- Email: [email protected]
Three-dimensional carbon nanotubes for high capacity lithium-ion batteries
Journal of Power Sources 299 (2015) 465e471
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Three-dimensional carbon nanotubes for high capacity lithium-ion batteries Chiwon Kang a, Mumukshu Patel a, Baskaran Rangasamy a, 1, Kyu-Nam Jung c, Changlei Xia b, Sheldon Shi b, Wonbong Choi a, b, * a b c
Department of Materials Science and Engineering, University of North Texas, North Texas Discovery Park 3940 North Elm St., Denton, TX 76207, USA Department of Mechanical and Energy Engineering, University of North Texas, North Texas Discovery Park, 3940 North Elm St., Denton, TX 76207, USA Energy Efficiency and Materials Research Division, Korea Institute of Energy Research, 152 Gajeong-ro, Yuseong-gu, Daejeon 305-343, Republic of Korea
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
We show a novel structure of multistacked 3D CNTs for a higher loading of CNTs. The bulk density of multi-stacked 3D CNTs is twice as high as that of graphites. The multi-stacked 3D CNTs yield a stable and high reversible volumetric capacity.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 June 2015 Received in revised form 29 August 2015 Accepted 31 August 2015 Available online xxx
Carbon nanotubes (CNTs) have been considered as a potential anode material for next generation Lithium-ion batteries (LIBs) due to their high conductivity, flexibility, surface area, and lithium-ion insertion ability. However, the low mass loading and bulk density of carbon nanomaterials hinder their use in large-scale energy storage because their high specific capacity may not scale up linearly with the thickness of the electrode. To address this issue, a novel three-dimensional (3D) architecture is rationally designed by stacking layers of free-standing CNTs with the increased areal density to 34.9 mg cm2, which is around three-times higher than that of the state-of-the-art graphitic anodes. Furthermore, a thermal compression process renders the bulk density of the multi-stacked 3D CNTs to be increased by 1.85 g cm3, which yields an excellent volumetric capacity of 465 mAh cm3 at 0.5C. Our proposed strategy involving the stacking of 3D CNT based layers and post-thermal compression provides a powerful platform for the utilization of carbon nanomaterials in the advanced LIB technology. © 2015 Elsevier B.V. All rights reserved.
Keywords: 3-Dimensional free-standing carbon nanotubes Lithium ion batteries Volumetric capacity Areal capacity Bulk density Multi-layered anode stack
1. Introduction * Corresponding author. Department of Materials Science and Engineering, University of North Texas, North Texas Discovery Park 3940 North Elm St., Denton, TX 76207, USA. E-mail address: [email protected] (W. Choi). 1 Present address: Department of Physics, School of Basic and Applied Sciences, Central University of Tamilnadu, Thiruvarur, Tamilnadu, India. http://dx.doi.org/10.1016/j.jpowsour.2015.08.103 0378-7753/© 2015 Elsevier B.V. All rights reserved.
The lithium-ion battery (LIB) has been one of the most commonly used state-of-the-art energy storage systems since it was first commercialized in 1990. The commercial success of the LIB is mainly attributed to the unique features of high operating
466
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
potential (>3.0 V), high energy density (110e170 Wh kg1 and 230e400 Wh L1), appropriate cyclability, and cost effectiveness [1,2]. To further enhance the LIB performance, attempts have been made to investigate more efficient anode materials alternative to the conventional graphite anodes offering a theoretical specific capacity of 372 mAh g1. Among the potential carbon-based materials, the quest for carbon nanotubes (CNTs) has been particularly promising due to their excellent electrochemical and physicochemical properties [3]. The excellent electrochemical performances of CNTs for LIB are mainly associated with the CNTs' unique characteristics such as high surface area, short diffusion length for lithium-ions, and excellent electrical and mechanical properties [4e6]. Although CNTs have been proven as emerging anode material candidates to yield high specific capacity for LIBs, there are still legitimate limitations to be examined for the application of the CNTs to commercial products [7]. First, the previously reported CNTs with high specific capacity are, in many cases, dimensionally confined in thin film geometries that are not suitable for large-scale LIBs. Second, the specific capacity does not scale proportionally with the thickness of CNTs when the weight of electroactive CNTs is negligibly low relative to the total weight of an LIB cell. Third, CNTs have intrinsically low volumetric capacity that is a critical property for large-scale energy storage systems [8]. As a key property to enhance volumetric capacity, the bulk density (mass divided by total volume of solid matter) of most nanomaterial-based electrodes including CNTs is less than 1 g cm3, which is a mediocre density compared to that of commercial-grade graphite (0.35~0.9 g cm3) [9,10]. Due to the above-mentioned limitations, there is no clear confirmation that CNTs are indeed pertinent anode nanomaterials for high volumetric capacities to be used in largescale energy storage systems. To address these challenging issues of the CNT anode, we have introduced a 3D micro-channeled copper (Cu) that has a high surface area to accommodate the large loading amounts of CNTs [11]. In addition, the high porosity of 3D CNTs is suited to facilitate lithium-ion diffusion and electron charge transfer through the electrolyte and bulk electrode into the 3D CNTs as electroactive materials; as a result, the high porosity can enhance cyclic performance and current rate capability of the electrode [12,13]. In recent years, different approaches have been applied to enhance areal density (mass per unit area) and capacity (storage capacity per unit area) as well as bulk density and volumetric capacity. T. Sharifi et al. made the nitrogen-doped CNTs grown on carbon paper substrate by using a catalytic chemical vapor deposition (CVD) method, and the areal density and capacity of the CNTs/carbon paper electrode are 9.6 mg cm2 and 1.95 mAh cm2 at 0.2C, respectively [14], which are comparable to those of commercially available LIBs [15]. D. T. Welna et al. fabricated a freestanding film of vertically aligned multi-wall CNTs on a nickel
substrate using a combined method of catalytic thermal CVD and sputter deposition, and the CNT anode showed the bulk density and volumetric capacity of 0.51 g cm3 and 395 mAh cm3 at 57 mA g1 [16], respectively, which are comparable to those of commercial graphites [10,17,18]. In this study, we report that multi-stacked 3D CNTs demonstrate a stable and reversible volumetric capacity higher than that of the state-of-the-art graphitic anode used in an LIB. The structure of multi-stacked 3D CNTs provides one of the most pragmatic approaches towards the enhanced LIB performance of CNTs as promising anode nanomaterials for advanced large-scale LIBs. 2. Experimental 2.1. Fabrication of multi-stacked 3D CNT anode In this work, a multi-stacked 3D CNT-based anode was prepared through the following key steps: (1) 3D CNTs were directly grown on a Cu mesh (>99% purity) with 50 mm thickness and 65 mm hole size (TWP Inc.) as a substrate by using a thermal catalytic CVD method; (2) The Cu substrate was removed by ferric chloride (FeCl3) etching solution (Transene Company, Inc.) to obtain 3D freestanding CNT layers; and (3) The 3D free-standing CNT layers were stacked through the vacuum annealing and thermal compression methods to fabricate multi-stacked 3D CNTs with a high bulk density. Fig. 1 schematically represents the process for the formation of the multi-stacked 3D CNT-based anode. In Fig. 1(a), CNTs were directly grown on Cu mesh by a CVD method, where nickel (Ni) and titanium (Ti) were deposited onto Cu, using a radio frequency (RF)-direct current (DC) magnetron-sputtering system, which served as a catalyst for CNT growth and a barrier layer against carbon diffusion into Cu during CNT growth, respectively. The Ni/Ti deposited Cu mesh was placed in a CVD system, and the mixture gas of H2 and C2H4 with a volume ratio of 1:2 flowed into the furnace for 50 min at the growth temperature of 750 C. The weights of each sample before and after CNT growth were measured to calculate the net weight of CNT as an electroactive material for a LIB. After 4 h of Cu etching, Cu mesh was completely dissolved into the concentrated FeCl3 solution, and the solution was left with 3D free-standing CNT structures (see Fig. 1(b)). The freestanding structures were thoroughly washed with de-ionized water and dried in a vacuum oven (Thermo Scientific Lindberg) at 120 C for 4 h as presented in Fig. 1(c). To prepare a highly dense multi-stacked 3D CNT anode, we employed polyvinylidene fluoride (PVDF, SigmaeAldrich) as a binder dissolved into N-Methyl-2pyrrolidone (NMP) solvent. The solution is composed of 1.4 wt% of PVDF and 98.6 wt% of NMP (the weight ratio of CNTs to PVDF is 8:2). Two pieces of free-standing CNT layers were sandwiched within the solution and subsequently heated at 120 C under a
Fig. 1. Process flow for the fabrication of multi-stacked 3D CNT anode. (a) CVD grown 3D CNTs on Cu mesh. (b) Chemical etching process of the Cu mesh via the FeCl3 etching solution. (c) 3D free-standing CNT structure after the etching process. (d) The multi-stacked 3D CNT anode fabricated by pressing the layers of 3D CNTs with the aid of PVDF binder.
