercalate both into the channels between the nanotubes and into the interstitial space of the nanotubes. In the ideal case, this gives an enhanced anode stoichiometry of LiC2, which is higher than those of conventional graphite anodes.17,18 Furthermore, Raman spectroscopy was used to investigate those doping experiments such as those of alkali metals including lithium-doped SWNTs, C60-peapod, and * To whom correspondence should be addressed. Telephone: 86-59183792835. Fax: 86-591-83792835. E-mail: [email protected].

MWNTs.19-22 However, some relevant issues are not clearly understood, such as the following: (1) Does the diffusion of Li ions inside interplanar spacing depend on the morphology of the NCMs, especially on the length of lithium reversible route? (2) Where is the main reversible area for lithium insertion/ extraction in SWNTs and C60-peapod? (3) Can the intercalation of C60 modify the capacity and cyclability of SWNT electrode? In this work, we systematically chose five kinds of NCMs with varied structures and morphology: single-walled carbon nanotubes (SWNTs), C60@SWNTs (C60-peapod), multiwalled carbon nanotubes (MWNTs), graphite nanoflakes (denoted as G-1), and graphite nanoparticles containing some MWNTs (denoted as G-2) as Li-insertion materials and consequently aim to answer the questions mentioned above. 2. Experimental Section 2.1. Preparation of NCMs. G-1, G-2, SWNTs, and C60peapod were all produced by an arc-discharged method based on our previous studies.23 Both G-1 and G-2 were synthesized as the follows. The arc discharge was created at currents of 80 and 60 A with a constant distance maintained between the electrodes in He atmosphere at pressures of 0.72 and 0.30 atm, respectively. The details of preparation for SWNTs and C60peapod were also presented in an earlier work.25 SWNTs with 99% purity and C60-peapod prepared by a gas phase method with ∼85% filling ratio were obtained. MWNTs were purchased from Shenzhen Nanotech Port (Shenzhen, China) and used as received. The structures and morphology of these five kinds of NCMs were observed by SEM (JSM6700F), TEM (JEOL 2010), and Raman spectroscopy (Renishaw, excited at 514.5 nm), respectively. 2.2. Electrochemical Test of Li /NCMs Cells. The electrochemical behaviors were measured via CR2025 coin-type test cells assembled in a dry argon-filled glovebox. The working electrode was fabricated by drying slurry (85 wt % active material, 5 wt % acetylene black, and 10 wt % poly(vinylidene fluoride) (PVDF)) dissolved in 1-methyl-2-pyrrolidinone (NMP) on a copper foil disk of 15 mm diameter at 80 °C under vacuum

10.1021/jp9061658 CCC: $40.75  2009 American Chemical Society Published on Web 09/29/2009

18432

J. Phys. Chem. C, Vol. 113, No. 42, 2009

Li et al.

Figure 1. SEM images of different NCMs: (a) G-1, (b) G-2, (c) MWNTs, (d) C60-peapod, and (e) SWNTs.

Figure 3. (a-e) Cyclic voltammetry curves and (f) the first charge cyclic voltammetry curves of five NCMs.

Figure 2. (a-e) TEM images and (f) models of different NCMs: (a) G-1, (b) G-2, (c) MWNTs, (d) C60-peapod, and (e) SWNTs. The insets show the HR-TEM images of part patterns in (a)-(e), respectively.

for 24 h. Typically about 2.0 mg of NCMs was used as the negative electrode. The test cell consisted of cathode and lithium sheet anode which were separated by a Celgard 2300 membrane and electrolyte of 1 M LiPF6 in EC:EMC:DMC (1:1:1 in volume). The cells were cycled by LAND2001A between 0.05 and 3.00 V (versus Li/Li+) at a current density of 50 mA g-1 at room temperature. Cyclic voltammetry (CV) tests system were performed on a CHI660C Electrochemical Workstation with a scan rate of 0.4 mV s-1. 3. Results and Discussion 3.1. Morphology and Structure of NCMs. Figure 1 shows typical SEM images of the NCMs. Parts a-e correspond to samples G-1, G-2, MWNTs, C60-peapod, and SWNTs, respectively. For the sample G-1, there are mainly thin graphene nanoflakes with the width ranging from 100 nm to several micrometers, and some sporadic graphene platelets, as shown in Figure 1a. Figure 1b shows that the MWNTs containing about 60% closed elliptical carbon nanocapsules are very straight with lengths of 100 nm-4 µm. Diversely, the commercial MWNTs have a uniform and porous structure in which the individual MWNTs are obviously visible (Figure 1c). Figures 1d and e show SEM images of sample C60-peapod and sample SWNTs produced by our recent method, which are of ultrahigh purity.23 There are no differences between C60-peapod and empty SWNTs from SEM observation. Figures 2a-e show the TEM patterns of the NCMs. The inserted figures show the HRTEM images of part patterns in Figures 2a-e, respectively. Also, the schematic representations for both MWNTs (Figure 2c) and SWNT bundles (Figures 2d and e) are shown in Figure 2f. The HR-TEM image in Figure

