Nanostructured Silicon Anodes for Lithium Ion Batteries
Lithium batteries
Nanostructured Silicon Anodes for Lithium Ion Rechargeable Batteries Ranganath Teki, Moni K. Datta, Rahul Krishnan, Thomas C. Parker, Toh-Ming Lu, Prashant N. Kumta, and Nikhil Koratkar*
Rechargeable lithium ion batteries are integral to today’s information-rich, mobile society. Currently they are one of the most popular types of battery used in portable electronics because of their high energy density and flexible design. Despite their increasing use at the present time, there is great continued commercial interest in developing new and improved electrode materials for lithium ion batteries that would lead to dramatically higher energy capacity and longer cycle life. Silicon is one of the most promising anode materials because it has the highest known theoretical charge capacity and is the second most abundant element on earth. However, silicon anodes have limited applications because of the huge volume change associated with the insertion and extraction of lithium. This causes cracking and pulverization of the anode, which leads to a loss of electrical contact and eventual fading of capacity. Nanostructured silicon anodes, as compared to the previously tested silicon film anodes, can help overcome the above issues. As arrays of silicon nanowires or nanorods, which help accommodate the volume changes, or as nanoscale compliant layers, which increase the stress resilience of silicon films, nanoengineered silicon anodes show potential to enable a new generation of lithium ion batteries with significantly higher reversible charge capacity and longer cycle life.
[] Prof. N. Koratkar Department of Mechanical, Aerospace, & Nuclear Engineering Rensselaer Polytechnic Institute, Troy, NY 12180 (USA) E-mail:
[email protected] R. Teki Department of Chemical & Biological Engineering Rensselaer Polytechnic Institute, Troy, NY 12180 (USA) R. Krishnan Department of Materials Science & Engineering Rensselaer Polytechnic Institute, Troy, NY 12180 (USA) T. C. Parker, T.-M. Lu Department of Physics, Applied Physics, & Astronomy Rensselaer Polytechnic Institute, Troy, NY 12180 (USA) M. K. Datta, P. N. Kumta Department of Mechanical Engineering & Materials Science and Chemical & Petroleum Engineering, Bioengineering University of Pittsburgh, Pittsburgh, PA 15260 (USA) DOI: 10.1002/smll.200900382 small 2009, x, No. x, 1–7
Final page numbers not assigned
Keywords:
lithium ion batteries nanorods nanostructured anodes oblique angle deposition silicon
1. Introduction Rechargeable lithium ion batteries are integral to the portable, entertainment, computing, and telecommunication equipment required by today’s information-rich, mobile society. They are currently one of the most popular types of battery for portable electronics, with one of the best energy-toweight ratios, suffering virtually no memory effect and a slow discharge rate when not in use.[1] Although originally intended for consumer and portable electronics, lithium ion batteries are now growing in popularity for use in next-generation wireless communication devices, plug-in hybrid electric vehicles, stationary storage batteries, microchips, defense applications, and even in all electric vehicles. Li-based rechargeable batteries were first proposed in the early 1960 s and since then the battery has undergone several transformations. Initially, metallic Li was used as the anode but it posed serious safety hazards because of dendritic Li growth
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
concepts
N. Koratkar et al.
during cycling.[2] It was discovered that pyrolytic carbon can insert Li quite effectively[3] and Sony introduced the first commercial C/LiCoO2-based Li ion battery in 1991.[4] Research is still ongoing to tailor Li ion batteries for highenergy applications (e.g., modern communication devices), for high-power applications (e.g., plug-in hybrid and electric vehicles), and long-cycle-life applications (e.g., uninterrupted power sources).[5] Li ion batteries have a higher energy density than most other types of rechargeable battery due to Li being one of the most electropositive elements ( 3.04 V versus standard hydrogen electrode) as well as the lightest electrochemically active metal. This means that for their size or weight they can store more energy than other rechargeable batteries. They also operate at higher voltages in comparison to other hitherto wellknown rechargeable systems, typically about 3.7 V for Li ion versus 1.2 V for nickel–metal hydride (NiMH) or nickel– cadmium (NiCd). They also have a lower self-discharge rate than other types of rechargeable battery. On the flip side, Li ion batteries are more complex to manufacture and hence more expensive than similar capacity NiMH or NiCd batteries due to the complexity of the anode and cathode material systems as well as the stringent requirements on the non-aqueous electrolytes needed for efficient functioning. The three primary functional components of a Li ion battery (Figure 1) are the anode, cathode, and electrolyte, for which a variety of materials may be used. Commercially, the most popular material for the anode is graphite or carbon.[6] The cathode generally falls into three categories: layered oxides such as LiCoO2, transition metal phosphates such as LiFePO4, or spinels such as LiMn2O4.[7] Liquid electrolytes in Li ion batteries consist of solid lithium–salt electrolytes, such as LiPF6, LiBF4, or LiClO4, in an organic solvent such as alkene carbonates or mixtures of carbonates, amides, or imides. The choice of material for the anode, cathode, and electrolyte combined with their physical, chemical, electronic, and electrochemical attributes affects the voltage, capacity, life, and safety of a Li ion battery. Both the anode and cathode are materials into which the ionized Li ions are inserted and
Figure 1. Schematic diagram of a conventional Li ion battery showing the anode, cathode, and electrolyte. Li ions are extracted from the cathode and inserted into the anode during charging. The process is reversed during discharging with Li ions being extracted from the anode and inserted into the cathode.