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
467
Fig. 2. Structural properties of multi-stacked 3D CNT anode. (a) FESEM plain-view image of randomly oriented and highly dense 3D CNTs directly grown on Cu mesh using a CVD method and schematic diagram of the 3D CNTs on Cu. (b) High resolution TEM (HRTEM) image of a single CNT comprising the turbostratic or low-ordered graphitization structure. (c) Raman spectra of a 3D CNT showing the intensity ratio of D-band to G-band peaks (ID/IG > 1). (d) Energy Dispersive X-ray Spectroscopy (EDS) spectra obtained from a spot on the 3D free-standing CNT structure shown in the SEM image of (a).
vacuum pressure of 760 mmHg to completely evaporate the NMP solvent. The processing step was repeated until the desired number of stacked layers was reached. Finally, the stacked layers were thermally compressed at 60 C and 5000 bar (Fig. 1(d)). The thickness and footprint areas of the as-prepared anode stack were measured by a digital caliper (VWR) to calculate bulk density for the measurement of volumetric capacity. For comparison, the control anode stack was also prepared at room temperature by using a precision press (MTI Inc.) with a pressure of 6 bar, without using the binder.
2.2. Structural characterization The as-grown CNTs were characterized by using a field emission scanning electron microscope (FESEM) (JEOL, JSM-7000F), energy dispersive X-ray spectrometer (EDS) (Thermo Electron Corporation, NORAN System SIX), Raman spectrometer (Arþ laser with l ¼ 514 nm, 33 mW power), and field emission transmission electron microscope (FETEM) (FEI, TECHNAI F20). The BrunauereEmmetteTeller (BET) specific surface area (SSA) and the total pore volume of the samples were measured by using a surface
Fig. 3. FESEM cross-sectional images of (a) the control multi-stacked 3D CNT anode without the treatment of binder and high pressure. (b) The denser structure of CNTs could be fabricated by applying the binder and high compressive pressure at 60 C.
468
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
area/pore size analyzer (3Flex 3500, Micromeritics Instrument Corp.) at 77 K. Before characterizing the pore volume, the samples were degassed at 100 C for 3 days at ~0.1 mbar by employing a degasser (VacPrep 061, Micromeritics Instrument Corp.) followed by in-situ degassing in the surface area/pore size analyzer. The total pore volume was measured using a Non-Local Density Functional Theory (NLDFT) model based on the amount of nitrogen adsorbed up to P/P0 ~ 0.99. 2.3. Electrochemical performance The electrochemical performance of the multi-stacked 3D CNTbased anode was measured in a coin cell (CR 2032, Wellcos Ltd.). Half-cell assembly was fabricated in an argon-filled glovebox (VAC atmosphere Ltd. and MBraun) under extremely low levels of humidity and oxygen (<0.5 ppm). The 3D-CNT anode was used as a working electrode and Li metal foil served as a counter and reference electrode. 1 mol dm3 solution of lithium hexafluorophosphate (LiPF6) salt in 1:1:1 (volume ratio) mixture solvent of ethylene carbonate (EC), dimethylene carbonate (DMC), and diethylene carbonate (DEC) (MTI Corp.) was used as an electrolyte. A typical polypropylene (PP) based membrane (Separator-2400, Wellcos Ltd.) was used as a separator. The cell was finally assembled using a CR 2032 coin-cell crimping tool (Hohsen Corp.). The charge (delithiation) and discharge (lithiation) cycling tests were performed in a multi-channel battery testing unit (MACCOR-series 4000) at room temperature in the voltage window of 0.01e3.0 V. 2.4. Electrochemical impedance spectroscopy (EIS) measurement The EIS measurement for the 3D-CNT anode after 100 cycles was conducted at a series of open-circuit voltages (OCVs) in a frequency range from 0.01 Hz to 106 Hz using a potentiostat (Reference 3000, Gamry Instrument). Each OCV determines the depth of discharge (DOD) obtained by the lithium insertion from the anode stack. The perturbation amplitude was set as ±5 mV and ten points per decade were collected. The impedance data of real and imaginary parts were plotted as a function of frequency in a complex plane diagram (Nyquist plot) and were fitted using a computer software (Echem Analyst™, Gamry Instrument).
469
grown on Cu, showing highly dense and randomly oriented CNTs. At a microscopic level, the CNT has a low-ordered graphitization structure, as confirmed by a high resolution transmission electron microscopy (HRTEM) (see Fig. 2(b)) [19]. As seen in Fig. 2(c), typical Raman spectra of a 3D CNT indicate two strong peaks at ~1350 cm1 and ~1580 cm1, which are ascribed to D-band (defect related) and G-band (E2g in-plane vibration mode). The intensity ratio (ID/IG) is higher than 1.0, implying that the CNTs are composed of low-ordered graphitization structures [20]. The XRD patterns also identify the low-ordered graphitization structures of CNTs based on the broad peak related to hexagonal carbon (002) appearing at 2q ¼ 25.5 (JCPDS card no. 26-1079) as presented in Figure S2. It is known that the disorder structures, like the as-grown CNTs in this study, show good cycling stability by circumventing the problem associated with graphene layer exfoliation by solvent intercalation occurring in highly crystalline graphite [21]. The energy dispersive x-ray spectroscopy (EDS) analysis shows quite a strong carbon (C) peak (higher than 93 wt%) and a weak Cu peak (lower than 3 wt%), demonstrating that most of the Cu was etched away (see the inset table in Fig. 2(d)). The average atomic percentage of C and Cu elements in four different spots is found to be 99.4 at% and 0.4 at%, respectively. In addition, 1.1 wt% Ti originates from the barrier layer that formed in TiC layer during the CNT growth process [22]. Fig. 3(a) shows stacked layers with 3D free-standing CNTs after applying low pressure of 6 bar. The anode stack exhibits the overall thickness of 780 mm with the areal and bulk densities of 34.9 mg cm2 and 0.45 g cm3, respectively. Such high areal density is around three times as high as commercial-grade graphite anodes [17]. The applied pressure causes no severe fracture on the CNTs as evident in the inset of Fig. 3(a). To further enhance the CNTs' density, polyvinylidene fluoride (PVDF) is introduced as a binder and a hot press is subsequently followed, as mentioned in the experimental section. The fabricated anode stack shows the overall thickness of 50 mm, increasing the density by 1.8 g cm3 (see Fig. 3(b)). The denser CNT structure after thermal-press is observed in the inset of Fig. 3(b). The bulk densities for the three different anode stacks of 1S (1 layer), 3S (3 layers), and 5S (5 layers) are 1.3, 1.85, and 1.8 g cm3, respectively. 3.2. Li-ion battery performance of multi-stacked 3D CNT anode
3. Results and discussion 3.1. Morphology and structural properties of multi-stacked 3D CNT anode Fig. 2(a) shows the field emission scanning electron microscope (FESEM) image of the as-prepared 3D-CNTs on Cu, illustrating densely packed and randomly oriented CNTs. Areal loading density of the 3D-CNTs is 2.72 mg cm2, which is 400% higher than that of 2D-CNTs [11]. The loading density can be further enhanced by stacking the 3D-CNTs via the thermal compression process as presented in the next section. Such high improvement of the density is mainly attributed to the surface area enhancement (almost 60%) of the 3D Cu current collector. The average diameter and thickness of CNTs are ~250 nm and 30 mm, respectively (analyzed by the ImageJ software, Figure S1 in supplementary information). Fig. 2(a) represents the SEM image of 3D-CNTs directly
The electrochemical performance of the multi-stacked 3D CNTs anode is carried out by galvanostatic charge and discharge cycling tests at various current densities within a voltage window of 0.01e3.0 V. Note that the 1 C-rate is set as 372 mA g1, based on the theoretical specific capacity of graphite. The variation of specific capacity as a function of the cycle number is shown in Fig. 4(a). The specific capacities for the 1S, 3S, and 5S at 0.5C are in the range of 240~264 mAh g1, 239~259 mAh g1, and 211~238 mAh g1, respectively. The specific capacities obtained during the first charge and discharge cycles show the irreversible capacity loss (Figure S3), which is attributed to the solid electrolyte interphase (SEI) layer formed on the surface of CNTs [4,5]. The higher specific capacity of 1S mainly stems from the higher specific surface area exposed to the electrolyte, allowing lithium-ions to be efficiently intercalated into the CNT structures, compared to the densely packed 3S and 5S. Nevertheless, the good cycling performance of the 3S and 5S is
Fig. 4. Electrochemical performance of the multi-stacked 3D CNT anode as a function of the number of stacked layers. 1S, 3S, and 5S of CNTs have the areal densities of 2.6, 4.5, and 9.0 mg cm2 and the bulk densities of 1.3, 1.85, and 1.8 g cm3 as electroactive anode materials for LIBs, respectively. Cycling performance of all the multi-stacked 3D CNT anode with respect to (a) specific and (b) areal capacities at 0.5C. (c) Voltage profiles of 3S versus specific and volumetric capacities as a function of charge and discharge cycle numbers at 0.5C with an operation voltage range of 0.01e3.0 V. (d) C-rate capability of 3S in regard to specific and volumetric capacities at various current rates. (e) Electrochemical impedance spectra of 3S at the discharged state of 0.01 V after 100 cycles (The inset is the equivalent circuit model to fit the EIS.). (f) The variation of each resistance (Re, Rf, and Rct) with DOD of 3S.