2a clearly shows that the thickness of the graphene nanoflakes is approximately 1.5 nm, indicating that it is composed of approximately 4-7 stacked individual monatomic graphene layers. The HR-TEM image for G-2 is shown in the inset of Figure 2b. A large number of lattice fringes of regular spacing can be clearly observed, showing highly graphitized structures. Further investigation reveals that the caps of some MWNTs (2-5 nm in diameter) are partially opened, and there is a small quantity of structural defects along the sidewalls. Figure 2c reveals that the outer diameter of commercial MWNTs ranges from 10 to 30 nm and the inner diameter ranges from 5 to 10 nm. The inset of Figure 2c shows a typical HR-TEM pattern of MWNTs: a fiber with a stacking morphology of truncated conical graphene, large amounts of open edges on the outer surface, and central channels with symmetrical nodes (see “model 1” in Figure 2f for details). Such a structure makes the interior cores and interlayer space accessible for Li ions insertion/extraction. Figures 2d and e show TEM images of C60peapod and SWNTs, respectively. The inset of Figure 2d indicates a typical HR-TEM image of the C60-encapsulated nanotubes. The two parallel dark lines correspond to the SWNT walls. Round objects are attributed to the individual C60 molecules which are aligned linearly along the tube axis. The filling rate of C60 peas is roughly estimated as high as ∼85% by HR-TEM observation. Webs of bundles for either SWNTs or C60-peapod with outer diameters of 10-40 nm are shown in Figures 2d and e. Furthermore, their hexagonal schematic illustration is pictured as “model 2” shown in Figure 2f. The typical diameter of a SWNT is ∼1.4 nm, and the hole composed by packed SWNTs is ∼0.6 nm. Hereby, Li ions could insert and extract there reversibly because of its smaller size (about 0.076 nm in diameter). 3.2. Cyclic and Li Ions Storage Characteristics of NCM Electrodes. Cyclic voltammograms (CVs) for five NCMs are given in Figures 3a-e. The CVs for all NCMs clearly show peaks that correspond to the insertion and extraction of Li ions. The first discharge cycles in Figure 3 for these five NCMs are characterized by peaks ranging from 0.3 to 0.7 V, corresponding to the voltage plateaus in the discharge curves shown in Figure 4. The plateaus presumably arised from the irreversible insertion of Li ions into the NCMs. However, it can be seen from Figures 3a and b that no distinct peaks are detected in both G-1 and G-2, indicating a significantly lower amount of Li ions for the formation of the solid electrolyte interface (SEI) in these two NCMs due to their densely packed structures and few surface

Properties of Nanocarbon Materials

Figure 4. First discharge-charge curves of Li insertion/extraction into/ from the five kinds of NCMs.