2 www.small-journal.com
extracted. The nature of the insertion and extraction can vary from electrochemical intercalation in the case of layered oxides and graphite causing changes in the crystallographic and electronic structure to alloy formation in the case of tin and silicon. When the cell is being charged, Li ions are extracted from the cathode, transported through the Li ion conducting electrolyte, and inserted into the anode. While discharging the Li ions are extracted from the anode and inserted into the cathode while again migrating through the electrolyte under an opposite electrochemical potential gradient. The performance of the anode and cathode are typically quantified in terms of their charge capacity per unit weight of the materials comprising the electrode.
2. Silicon as Li Ion Battery Anode Silicon is one of the most promising anode materials because it has the highest-known theoretical charge capacity of 4200 mAh g 1, which is more than ten times that of existing graphite anodes (which have a capacity of 372 mAh g 1 corresponding to the formation of LiC6), and various other oxide and nitride materials.[8] Silicon is also the second most abundant element on earth. Because of these attributes, a great deal of attention has been given to using Si as a Li ion cell anode material. However, Si anodes have limited applications because of the huge volume change (order of 400%) associated with the insertion and extraction of Li (each Si atom can accommodate 4.4 Li atoms leading to formation of Li22Si5 alloy).[9] The stresses induced by these volume changes cause cracking and pulverization of the Si anode, which leads to loss of electrical contact and eventual fading of capacity. This is illustrated in Figure 2a, which shows the specific capacity of amorphous Si film anodes of thickness 1 mm and 250 nm.[10] The films exhibit near theoretical capacity for a limited number of cycles after which the capacity begins to fade dramatically. Figure 2b and c shows the stress-induced break up of the film into smaller islands, causing pile up of these islands to form pillars, which eventually delaminate (i.e., peel off the underlying electrode), resulting in loss of electrical contact. Ohara et al. tested a Si film deposited on a Ni substrate.[11] Ni develops a passivating layer that acts as a good binding agent between the substrate and the Si film due to the strong affinity of Si to oxygen. Such films were shown to have capacities as high as 1700–2200 mAh g 1 for 750 cycles at a 2 C charge/discharge rate. Some other approaches used to overcome the issues with Si anodes include pure Si micro- and nanoscale powder anodes, Si dispersed in an active/inactive matrix, Si mixed with different binders, and Si thin films.[12] Bulk Si anodes fabricated from micrometer-sized Si particles show a large irreversible capacity and poor capacity retention, mainly due to loss of electrical conductivity from the large volume expansion. Generation of nanocomposites comprising a nanocrystalline Si powder coated with a thin layer of amorphous carbon serving as an interfacial adhesion layer, together embedded in a mechanically softer but more ductile carbon matrix that is relatively inactive in the electrochemical potential window of 0.02–1.2 V, has led to moderate success of sustaining reversible capacities on the order of 700 mAh g 1 at a C rate of C/4.[13,14] Carbon-coated
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2009, x, No. x, 1–7
Final page numbers not assigned
fading by maintaining electrical contact between the silicon particles. Similarly, Ng et al. have obtained a reversible capacity of 1450 mAh g 1 with spheroidal carbon-coated silicon nanocomposite anode structures at a cycling rate of 100 mAh g 1.[16] Incorporation of carbon nanotubes have led to reversible capacities as high as 1000 mAh g 1[17] under a constant current density of 250 mA cm 2. However, it has been reported[18] that nanometer-sized Si particles in composites tend to agglomerate after the insertion/extraction of Li ions and this size increase results in poor Li insertion/extraction kinetics. Si micropillar array anodes (0.58 mm in diameter and 0.81 mm in height) have also been tested. Despite their reasonable capacity retention, they exhibited relatively low Faradaic efficiency through 50 cycles, which would limit their practical utility.[19] It should also be noted that in order to accommodate such Si anodes in practical cells, it is important that their capacities match that of the commercial standard cathodes such as LiCoO2. Yin et al. showed that Si films of thickness greater than 4 mm provide geometric capacities of 2.6 mAh cm 2, which match that of the cathode (2 mAh cm 2).[20] These films were grown by electron-beam deposition on a Cu substrate with a concave–convex surface. As a result, the thick Si layer was reported to have good adhesion to Cu, providing capacity matching over a long cycle life.