470
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
mainly due to the reinforced bonding strength of CNT networks through the binder-assisted hot press [23]. Fig. 4(b) illustrates the areal capacity versus number of cycles, showing that the areal capacity proportionally increases with the number of stacked layers. 5S delivers an average areal capacity of 1.92 mAh cm2, meeting the commercial standard, which is 173% and 275% higher than those of 3S and 1S, respectively [15]. Such increment of the areal capacity implies that the anode stack can be a promising strategy to increase the loading amount of carbon nanomaterials and corresponding total capacity. The enhancement of bulk density, resulted from the high compression of the anode stack, is beneficial for increasing the volumetric capacity that is a crucial parameter for the large-scale implementation of carbon nanomaterials in LIBs. In this study, the volumetric capacity of multi-stacked 3D CNTs is determined by multiplying the cycled specific capacity (see Fig. 4(a)) by the bulk density (1.3, 1.85, and 1.8 g cm3 for 1S, 3S, and 5S), and the 3S sample yields the highest average volumetric capacity among the tested anodes (323, 456, and 385 mAh cm3 for 1S, 3S, and 5S). After the initial cycles, a good cycling stability (after the 3rd cycle) with the capacity retention of 93% is also achieved. The excellent cyclic stability can be attributed to the high structural integrity of CNT networks in the anode stack made by a thermal compression process. Furthermore, the Coulombic efficiency of each sample reaches above 98% after the 4th cycle, indicative of the high capacity reversibility as evident by Figure S4. It is noted that the controlled stacking process, using a binder and thermal press, is a crucial step to generate reliable cycling performance as evident from the degradation of cycle stability of control sample, which is made through the fabrication processing without binder and thermal compressive pressure, in Figure S5. However, no macro- or microscopic structural damage of the CNTs is observed after 150 cycling at 1C as seen in the SEM and HRTEM images (Figure S6), compared to those of the as-grown CNTs (Figure S1(a) and Fig. 2(b)). Fig. 4(c) exhibits the charge and discharge voltage profiles of the 3S cell at 0.5C for the various cycle numbers. The voltage profiles gradually drop and rise in the voltage level lower than 0.25 V, which is different from those of carbon- and lithium-based crystalline alloys (e.g. lithiated graphite with LiC6) [24]. The main cause for the characteristic voltage profiles is attributed to the microstructure of highly defective CNTs containing a wide range of possible sites with different energies where lithium-ion can reside [25]. The reversible volumetric capacity is in the range of 443~479 mAh cm3, which is superior to those of the commercialegrade graphitic anodes (233~453 mAh cm3) [10,17,18]. The variations of specific and volumetric capacities of the 3S cell at the different C rates are shown in Fig. 4(d). Here we only present the data from the 3S sample for simplicity's sake. We found that the volumetric capacities of 3S with the different C rates are higher than those of the 1S and 5S. We attributed the lower volumetric capacities of 1S and 5S samples to the lower bulk densities and the increased internal resistance, respectively [26,27]. The specific and
volumetric capacities decrease in a staircase fashion with increase of C-rate from 0.1C to 3C. These results are common in LIB cells and mainly due to a decrease in ionic conductivity of the electrolyte with the C-rate increase [28]. At the different C-rates of 0.2C, 0.5C, 1C, and 3C, the reversible capacities reach 312 mAh g1 (578 mAh cm3), 251 mAh g1 (465 mAh cm3), 211 mAh g1 (390 mAh cm3), and 155 mAh g1 (287 mAh cm3), respectively. Furthermore, the recovery of the capacity after the C-rate test up to 3C represents the preservation of the structural integrity of 3S. As shown in Table 1, the volumetric capacities obtained from our multi-stacked 3D-CNT anodes are superior to those of the recently reported carbon nanotubes, porous graphite, and commercially available graphitized mesocarbons. By considering the outstanding features of the anode, it may open up new avenues for the application of carbon nanomaterials to next generation LIBs by providing a higher volumetric capacity. 3.3. Electrochemical impedance spectroscopy (EIS) analysis of multi-stacked 3D CNT anode The Nyquist plot of the 3S after 100 cycles is measured at 0.01 V in the frequency range of 0.01e106 Hz at an amplitude of 5 mV (Fig. 4(e)). The equivalent circuit model is developed as in the inset of Fig. 4(e). The electrolyte resistance (Re) of 15U is determined by the high-frequency intercept with the real part axis of the plot. There are two depressed semi-circles observed in the Nyquist plot: at the high frequency region of 104 ~ 106 Hz, the resistance (Rf) of an SEI layer formation on the CNTs and Li metal is 13U, which is equivalent to the semi-circle diameter of the high frequency region; at the intermediate frequency region of 1 ~ 104 Hz, the charge transfer resistance (Rct) created by the lithium insertion at the electrodes interfaces is measured to be 22U, which is comparable to the previous reports of carbon-based anode materials [30,31]. Note that the impedances of both multi-stacked 3D CNTs and Li metal electrodes are considered for the accurate modeling and analysis of the interfacial charge transport mechanism in the tested cell [32]. As seen in Fig. 4(f), the values of Re, Rf, and Rct are not significantly varied as a function of depth of discharge (DOD), which is indicative of stable resistance of not only the SEI layer on the CNTs and Li metal, but also the CNT/electrolyte and Li metal/electrolyte interfaces (The Nyquist plots and the values of Re, Rf, and Rct for each value of OCV are presented in Figure S7, and Table S1, respectively.). At the low frequency region below 1 Hz, solid state diffusion of lithium-ions into the bulk of CNTs is indicated by a “Warburg impedance (Zw).” 3.4. BET surface area and pore volume of multi-stacked 3D CNT anode It is important to recall that the high surface area and porosity of 3D CNTs are beneficial to the promoted transfer rate of lithium-ions
Table 1 Comparison of material properties and charge capacities of commercially available graphite and lab-scale production of porous graphite and carbon nanotube based electroactive materials for LIB anodes. Material property Graphitized mesocarbon (MCMB 25e28) Nitrogen-doped carbon nanotubes/carbon paper Vertically aligned carbon nanotube Porous graphite Ours
Specific capacity [mAh g1]
Bulk density [g cm3]
Volumetric capacity [mAh cm3]
C-rate (C) or current density [mA g1]
Reference
335 204
0.9 0.24
301.5 48
N/A 0.2C
[10] [14]
782 309 312 251 211
0.51 1.5 1.85
395 463 578 465 390
57 mA g1 0.2C 0.2C 0.5C 1C
[16] [29]
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
and electrons, which is responsible for the enhanced cyclic performance and C-rate capability of the electrode [33]. The BET surface area of the multi-stacked 3D CNTs (29 m2 g1) is lower than that of the pristine 3D CNTs (63 m2 g1) due to the binderpassivated surface and compaction of CNTs by the post-treated binder at high pressure; however, it is still comparable to that of multi-wall carbon nanotube-based free-standing films (50 m2 g1) previously investigated [34]. According to Non-Local Density Functional Theory (NLDFT) model [35], the total pore volumes of the pristine 3D CNTs and the multi-stacked 3D CNTs samples are 0.15 cm3 g1 and 0.18 cm3 g1, respectively, which are greater than that of the multi-wall carbon nanotube-based free-standing films (0.13 cm3 g1) [34]. The BET surface area and total pore volume of the multi-stacked 3D CNTs confirm that there is no problem with the effective transfer of lithium-ions and electrons through an electrolyte and an electrode into the thick electroactive materials of our proposed multi-stacked CNTs. 4. Conclusion We have designed the multi-stacked 3D-CNT anodes to address the issues of low areal and volumetric capacities of nanostructured carbon materials for the implementation in large-scale advanced LIB. The design involves the hot-press assisted stacking method of 3D free-standing CNT layers to achieve the high areal and bulk densities while maintaining the high structural integrity and electrical conductivity of the CNTs. The multi-stacked 3D-CNT anodes showed the bulk density of 1.85 g cm3 as one of the highest values reported from carbon materials and delivered the average volumetric capacity of 646 mAh cm3 at 0.1C that is higher than that of the state-of-the-art graphitic anode used in LIB. In addition, the proposed process for the anode stack is compatible with the conventional anode fabrication process. With such excellent electrochemical performance, the proposed multi-stacked 3D-CNT anodes could overcome the current limitation, low volumetric capacity, of the CNT-based anode materials for the application in the nextgeneration LIBs. Acknowledgment A part of this work was supported by the Korea Institute of Energy Research (KIER) (B5-2498). W.C. acknowledges partial support from the KIST Institutional Program. We are grateful to Mr. D. Kim and Mr. J. Smith who provided us with the schematic diagrams. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpowsour.2015.08.103. References [1] A. Manthiram, Materials aspects: an overview, in: G.A. Nazri, G. Pistoia (Eds.), Lithium Batteries : Science and Technology, Springer, New York, 2004, pp. 1e11. [2] J.M. Tarascon, M. Armand, Issues and challenges facing rechargeable lithium batteries, Nature 414 (2001) 359e367. [3] P.M. Ajayan, O. Zhou, Applications of carbon nanotubes, in: M.S. Dresselhaus, G. Dresselhaus, P.H. Avouris (Eds.), Carbon Nanotubes Synthesis, Structure, Properties, and Application, Springer, New York, 2001, pp. 401e404. [4] B.J. Landi, M.J. Ganter, C.D. Cress, R.A. DiLeo, R.P. Raffaelle, Carbon nanotubes for lithium ion batteries, Energy. Environ. Sci. 2 (2009) 638e654. [5] C.D.L. Casas, W. Li, A review of application of carbon nanotubes for lithium ion battery anode material, J. Power. Sources 208 (2012) 74e85. [6] X.X. Wang, J.N. Wang, H. Chang, Y.F. Zhang, Preparation of short carbon nanotubes and application as an electrode material in Li-ion batteries, Adv. Funct. Mater. 17 (2007) 3613e3618.