functional groups.24 Furthermore, the CVs repeatedly emerged at the same potential for all five NCMs after the second discharge cycle, showing that the NCM electrodes exhibit excellent cyclability. In the first charge CVs (Figure 3f), peaks around 0.60 V are observed for samples G-1 and G-2, and that of 0.30, 0.40, 1.25, and 2.25 V for the latter three NCMs, i.e., MWNTs, C60-peapod, and SWNTs. On one hand, the peaks around 0.30, 0.40, and 0.60 V are respectively attributed to the extraction of Li ions from between rolled graphitic layers and trigonal interstitial channels and between the graphitic layers. Meanwhile, the peak at 0.30 V for MWNTs is shifted to 0.60 V for samples G-1 and G-2, suggesting that the extraction of Li ions from the inner sites is greatly hindered due to the strong tendency of graphite to accumulate Li ions.12,24 Based on the SEM and TEM observation, the length of Li ions reversible route was estimated to be larger than 250 nm for samples G-1 and G-2, ∼100 nm for C60-peapod and SWNTs, and ∼50 nm for MWNTs, respectively. It also appeared that the insertion of Li ions into the inner sites of the MWNTs was easier than that of the other four NCMs because of the shortest length of the Li ions reversible route, and this resulted in an enhanced reversible capacity (Crev) (see Figure 5b for details). On the other hand, the peaks around 1.25 and 2.25 V for the latter three NCMs were attributed to the extraction of Li ions from the inner cores of the CNTs and bonding to the surface functional groups in the CNTs, respectively.11,18 Therefore, the amount of extracted Li ions from those two types of reversible sites in the latter three NCMs was larger than that in both G-1 and G-2, which did not obviously exhibit these two peaks (around 1.25 and 2.25 V), resulting in a larger Crev for the latter three NCMs. Figure 4 shows the first charge-discharge voltage profiles of the five NCM electrodes. The voltage dropped rapidly and formed a plateau during the first discharge process, which was attributed to the decomposition of the electrolyte and the formation of the SEI film on each NCM surface. When the

J. Phys. Chem. C, Vol. 113, No. 42, 2009 18433

Figure 5. Cycle performances and Coulombic efficiency of five NCM electrodes as a function of cycle number.

voltage gradually decreased, all the second plateaus at about 0.15 V appeared in the NCM electrodes. Furthermore, during the charge process, prominent voltage plateaus were exhibited at about 0.55 V for samples G-1 and G-2, 0.30 V for MWNTs, and 0.40 V for C60-peapod and SWNTs, which was consistent with the results of CVs (shown in Figure 3f). The Coulombic efficiency and the electrochemical cycle properties of these five NCM electrodes over 50 cycles at the current density of 50 mA g-1 were compared in Figure 5. They all exhibited excellent Coulombic efficiency with larger than 95% retention after 50 cycles. Compared with the latter three electrodes, the values of Coulombic efficiency for the G-1 and G-2 electrodes were larger. Meanwhile, the Crev values for G-1 and G-2 remain persistently about 61 and 74 mA h g-1 after several unstable cycles as shown in Figure 5a. It was ascribed to the formation of the lower active functional groups, which cannot prompt the decomposition of the electrolyte and the formation of the lower SEI film on the surface of these two samples. Such a result was consistent with that of previous work on raw NCMs synthesized by the arc-discharge method.24 Inversely, the reversible capacities for the latter three NCM electrodes corresponding to MWNTs, C60-peapod, and SWNTs were larger and found to be 214, 203, and 202 mA h g-1 after 50 cycles (Figures 5b and c). It should be noted that C60-peapod and SWNTs presented obviously dissimilar charge capacities: SWNTs exhibited the initial capacity of 867 mA h g-1 and the irreversible capacity of 534 mA h g-1, and the C60-peapod exhibits those of 669 and 403 mA h g-1. The reason for this should be that the volume of inner space for SWNTs was larger than that of C60-peapod, and also the conductance for SWNTs was larger than that of C60-peapod. In addition, the conductance of C60-peapod enhanced after charge-discharge circulations, which can also improve its cycle performance.13 Furthermore, Figure 5c shows that the charge capacities for these two samples decreased tardily in the initial 15 cycles and then turned to the same values of about 205 mA h g-1. This suggested that the

18434

J. Phys. Chem. C, Vol. 113, No. 42, 2009

Li et al.

TABLE 1: Comparison of the Electrochemical Properties and the Structural Parameters of NCMs with the Different Types of NCMs Including Those of CNTs Reported in Other Literaturea samples

reversible route (nm)

extraction potential (V)

initial capacity (mA h/g)

capcity (mA h/g)

coulombic efficiency (%)

MWNTs C60-peapod SWNTs G-1 G-2 CNTs in ref 24 (0.2 mA/cm2, after 30 cycles)

∼50 ∼100 ∼100 ∼250 ∼250 s

∼0.55 ∼0.40 ∼0.40 ∼0.30 ∼0.30 0.30

862 669 867 164 173 180

214 203 202 74 61 50

96 97 97 99 99 almost 100

a

Each sample was cycled 50 times at 50 mA/g in this work.