471
[7] Y. Gogotsi, P. Simon, True performance metrics in electrochemical energy storage, Science 334 (2011) 917e918. [8] C. Zhang, W. Lv, Y. Tao, Q.H. Yang, Towards superior volumetric performance: design and preparation of novel carbon materials for energy storage, Energy. Environ. Sci. 8 (2015) 1390e1403. [9] Z. Chen, D. Zhang, X. Wang, X. Jia, F. Wei, H. Li, Y. Lu, High-performance energy-storage architectures from carbon nanotubes and nanocrystal building blocks, Adv. Mater. 24 (2012) 2030e2036. k, D. Goers, M.E. Spahr, Carbon materials in lithium-ion batteries, in: [10] P. Nova guin, E. Fra˛ ckowiak (Eds.), Carbons for Electrochemical Energy Storage F. Be and Conversion Systems, CRC Press, Boca Raton, 2010, p. 300. [11] C.W. Kang, I. Lahiri, R. Baskaran, W.G. Kim, Y.K. Sun, W.B. Choi, 3-dimensional carbon nanotube for Li-ion battery anode, J. Power. Sources 219 (2012) 364e370. [12] H.G. Zhang, P.V. Braun, Three-dimensional metal scaffold supported bicontinuous silicon battery anodes, Nano. Lett. 12 (2012) 2778e2783. [13] S.J. Dillon, K. Sun, Microstructural design considerations for Li-ion battery systems, Curr. Opin. Solid. State. Mater. Sci. 16 (2012) 153e162. € m, K. Edstro €m, T. Wågberg, [14] T. Sharifi, M. Valvo, E. Gracia-Espino, R. Sandstro Hierarchical self-assembled structures based on nitrogen-doped carbon nanotubes as advanced negative electrodes for Li-ion batteries and 3D microbatteries, J. Power. Sources 279 (2015) 581e592. [15] L.F. Cui, L. Hu, J.W. Choi, Y. Cui, Light-weight free-standing carbon nanotubesilicon films for anodes of lithium ion batteries, ACS. Nano 4 (2010) 3671e3678. [16] D.T. Welna, L. Qu, B.E. Taylor, L. Dai, M.F. Durstock, Vertically aligned carbon nanotube electrodes for lithium-ion batteries, J. Power. Sources 196 (2011) 1455e1460. [17] R. Moshtev, B. Johnson, State of the art of commercial Li ion batteries, J. Power. Sources 91 (2000) 86e91. [18] Q. Xiao, Y. Fan, X. Wang, R.A. Susantyokoa, Q. Zhang, A multilayer Si/CNT coaxial nanofiber LIB anode with a high areal capacity, Energy. Environ. Sci. 7 (2014) 655e661. [19] J.K. Park, Materials for lithium secondary batteries, in: J.K. Park (Ed.), Principles and Applications of Lithium Secondary Batteries, Wiley-VCH, Weinheim, 2012, pp. 92e118. [20] S. Osswald, M. Havel, Y. Gogotsi, Monitoring oxidation of multiwalled carbon nanotubes by Raman spectroscopy, J. Raman. Spectrosc. 38 (2007) 728e736. [21] M. Winter, K.C. Moeller, J.O. Besenhard, Carbonaceous and graphitic anodes, in: G.A. Nazri, G. Pistoia (Eds.), Lithium Batteries : Science and Technology, Springer, New York, 2004, p. 177. [22] I. Lahiri, S.W. Oh, J.Y. Hwang, S.J. Cho, Y.K. Sun, R. Banerjee, W.B. Choi, High capacity and excellent stability of lithium ion battery anode using interfacecontrolled binder-free multiwall carbon nanotubes grown on copper, ACS. Nano 4 (2010) 3440e3446. [23] B. Marinho, M. Ghislandi, E. Tkalya, C.E. Koning, G.D. With, Electrical conductivity of compacts of graphene, multi-wall carbon nanotubes, carbon black, and graphite powder, Powder. Technol. 221 (2012) 351e358. [24] Y. Rachid, Thermodynamics of electrode materials for lithium-ion batteries, in: K. Ozawa (Ed.), Lithium Ion Rechargeable Batteries, Wiley-VCH, Weinheim, 2009, pp. 67e86. [25] R.A. Huggins, Negative electrodes in lithium cells, in: R.A. Huggins (Ed.), Advanced Batteries : Materials Sciences Aspects, Springer, New York, 2010, pp. 134e135. [26] A. Shah, M.N. Ates, S. Kotz, J. Seo, M. Abraham, S. Somu, A. Busnaina, A layered carbon nanotube architecture for high power lithium ion batteries, J. Electrochem. Soc. 161 (6) (2014) A989eA995. [27] L. Hu, F.L. Mantia, H. Wu, X. Xie, J. McDonough, M. Pasta, Y. Cui, Lithium-ion textile batteries with large areal mass loading, Adv. Energy Mater. 1 (2011) 1012e1017. [28] J.F. Fauvarque, P. Simon, Principles of electrochemistry and electrochemical guin, E. Fra˛ ckowiak (Eds.), Carbons for Electrochemical methods, in: F. Be Energy Storage and Conversion Systems, CRC Press, Boca Raton, 2010, pp. 7e16. [29] S. Jeong, J.P. Lee, M. Ko, G. Kim, S. Park, J. Cho, Etched graphite with internally grown Si nanowires from pores as an anode for high density Li-ion batteries, Nano Lett. 13 (2013) 3403e3407. [30] S. Yang, J. Huo, H. Song, X. Chen, A comparative study of electrochemical properties of two kinds of carbon nanotubes as anode materials for lithium ion batteries, Electrochim. Acta 53 (2008) 2238e2244. guin, F. Chevallier, C. Vix-Guterl, S. Saadallah, V. Bertagna, J.N. Rouzaud, [31] F. Be E. Frackowiak, Correlation of the irreversible lithium capacity with the active surface area of modified carbons, Carbon 43 (2005) 2160e2167. [32] J.Y. Song, H.H. Lee, Y.Y. Wang, C.C. Wan, Two- and three-electrode impedance spectroscopy of lithium-ion batteries, J. Power. Sources 111 (2002) 255e267. [33] L. Qie, W.-M. Chen, Z.-H. Wang, Q.-G. Shao, X. Li, L.-X. Yuan, X.-L. Hu, W.X. Zhang, Y.-H. Huang, Nitrogen-doped porous carbon nanofiber webs as anodes for lithium-ion batteries with a superhigh capacity and rate capability, Adv. Mater. 24 (2012) 2047e2050. [34] S.Y. Chew, S.H. Ng, J. Wang, P. Nov ak, F. Krumeich, S.L. Chou, J. Chen, H.K. Liu, Flexible free-standing carbon nanotube films for model lithium-ion batteries, Carbon 47 (2009) 2976e2983. [35] J. Landers, G.Y. Gor, A.,V. Neimark, Density functional theory methods for characterization of porous materials, Colloids. Surf. A 437 (2013) 3e32.
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Three-dimensional carbon nanotubes for high capacity lithium-ion batteries Chiwon Kang a, Mumukshu Patel a, Baskaran Rangasamy a, 1, Kyu-Nam Jung c, Changlei Xia b, Sheldon Shi b, Wonbong Choi a, b, * a b c
Department of Materials Science and Engineering, University of North Texas, North Texas Discovery Park 3940 North Elm St., Denton, TX 76207, USA Department of Mechanical and Energy Engineering, University of North Texas, North Texas Discovery Park, 3940 North Elm St., Denton, TX 76207, USA Energy Efficiency and Materials Research Division, Korea Institute of Energy Research, 152 Gajeong-ro, Yuseong-gu, Daejeon 305-343, Republic of Korea
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
We show a novel structure of multistacked 3D CNTs for a higher loading of CNTs. The bulk density of multi-stacked 3D CNTs is twice as high as that of graphites. The multi-stacked 3D CNTs yield a stable and high reversible volumetric capacity.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 June 2015 Received in revised form 29 August 2015 Accepted 31 August 2015 Available online xxx
Carbon nanotubes (CNTs) have been considered as a potential anode material for next generation Lithium-ion batteries (LIBs) due to their high conductivity, flexibility, surface area, and lithium-ion insertion ability. However, the low mass loading and bulk density of carbon nanomaterials hinder their use in large-scale energy storage because their high specific capacity may not scale up linearly with the thickness of the electrode. To address this issue, a novel three-dimensional (3D) architecture is rationally designed by stacking layers of free-standing CNTs with the increased areal density to 34.9 mg cm2, which is around three-times higher than that of the state-of-the-art graphitic anodes. Furthermore, a thermal compression process renders the bulk density of the multi-stacked 3D CNTs to be increased by 1.85 g cm3, which yields an excellent volumetric capacity of 465 mAh cm3 at 0.5C. Our proposed strategy involving the stacking of 3D CNT based layers and post-thermal compression provides a powerful platform for the utilization of carbon nanomaterials in the advanced LIB technology. © 2015 Elsevier B.V. All rights reserved.
Keywords: 3-Dimensional free-standing carbon nanotubes Lithium ion batteries Volumetric capacity Areal capacity Bulk density Multi-layered anode stack
1. Introduction * Corresponding author. Department of Materials Science and Engineering, University of North Texas, North Texas Discovery Park 3940 North Elm St., Denton, TX 76207, USA. E-mail address: [email protected] (W. Choi). 1 Present address: Department of Physics, School of Basic and Applied Sciences, Central University of Tamilnadu, Thiruvarur, Tamilnadu, India. http://dx.doi.org/10.1016/j.jpowsour.2015.08.103 0378-7753/© 2015 Elsevier B.V. All rights reserved.
The lithium-ion battery (LIB) has been one of the most commonly used state-of-the-art energy storage systems since it was first commercialized in 1990. The commercial success of the LIB is mainly attributed to the unique features of high operating
466
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
potential (>3.0 V), high energy density (110e170 Wh kg1 and 230e400 Wh L1), appropriate cyclability, and cost effectiveness [1,2]. To further enhance the LIB performance, attempts have been made to investigate more efficient anode materials alternative to the conventional graphite anodes offering a theoretical specific capacity of 372 mAh g1. Among the potential carbon-based materials, the quest for carbon nanotubes (CNTs) has been particularly promising due to their excellent electrochemical and physicochemical properties [3]. The excellent electrochemical performances of CNTs for LIB are mainly associated with the CNTs' unique characteristics such as high surface area, short diffusion length for lithium-ions, and excellent electrical and mechanical properties [4e6]. Although CNTs have been proven as emerging anode material candidates to yield high specific capacity for LIBs, there are still legitimate limitations to be examined for the application of the CNTs to commercial products [7]. First, the previously reported CNTs with high specific capacity are, in many cases, dimensionally confined in thin film geometries that are not suitable for large-scale LIBs. Second, the specific capacity does not scale proportionally with the thickness of CNTs when the weight of electroactive CNTs is negligibly low relative to the total weight of an LIB cell. Third, CNTs have intrinsically low volumetric capacity that is a critical property for large-scale energy storage systems [8]. As a key property to enhance volumetric capacity, the bulk density (mass divided by total volume of solid matter) of most nanomaterial-based electrodes including CNTs is less than 1 g cm3, which is a mediocre density compared to that of commercial-grade graphite (0.35~0.9 g cm3) [9,10]. Due to the above-mentioned limitations, there is no clear confirmation that CNTs are indeed pertinent anode nanomaterials for high volumetric capacities to be used in largescale energy storage systems. To address these challenging issues of the CNT anode, we have introduced a 3D micro-channeled copper (Cu) that has a high surface area to accommodate the large loading amounts of CNTs [11]. In addition, the high porosity of 3D CNTs is suited to facilitate lithium-ion diffusion and electron charge transfer through the electrolyte and bulk electrode into the 3D CNTs as electroactive materials; as a result, the high porosity can enhance cyclic performance and current rate capability of the electrode [12,13]. In recent years, different approaches have been applied to enhance areal density (mass per unit area) and capacity (storage capacity per unit area) as well as bulk density and volumetric capacity. T. Sharifi et al. made the nitrogen-doped CNTs grown on carbon paper substrate by using a catalytic chemical vapor deposition (CVD) method, and the areal density and capacity of the CNTs/carbon paper electrode are 9.6 mg cm2 and 1.95 mAh cm2 at 0.2C, respectively [14], which are comparable to those of commercially available LIBs [15]. D. T. Welna et al. fabricated a freestanding film of vertically aligned multi-wall CNTs on a nickel
substrate using a combined method of catalytic thermal CVD and sputter deposition, and the CNT anode showed the bulk density and volumetric capacity of 0.51 g cm3 and 395 mAh cm3 at 57 mA g1 [16], respectively, which are comparable to those of commercial graphites [10,17,18]. In this study, we report that multi-stacked 3D CNTs demonstrate a stable and reversible volumetric capacity higher than that of the state-of-the-art graphitic anode used in an LIB. The structure of multi-stacked 3D CNTs provides one of the most pragmatic approaches towards the enhanced LIB performance of CNTs as promising anode nanomaterials for advanced large-scale LIBs. 2. Experimental 2.1. Fabrication of multi-stacked 3D CNT anode In this work, a multi-stacked 3D CNT-based anode was prepared through the following key steps: (1) 3D CNTs were directly grown on a Cu mesh (>99% purity) with 50 mm thickness and 65 mm hole size (TWP Inc.) as a substrate by using a thermal catalytic CVD method; (2) The Cu substrate was removed by ferric chloride (FeCl3) etching solution (Transene Company, Inc.) to obtain 3D freestanding CNT layers; and (3) The 3D free-standing CNT layers were stacked through the vacuum annealing and thermal compression methods to fabricate multi-stacked 3D CNTs with a high bulk density. Fig. 1 schematically represents the process for the formation of the multi-stacked 3D CNT-based anode. In Fig. 1(a), CNTs were directly grown on Cu mesh by a CVD method, where nickel (Ni) and titanium (Ti) were deposited onto Cu, using a radio frequency (RF)-direct current (DC) magnetron-sputtering system, which served as a catalyst for CNT growth and a barrier layer against carbon diffusion into Cu during CNT growth, respectively. The Ni/Ti deposited Cu mesh was placed in a CVD system, and the mixture gas of H2 and C2H4 with a volume ratio of 1:2 flowed into the furnace for 50 min at the growth temperature of 750 C. The weights of each sample before and after CNT growth were measured to calculate the net weight of CNT as an electroactive material for a LIB. After 4 h of Cu etching, Cu mesh was completely dissolved into the concentrated FeCl3 solution, and the solution was left with 3D free-standing CNT structures (see Fig. 1(b)). The freestanding structures were thoroughly washed with de-ionized water and dried in a vacuum oven (Thermo Scientific Lindberg) at 120 C for 4 h as presented in Fig. 1(c). To prepare a highly dense multi-stacked 3D CNT anode, we employed polyvinylidene fluoride (PVDF, SigmaeAldrich) as a binder dissolved into N-Methyl-2pyrrolidone (NMP) solvent. The solution is composed of 1.4 wt% of PVDF and 98.6 wt% of NMP (the weight ratio of CNTs to PVDF is 8:2). Two pieces of free-standing CNT layers were sandwiched within the solution and subsequently heated at 120 C under a
Fig. 1. Process flow for the fabrication of multi-stacked 3D CNT anode. (a) CVD grown 3D CNTs on Cu mesh. (b) Chemical etching process of the Cu mesh via the FeCl3 etching solution. (c) 3D free-standing CNT structure after the etching process. (d) The multi-stacked 3D CNT anode fabricated by pressing the layers of 3D CNTs with the aid of PVDF binder.