energy side with the decrease of the Li content, and then upshifts faintly.19 Moreover, the shift tendency for C60-peapod was similar to that of SWNTs (Figure 6c). However, it should be noted that the G-band at 0.15 V (corresponding to the maximal Li storage) for SWNTs was downshifted from 1593.2 cm-1 to 1592.3 cm-1, while that of C60-peapod at 0.11 V upshifted to 1594.7 cm-1. These shifts were probably derived from the competition between the upshift of the G-band caused by the stress induced by the insertion of the Li ions in interstitials of tubes and the downshift caused by the effect of the charge transfer from the Li ions inserted inside the tubes.20 Thus, compared with the C60-peapod, there was larger capacity for the insertion of the Li ions in the inner channels of SWNTs. As the charge progress was ongoing, both samples showed a similar downshift tendency, indicating that most inserted Li ions can reversibly extract from trigonal interstitial channels of SWNTs as also shown in model 2 of the inset of Figure 2f. And then, the latter faint upshift for both samples may be caused by the solvated Li ions from the SEI formed in the inner channels, which lead to the extraction of Li ions. This standpoint was also supported by Kawasaki et al.13 Figure 6. (a) Raman spectra of the SWNT-slurry, SWNTs, C60-peapod, and C60-peapod-slurry, (b, c) Raman spectra of SWNT and C60-peapod electrodes during the charge progresses.

cycle performance of these two electrodes did not depend on the inner structure of the tubes, and their dominant reversible sites for Li ions storage should be the sites between adjacent tubes and the trigonal interstitial channels. It can be also explained by a Raman result discussed in section 3.3. For comparison, the electrochemical properties and the structural parameters of NCMs with the different types of NCMs including those of CNTs reported in other literature are summarized in Table 1. 3.3. Mechanism of Li Storage in SWNT and C60-peapod Electrodes. One way to increase our understanding about the structures of SWNTs and C60-peapod, the Li storage behavior, the charge-transfer effects between the graphitic layers and Li ions is by using Raman detection. Figure 6a presents the Raman spectra of SWNT-slurry, SWNTs, C60-peapod, and C60-peapodslurry. The Raman spectra of C60-peapod display the superposition of lines of the SWNTs and the fullerene peas, while the SWNTs dominate the spectrum. The arrow in the inset of Figure 6a indicated intense C60 lines, which can be also observed in bare C60. Moreover, peak positions for G-band downshift after encapsulation of C60. The changes in the Raman spectra of the Li storage in samples SWNTs and C60-peapod as a function of potentials ranging from about 0.10 to 2.50 V corresponding to the Li insertion contents during the charge processes after more than 50 cycles are shown in Figures 6b and c. From Figure 6b, we noted that the peak of the G-band for the Li-SWNT initially shifts toward the lower

4. Conclusions In this paper we systematically investigated the electrochemical performance and the Li storage mechanism of five kinds of NCMs for Li insertion/extraction. The results revealed the following: (1) Lithium oxidation potentials increased with the increase of the length of lithium reversible insertion/extraction route; in other words, the diffusion of Li ions inside interplanar spacing depended on the morphology of the NCMs. (2) For both SWNT and C60-peapod electrodes, lithium can enter the inner space of the tubes, the trigonal interstitial channels of the bundles, and the intercalation sites between adjacent tubes, and the dominant reversible sites for lithium storage were the trigonal interstitial channels. (3) The intercalation of C60 can evidently improve the capacity and cyclability of SWNT electrode. Meanwhile, the electrochemical performance including chargedischarge capacities, Coulombic efficiency, and cyclability strongly depended on both the structures and the subsequent treatments for each NCM electrode. Acknowledgment. We acknowledge the financial support provided by Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences (grant no. SZD09003) and the National Key Project on Basic Research (grant no. 2009CB939801) and the Scientific Research Foundation for the Returned Overseas Chinese Scholars, Ministry of Human Resources and Social Security of China. The authors thank Mr. YZ. Li, Mrs. LH. Zhou, and Mr. F. Bao for helping with the Raman spectroscopy, SEM, and TEM, respectively.