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
467
Fig. 2. Structural properties of multi-stacked 3D CNT anode. (a) FESEM plain-view image of randomly oriented and highly dense 3D CNTs directly grown on Cu mesh using a CVD method and schematic diagram of the 3D CNTs on Cu. (b) High resolution TEM (HRTEM) image of a single CNT comprising the turbostratic or low-ordered graphitization structure. (c) Raman spectra of a 3D CNT showing the intensity ratio of D-band to G-band peaks (ID/IG > 1). (d) Energy Dispersive X-ray Spectroscopy (EDS) spectra obtained from a spot on the 3D free-standing CNT structure shown in the SEM image of (a).
vacuum pressure of 760 mmHg to completely evaporate the NMP solvent. The processing step was repeated until the desired number of stacked layers was reached. Finally, the stacked layers were thermally compressed at 60 C and 5000 bar (Fig. 1(d)). The thickness and footprint areas of the as-prepared anode stack were measured by a digital caliper (VWR) to calculate bulk density for the measurement of volumetric capacity. For comparison, the control anode stack was also prepared at room temperature by using a precision press (MTI Inc.) with a pressure of 6 bar, without using the binder.
2.2. Structural characterization The as-grown CNTs were characterized by using a field emission scanning electron microscope (FESEM) (JEOL, JSM-7000F), energy dispersive X-ray spectrometer (EDS) (Thermo Electron Corporation, NORAN System SIX), Raman spectrometer (Arþ laser with l ¼ 514 nm, 33 mW power), and field emission transmission electron microscope (FETEM) (FEI, TECHNAI F20). The BrunauereEmmetteTeller (BET) specific surface area (SSA) and the total pore volume of the samples were measured by using a surface
Fig. 3. FESEM cross-sectional images of (a) the control multi-stacked 3D CNT anode without the treatment of binder and high pressure. (b) The denser structure of CNTs could be fabricated by applying the binder and high compressive pressure at 60 C.
468
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
area/pore size analyzer (3Flex 3500, Micromeritics Instrument Corp.) at 77 K. Before characterizing the pore volume, the samples were degassed at 100 C for 3 days at ~0.1 mbar by employing a degasser (VacPrep 061, Micromeritics Instrument Corp.) followed by in-situ degassing in the surface area/pore size analyzer. The total pore volume was measured using a Non-Local Density Functional Theory (NLDFT) model based on the amount of nitrogen adsorbed up to P/P0 ~ 0.99. 2.3. Electrochemical performance The electrochemical performance of the multi-stacked 3D CNTbased anode was measured in a coin cell (CR 2032, Wellcos Ltd.). Half-cell assembly was fabricated in an argon-filled glovebox (VAC atmosphere Ltd. and MBraun) under extremely low levels of humidity and oxygen (<0.5 ppm). The 3D-CNT anode was used as a working electrode and Li metal foil served as a counter and reference electrode. 1 mol dm3 solution of lithium hexafluorophosphate (LiPF6) salt in 1:1:1 (volume ratio) mixture solvent of ethylene carbonate (EC), dimethylene carbonate (DMC), and diethylene carbonate (DEC) (MTI Corp.) was used as an electrolyte. A typical polypropylene (PP) based membrane (Separator-2400, Wellcos Ltd.) was used as a separator. The cell was finally assembled using a CR 2032 coin-cell crimping tool (Hohsen Corp.). The charge (delithiation) and discharge (lithiation) cycling tests were performed in a multi-channel battery testing unit (MACCOR-series 4000) at room temperature in the voltage window of 0.01e3.0 V. 2.4. Electrochemical impedance spectroscopy (EIS) measurement The EIS measurement for the 3D-CNT anode after 100 cycles was conducted at a series of open-circuit voltages (OCVs) in a frequency range from 0.01 Hz to 106 Hz using a potentiostat (Reference 3000, Gamry Instrument). Each OCV determines the depth of discharge (DOD) obtained by the lithium insertion from the anode stack. The perturbation amplitude was set as ±5 mV and ten points per decade were collected. The impedance data of real and imaginary parts were plotted as a function of frequency in a complex plane diagram (Nyquist plot) and were fitted using a computer software (Echem Analyst™, Gamry Instrument).
469
grown on Cu, showing highly dense and randomly oriented CNTs. At a microscopic level, the CNT has a low-ordered graphitization structure, as confirmed by a high resolution transmission electron microscopy (HRTEM) (see Fig. 2(b)) [19]. As seen in Fig. 2(c), typical Raman spectra of a 3D CNT indicate two strong peaks at ~1350 cm1 and ~1580 cm1, which are ascribed to D-band (defect related) and G-band (E2g in-plane vibration mode). The intensity ratio (ID/IG) is higher than 1.0, implying that the CNTs are composed of low-ordered graphitization structures [20]. The XRD patterns also identify the low-ordered graphitization structures of CNTs based on the broad peak related to hexagonal carbon (002) appearing at 2q ¼ 25.5 (JCPDS card no. 26-1079) as presented in Figure S2. It is known that the disorder structures, like the as-grown CNTs in this study, show good cycling stability by circumventing the problem associated with graphene layer exfoliation by solvent intercalation occurring in highly crystalline graphite [21]. The energy dispersive x-ray spectroscopy (EDS) analysis shows quite a strong carbon (C) peak (higher than 93 wt%) and a weak Cu peak (lower than 3 wt%), demonstrating that most of the Cu was etched away (see the inset table in Fig. 2(d)). The average atomic percentage of C and Cu elements in four different spots is found to be 99.4 at% and 0.4 at%, respectively. In addition, 1.1 wt% Ti originates from the barrier layer that formed in TiC layer during the CNT growth process [22]. Fig. 3(a) shows stacked layers with 3D free-standing CNTs after applying low pressure of 6 bar. The anode stack exhibits the overall thickness of 780 mm with the areal and bulk densities of 34.9 mg cm2 and 0.45 g cm3, respectively. Such high areal density is around three times as high as commercial-grade graphite anodes [17]. The applied pressure causes no severe fracture on the CNTs as evident in the inset of Fig. 3(a). To further enhance the CNTs' density, polyvinylidene fluoride (PVDF) is introduced as a binder and a hot press is subsequently followed, as mentioned in the experimental section. The fabricated anode stack shows the overall thickness of 50 mm, increasing the density by 1.8 g cm3 (see Fig. 3(b)). The denser CNT structure after thermal-press is observed in the inset of Fig. 3(b). The bulk densities for the three different anode stacks of 1S (1 layer), 3S (3 layers), and 5S (5 layers) are 1.3, 1.85, and 1.8 g cm3, respectively. 3.2. Li-ion battery performance of multi-stacked 3D CNT anode
3. Results and discussion 3.1. Morphology and structural properties of multi-stacked 3D CNT anode Fig. 2(a) shows the field emission scanning electron microscope (FESEM) image of the as-prepared 3D-CNTs on Cu, illustrating densely packed and randomly oriented CNTs. Areal loading density of the 3D-CNTs is 2.72 mg cm2, which is 400% higher than that of 2D-CNTs [11]. The loading density can be further enhanced by stacking the 3D-CNTs via the thermal compression process as presented in the next section. Such high improvement of the density is mainly attributed to the surface area enhancement (almost 60%) of the 3D Cu current collector. The average diameter and thickness of CNTs are ~250 nm and 30 mm, respectively (analyzed by the ImageJ software, Figure S1 in supplementary information). Fig. 2(a) represents the SEM image of 3D-CNTs directly
The electrochemical performance of the multi-stacked 3D CNTs anode is carried out by galvanostatic charge and discharge cycling tests at various current densities within a voltage window of 0.01e3.0 V. Note that the 1 C-rate is set as 372 mA g1, based on the theoretical specific capacity of graphite. The variation of specific capacity as a function of the cycle number is shown in Fig. 4(a). The specific capacities for the 1S, 3S, and 5S at 0.5C are in the range of 240~264 mAh g1, 239~259 mAh g1, and 211~238 mAh g1, respectively. The specific capacities obtained during the first charge and discharge cycles show the irreversible capacity loss (Figure S3), which is attributed to the solid electrolyte interphase (SEI) layer formed on the surface of CNTs [4,5]. The higher specific capacity of 1S mainly stems from the higher specific surface area exposed to the electrolyte, allowing lithium-ions to be efficiently intercalated into the CNT structures, compared to the densely packed 3S and 5S. Nevertheless, the good cycling performance of the 3S and 5S is
Fig. 4. Electrochemical performance of the multi-stacked 3D CNT anode as a function of the number of stacked layers. 1S, 3S, and 5S of CNTs have the areal densities of 2.6, 4.5, and 9.0 mg cm2 and the bulk densities of 1.3, 1.85, and 1.8 g cm3 as electroactive anode materials for LIBs, respectively. Cycling performance of all the multi-stacked 3D CNT anode with respect to (a) specific and (b) areal capacities at 0.5C. (c) Voltage profiles of 3S versus specific and volumetric capacities as a function of charge and discharge cycle numbers at 0.5C with an operation voltage range of 0.01e3.0 V. (d) C-rate capability of 3S in regard to specific and volumetric capacities at various current rates. (e) Electrochemical impedance spectra of 3S at the discharged state of 0.01 V after 100 cycles (The inset is the equivalent circuit model to fit the EIS.). (f) The variation of each resistance (Re, Rf, and Rct) with DOD of 3S.
470
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
mainly due to the reinforced bonding strength of CNT networks through the binder-assisted hot press [23]. Fig. 4(b) illustrates the areal capacity versus number of cycles, showing that the areal capacity proportionally increases with the number of stacked layers. 5S delivers an average areal capacity of 1.92 mAh cm2, meeting the commercial standard, which is 173% and 275% higher than those of 3S and 1S, respectively [15]. Such increment of the areal capacity implies that the anode stack can be a promising strategy to increase the loading amount of carbon nanomaterials and corresponding total capacity. The enhancement of bulk density, resulted from the high compression of the anode stack, is beneficial for increasing the volumetric capacity that is a crucial parameter for the large-scale implementation of carbon nanomaterials in LIBs. In this study, the volumetric capacity of multi-stacked 3D CNTs is determined by multiplying the cycled specific capacity (see Fig. 4(a)) by the bulk density (1.3, 1.85, and 1.8 g cm3 for 1S, 3S, and 5S), and the 3S sample yields the highest average volumetric capacity among the tested anodes (323, 456, and 385 mAh cm3 for 1S, 3S, and 5S). After the initial cycles, a good cycling stability (after the 3rd cycle) with the capacity retention of 93% is also achieved. The excellent cyclic stability can be attributed to the high structural integrity of CNT networks in the anode stack made by a thermal compression process. Furthermore, the Coulombic efficiency of each sample reaches above 98% after the 4th cycle, indicative of the high capacity reversibility as evident by Figure S4. It is noted that the controlled stacking process, using a binder and thermal press, is a crucial step to generate reliable cycling performance as evident from the degradation of cycle stability of control sample, which is made through the fabrication processing without binder and thermal compressive pressure, in Figure S5. However, no macro- or microscopic structural damage of the CNTs is observed after 150 cycling at 1C as seen in the SEM and HRTEM images (Figure S6), compared to those of the as-grown CNTs (Figure S1(a) and Fig. 2(b)). Fig. 4(c) exhibits the charge and discharge voltage profiles of the 3S cell at 0.5C for the various cycle numbers. The voltage profiles gradually drop and rise in the voltage level lower than 0.25 V, which is different from those of carbon- and lithium-based crystalline alloys (e.g. lithiated graphite with LiC6) [24]. The main cause for the characteristic voltage profiles is attributed to the microstructure of highly defective CNTs containing a wide range of possible sites with different energies where lithium-ion can reside [25]. The reversible volumetric capacity is in the range of 443~479 mAh cm3, which is superior to those of the commercialegrade graphitic anodes (233~453 mAh cm3) [10,17,18]. The variations of specific and volumetric capacities of the 3S cell at the different C rates are shown in Fig. 4(d). Here we only present the data from the 3S sample for simplicity's sake. We found that the volumetric capacities of 3S with the different C rates are higher than those of the 1S and 5S. We attributed the lower volumetric capacities of 1S and 5S samples to the lower bulk densities and the increased internal resistance, respectively [26,27]. The specific and
volumetric capacities decrease in a staircase fashion with increase of C-rate from 0.1C to 3C. These results are common in LIB cells and mainly due to a decrease in ionic conductivity of the electrolyte with the C-rate increase [28]. At the different C-rates of 0.2C, 0.5C, 1C, and 3C, the reversible capacities reach 312 mAh g1 (578 mAh cm3), 251 mAh g1 (465 mAh cm3), 211 mAh g1 (390 mAh cm3), and 155 mAh g1 (287 mAh cm3), respectively. Furthermore, the recovery of the capacity after the C-rate test up to 3C represents the preservation of the structural integrity of 3S. As shown in Table 1, the volumetric capacities obtained from our multi-stacked 3D-CNT anodes are superior to those of the recently reported carbon nanotubes, porous graphite, and commercially available graphitized mesocarbons. By considering the outstanding features of the anode, it may open up new avenues for the application of carbon nanomaterials to next generation LIBs by providing a higher volumetric capacity. 3.3. Electrochemical impedance spectroscopy (EIS) analysis of multi-stacked 3D CNT anode The Nyquist plot of the 3S after 100 cycles is measured at 0.01 V in the frequency range of 0.01e106 Hz at an amplitude of 5 mV (Fig. 4(e)). The equivalent circuit model is developed as in the inset of Fig. 4(e). The electrolyte resistance (Re) of 15U is determined by the high-frequency intercept with the real part axis of the plot. There are two depressed semi-circles observed in the Nyquist plot: at the high frequency region of 104 ~ 106 Hz, the resistance (Rf) of an SEI layer formation on the CNTs and Li metal is 13U, which is equivalent to the semi-circle diameter of the high frequency region; at the intermediate frequency region of 1 ~ 104 Hz, the charge transfer resistance (Rct) created by the lithium insertion at the electrodes interfaces is measured to be 22U, which is comparable to the previous reports of carbon-based anode materials [30,31]. Note that the impedances of both multi-stacked 3D CNTs and Li metal electrodes are considered for the accurate modeling and analysis of the interfacial charge transport mechanism in the tested cell [32]. As seen in Fig. 4(f), the values of Re, Rf, and Rct are not significantly varied as a function of depth of discharge (DOD), which is indicative of stable resistance of not only the SEI layer on the CNTs and Li metal, but also the CNT/electrolyte and Li metal/electrolyte interfaces (The Nyquist plots and the values of Re, Rf, and Rct for each value of OCV are presented in Figure S7, and Table S1, respectively.). At the low frequency region below 1 Hz, solid state diffusion of lithium-ions into the bulk of CNTs is indicated by a “Warburg impedance (Zw).” 3.4. BET surface area and pore volume of multi-stacked 3D CNT anode It is important to recall that the high surface area and porosity of 3D CNTs are beneficial to the promoted transfer rate of lithium-ions
Table 1 Comparison of material properties and charge capacities of commercially available graphite and lab-scale production of porous graphite and carbon nanotube based electroactive materials for LIB anodes. Material property Graphitized mesocarbon (MCMB 25e28) Nitrogen-doped carbon nanotubes/carbon paper Vertically aligned carbon nanotube Porous graphite Ours
Specific capacity [mAh g1]
Bulk density [g cm3]
Volumetric capacity [mAh cm3]
C-rate (C) or current density [mA g1]
Reference
335 204
0.9 0.24
301.5 48
N/A 0.2C
[10] [14]
782 309 312 251 211
0.51 1.5 1.85
395 463 578 465 390
57 mA g1 0.2C 0.2C 0.5C 1C
[16] [29]
C. Kang et al. / Journal of Power Sources 299 (2015) 465e471
and electrons, which is responsible for the enhanced cyclic performance and C-rate capability of the electrode [33]. The BET surface area of the multi-stacked 3D CNTs (29 m2 g1) is lower than that of the pristine 3D CNTs (63 m2 g1) due to the binderpassivated surface and compaction of CNTs by the post-treated binder at high pressure; however, it is still comparable to that of multi-wall carbon nanotube-based free-standing films (50 m2 g1) previously investigated [34]. According to Non-Local Density Functional Theory (NLDFT) model [35], the total pore volumes of the pristine 3D CNTs and the multi-stacked 3D CNTs samples are 0.15 cm3 g1 and 0.18 cm3 g1, respectively, which are greater than that of the multi-wall carbon nanotube-based free-standing films (0.13 cm3 g1) [34]. The BET surface area and total pore volume of the multi-stacked 3D CNTs confirm that there is no problem with the effective transfer of lithium-ions and electrons through an electrolyte and an electrode into the thick electroactive materials of our proposed multi-stacked CNTs. 4. Conclusion We have designed the multi-stacked 3D-CNT anodes to address the issues of low areal and volumetric capacities of nanostructured carbon materials for the implementation in large-scale advanced LIB. The design involves the hot-press assisted stacking method of 3D free-standing CNT layers to achieve the high areal and bulk densities while maintaining the high structural integrity and electrical conductivity of the CNTs. The multi-stacked 3D-CNT anodes showed the bulk density of 1.85 g cm3 as one of the highest values reported from carbon materials and delivered the average volumetric capacity of 646 mAh cm3 at 0.1C that is higher than that of the state-of-the-art graphitic anode used in LIB. In addition, the proposed process for the anode stack is compatible with the conventional anode fabrication process. With such excellent electrochemical performance, the proposed multi-stacked 3D-CNT anodes could overcome the current limitation, low volumetric capacity, of the CNT-based anode materials for the application in the nextgeneration LIBs. Acknowledgment A part of this work was supported by the Korea Institute of Energy Research (KIER) (B5-2498). W.C. acknowledges partial support from the KIST Institutional Program. We are grateful to Mr. D. Kim and Mr. J. Smith who provided us with the schematic diagrams. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpowsour.2015.08.103. References [1] A. Manthiram, Materials aspects: an overview, in: G.A. Nazri, G. Pistoia (Eds.), Lithium Batteries : Science and Technology, Springer, New York, 2004, pp. 1e11. [2] J.M. Tarascon, M. Armand, Issues and challenges facing rechargeable lithium batteries, Nature 414 (2001) 359e367. [3] P.M. Ajayan, O. Zhou, Applications of carbon nanotubes, in: M.S. Dresselhaus, G. Dresselhaus, P.H. Avouris (Eds.), Carbon Nanotubes Synthesis, Structure, Properties, and Application, Springer, New York, 2001, pp. 401e404. [4] B.J. Landi, M.J. Ganter, C.D. Cress, R.A. DiLeo, R.P. Raffaelle, Carbon nanotubes for lithium ion batteries, Energy. Environ. Sci. 2 (2009) 638e654. [5] C.D.L. Casas, W. Li, A review of application of carbon nanotubes for lithium ion battery anode material, J. Power. Sources 208 (2012) 74e85. [6] X.X. Wang, J.N. Wang, H. Chang, Y.F. Zhang, Preparation of short carbon nanotubes and application as an electrode material in Li-ion batteries, Adv. Funct. Mater. 17 (2007) 3613e3618.