Properties of Nanocarbon Materials References and Notes (1) Kim, M. G.; Cho, J. AdV. Funct. Mater. 2009, 19, 1. (2) Shukla, A. K.; Kumar, T. P. Curr. Sci. (India) 2008, 94, 314. (3) Liu, H. K.; Wang, G. X.; Guo, Z. P.; Wang, J. Z.; Konstantinov, K. J. Nanosci. Nanotechnol. 2006, 6, 1. (4) Yang, S. B.; Huo, J. P.; Song, H. H.; Chen, X. H. Electrochim. Acta 2008, 53, 2238. (5) Wang, X.; Wang, J. N.; Chang, H.; Zhang, Y. F. AdV. Funct. Mater. 2007, 17, 3613. (6) Eom, J. Y.; Kwon, H. S.; Liu, J.; Zhou, O. Carbon 2004, 42, 2589. (7) Masarpu, C.; Subramanian, V.; Zhu, H.; Wei, B. AdV. Funct. Mater. 2009, 19, 1. (8) Shimoda, H.; Gao, B.; Tang, X. P.; Kleinhammes, A.; Fleming, L.; Wu, Y.; Zhou, O. Phys. ReV. Lett. 2002, 88 (1), 015502. (9) Ng, S. H.; Wang, J.; Guo, Z. P.; Chen, J.; Wang, G.; Liu, H. K. Electrochim. Acta. 2005, 51, 23. (10) Gao, B.; Kleinhammes, A. X.; Tang, P.; Bower, C.; Fleming, L.; Wu, Y.; Zhou, O. Chem. Phys. Lett. 1999, 307, 153. (11) Gao, B.; Bower, C.; Lorentzen, J. D.; Fleming, L.; Kleinhammes, A.; Tang, X. P.; McNei, L. E.; Wu, Y.; Zhou, O. Chem. Phys. Lett. 2000, 327, 69. (12) Eom, J. Y.; Kwon, H. S. J. Mater. Res. 2008, 23, 2458. (13) Kawasaki, S.; Iwai, Y.; Hirose, M. Mater. Res. Bull. 2009, 41, 415. (14) Khantha, M.; Cordero, N. A.; Alonso, J. A.; Cawkwell, M.; Girifalco, L. A. Phys. ReV. B 2008, 78, 115430.

J. Phys. Chem. C, Vol. 113, No. 42, 2009 18435 (15) Girifalco, L. A.; Hodak, M. L. Phys. ReV. B 2000, 62, 13104. (16) Martinez, J. I.; Cabria, I.; Lopez, M. J.; Alonso, J. A. J. Phys. Chem. C 2009, 113 (3), 939. (17) Zhao, J. J.; Buldum, A.; Han, J.; Lu, J. P. Phys. ReV. Lett. 2000, 85 (8), 1706. (18) Meunier, V.; Kephart, J.; Roland, C.; Bernholc, J. Phys. ReV. Lett. 2002, 88 (7), 075506. (19) Kim, Y. A.; Kojima, M.; Muramatsu, H.; Umemoto, S.; Watanabe, T.; Yoshida, K.; Sato, K.; Ikeda, T.; Hayashi, T.; Endo, M.; Terrones, M.; Dresselhaus, M. S. Small 2006, 2, 667. (20) Iwaia, Y.; Hirosea, M.; Kanoa, R.; Kawasakia, S.; Hattorib, Y.; Takahashic, K. J. Phys. Chem. Solids 2008, 69, 1199. (21) Kavan, L.; Kalbac, M.; Zukalova, M.; Dunsch, L. Carbon 2006, 44, 99. (22) Kim, Y. A.; Kojima, M.; Muramatsu, H.; Shimamoto, D.; Hayashi, T.; Endo, M.; Terrones, M.; Dresselhaus, M. J. Raman Spectrosc. 2008, 39, 1183. (23) Wu, C.; Li, J.; Dong, G.; Guan, L. J. Phys. Chem. C 2009, 113 (9), 3612. (24) Yang, S.; Song, H.; Chen, X.; Okotrub, A. V.; Bulusheva, L. G. Electrochim. Acta 2007, 52, 5286. (25) Guan, L.; Suenaga, K.; Okazaki, T.; Shi, Z.; Gu, Z.; Iijima, S. J. Am. Chem. Soc. 2007, 129, 8954.

JP9061658