471
[7] Y. Gogotsi, P. Simon, True performance metrics in electrochemical energy storage, Science 334 (2011) 917e918. [8] C. Zhang, W. Lv, Y. Tao, Q.H. Yang, Towards superior volumetric performance: design and preparation of novel carbon materials for energy storage, Energy. Environ. Sci. 8 (2015) 1390e1403. [9] Z. Chen, D. Zhang, X. Wang, X. Jia, F. Wei, H. Li, Y. Lu, High-performance energy-storage architectures from carbon nanotubes and nanocrystal building blocks, Adv. Mater. 24 (2012) 2030e2036. k, D. Goers, M.E. Spahr, Carbon materials in lithium-ion batteries, in: [10] P. Nova guin, E. Fra˛ ckowiak (Eds.), Carbons for Electrochemical Energy Storage F. Be and Conversion Systems, CRC Press, Boca Raton, 2010, p. 300. [11] C.W. Kang, I. Lahiri, R. Baskaran, W.G. Kim, Y.K. Sun, W.B. Choi, 3-dimensional carbon nanotube for Li-ion battery anode, J. Power. Sources 219 (2012) 364e370. [12] H.G. Zhang, P.V. Braun, Three-dimensional metal scaffold supported bicontinuous silicon battery anodes, Nano. Lett. 12 (2012) 2778e2783. [13] S.J. Dillon, K. Sun, Microstructural design considerations for Li-ion battery systems, Curr. Opin. Solid. State. Mater. Sci. 16 (2012) 153e162. € m, K. Edstro €m, T. Wågberg, [14] T. Sharifi, M. Valvo, E. Gracia-Espino, R. Sandstro Hierarchical self-assembled structures based on nitrogen-doped carbon nanotubes as advanced negative electrodes for Li-ion batteries and 3D microbatteries, J. Power. Sources 279 (2015) 581e592. [15] L.F. Cui, L. Hu, J.W. Choi, Y. Cui, Light-weight free-standing carbon nanotubesilicon films for anodes of lithium ion batteries, ACS. Nano 4 (2010) 3671e3678. [16] D.T. Welna, L. Qu, B.E. Taylor, L. Dai, M.F. Durstock, Vertically aligned carbon nanotube electrodes for lithium-ion batteries, J. Power. Sources 196 (2011) 1455e1460. [17] R. Moshtev, B. Johnson, State of the art of commercial Li ion batteries, J. Power. Sources 91 (2000) 86e91. [18] Q. Xiao, Y. Fan, X. Wang, R.A. Susantyokoa, Q. Zhang, A multilayer Si/CNT coaxial nanofiber LIB anode with a high areal capacity, Energy. Environ. Sci. 7 (2014) 655e661. [19] J.K. Park, Materials for lithium secondary batteries, in: J.K. Park (Ed.), Principles and Applications of Lithium Secondary Batteries, Wiley-VCH, Weinheim, 2012, pp. 92e118. [20] S. Osswald, M. Havel, Y. Gogotsi, Monitoring oxidation of multiwalled carbon nanotubes by Raman spectroscopy, J. Raman. Spectrosc. 38 (2007) 728e736. [21] M. Winter, K.C. Moeller, J.O. Besenhard, Carbonaceous and graphitic anodes, in: G.A. Nazri, G. Pistoia (Eds.), Lithium Batteries : Science and Technology, Springer, New York, 2004, p. 177. [22] I. Lahiri, S.W. Oh, J.Y. Hwang, S.J. Cho, Y.K. Sun, R. Banerjee, W.B. Choi, High capacity and excellent stability of lithium ion battery anode using interfacecontrolled binder-free multiwall carbon nanotubes grown on copper, ACS. Nano 4 (2010) 3440e3446. [23] B. Marinho, M. Ghislandi, E. Tkalya, C.E. Koning, G.D. With, Electrical conductivity of compacts of graphene, multi-wall carbon nanotubes, carbon black, and graphite powder, Powder. Technol. 221 (2012) 351e358. [24] Y. Rachid, Thermodynamics of electrode materials for lithium-ion batteries, in: K. Ozawa (Ed.), Lithium Ion Rechargeable Batteries, Wiley-VCH, Weinheim, 2009, pp. 67e86. [25] R.A. Huggins, Negative electrodes in lithium cells, in: R.A. Huggins (Ed.), Advanced Batteries : Materials Sciences Aspects, Springer, New York, 2010, pp. 134e135. [26] A. Shah, M.N. Ates, S. Kotz, J. Seo, M. Abraham, S. Somu, A. Busnaina, A layered carbon nanotube architecture for high power lithium ion batteries, J. Electrochem. Soc. 161 (6) (2014) A989eA995. [27] L. Hu, F.L. Mantia, H. Wu, X. Xie, J. McDonough, M. Pasta, Y. Cui, Lithium-ion textile batteries with large areal mass loading, Adv. Energy Mater. 1 (2011) 1012e1017. [28] J.F. Fauvarque, P. Simon, Principles of electrochemistry and electrochemical guin, E. Fra˛ ckowiak (Eds.), Carbons for Electrochemical methods, in: F. Be Energy Storage and Conversion Systems, CRC Press, Boca Raton, 2010, pp. 7e16. [29] S. Jeong, J.P. Lee, M. Ko, G. Kim, S. Park, J. Cho, Etched graphite with internally grown Si nanowires from pores as an anode for high density Li-ion batteries, Nano Lett. 13 (2013) 3403e3407. [30] S. Yang, J. Huo, H. Song, X. Chen, A comparative study of electrochemical properties of two kinds of carbon nanotubes as anode materials for lithium ion batteries, Electrochim. Acta 53 (2008) 2238e2244. guin, F. Chevallier, C. Vix-Guterl, S. Saadallah, V. Bertagna, J.N. Rouzaud, [31] F. Be E. Frackowiak, Correlation of the irreversible lithium capacity with the active surface area of modified carbons, Carbon 43 (2005) 2160e2167. [32] J.Y. Song, H.H. Lee, Y.Y. Wang, C.C. Wan, Two- and three-electrode impedance spectroscopy of lithium-ion batteries, J. Power. Sources 111 (2002) 255e267. [33] L. Qie, W.-M. Chen, Z.-H. Wang, Q.-G. Shao, X. Li, L.-X. Yuan, X.-L. Hu, W.X. Zhang, Y.-H. Huang, Nitrogen-doped porous carbon nanofiber webs as anodes for lithium-ion batteries with a superhigh capacity and rate capability, Adv. Mater. 24 (2012) 2047e2050. [34] S.Y. Chew, S.H. Ng, J. Wang, P. Nov ak, F. Krumeich, S.L. Chou, J. Chen, H.K. Liu, Flexible free-standing carbon nanotube films for model lithium-ion batteries, Carbon 47 (2009) 2976e2983. [35] J. Landers, G.Y. Gor, A.,V. Neimark, Density functional theory methods for characterization of porous materials, Colloids. Surf. A 437 (2013) 3e32.