3. Use of Si Nanowires
Figure 2. Characterization of amorphous Si films as the battery anode. a) Specific capacity plotted as a function of cycle number. b) Stress-induced cracking of the film after a few cycles. c) Delamination and peeling of the film from the collector electrode after extended cycling. Reprinted with permission from Reference [10]. Copyright 2006, The Electrochemical Society.
silicon particles have also been shown to achieve capacities as high as 1000 mAh g 1 when charging and discharging at a constant current of 0.3 mA mg 1.[15] The carbon coating enhances structural stability and also prevents local capacity small 2009, x, No. x, 1–7
Final page numbers not assigned
Cui and co-workers recently published a seminal paper[8] that indicated that anodes comprised of Si nanowires (Figure 3a,b) were able to accommodate large strain without pulverization, provide good electronic contact, and showed short Li conduction distances. They achieved the theoretical charge capacity for Si anodes and maintained a discharge capacity close to 75% of the maximum, with little fading of capacity at a discharge rate of 0.05 C (Figure 3c). The Si nanowires were grown using a vapor–liquid–solid (VLS) process directly on stainless steel substrates, so that each nanowire was electrically connected to the metallic current collector. The nanowires averaging 89 nm in diameter allowed for the large volume changes without pulverization or loss of contact. The diameter of the nanowires was observed to increase on cycling (the average diameter increased to 141 nm) and the nanowires showed a drastic change in their atomic structure; the initially crystalline Si nanowires underwent a gradual transformation to amorphous LixSi. Similar crystalline-to-amorphous Si phase transitions upon reaction with Li have been reported by Kumta and co-workers and other researchers as well.[22] The main mechanism behind the stress alleviation appears to be that the nanowire array provides sufficient space between adjacent nanowires to accommodate the volume change associated with alloying and de-alloying of Li. Another novel feature of the nanowire architecture is that each individual nanowire is directly connected to the currentcollecting substrate thus enabling a robust electrical contact to be maintained. In addition to this, nanomechanics could also play an important role in preventing the fracture of the Si nanowires. This is because in general nanostructures can undergo superplastic deformation and thus endure larger
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
3
concepts
N. Koratkar et al.
stresses without failure compared to the equivalent material synthesized in the bulk because the dimensions of the nanostructure limit the size of the precursor crack, which typically initiates the fracture process. The fracture processes in
Figure 3. Testing of Si nanowires as the battery anode. a) Concept schematic of Si nanowire electrode assembled on the current collector. b) Scanning electron microscopy (SEM) image of Si nanowires that comprise the device anode. c) Capacity versus cycle number for various electrode configurations. The Si nanowires show stable capacity (3500 mAh g 1) without any degradation with increase in the number of cycles. Reprinted with permission from Reference [8]. Copyright 2008, Nature Publishing Group.
4 www.small-journal.com
nanocrystalline materials will therefore be mainly governed by the nanometer-sized grain-boundary dynamics, which increases the toughness and the stress to failure. In addition other phenomena such as the nature of the phase transitions at specific voltages, the conversion of the nanocrystalline and/or crystalline/microcrystalline Si (nc-Si) to amorphous Si (a-Si) and the adhesion effects at the nanoscale could also contribute to the cycling stability, and these aspects need to be investigated in depth. Cui and co-workers have also studied various other forms of nanowire electrode. They used germanium nanowires[23] instead of Si, since the room-temperature diffusivity of Li in Ge is 400 times that of Si, making it an attractive anode for highpower applications. They explored the use of spinel LiMn2O4 nanorods[24] as Li ion battery cathodes and found that these structures have a higher charge-storage capacity at high power rates compared to commercially available powders. In another interesting work the same group fabricated crystalline– amorphous core–shell Si nanowire-based anodes.[25] Such electrode architecture allows the amorphous shells to be selectively electrochemically active (Kumta and co-workers[22] have shown that amorphous Si is better able to withstand pulverization during cycling as compared to crystalline Si), while the crystalline Si core acts as a stable mechanical support and provides an efficient electrical conducting pathway. These Si nanowire-based core–shell anodes exhibited a high chargestorage capacity three times that of carbon with a 90% capacity retention over 100 cycles, and an excellent electrochemical performance. Shimizu et al. have grown high-density epitaxial nanowire arrays by using VLS growth in a porous anodic aluminum oxide (AAO) template.[21] The use of AAO allows for great flexibility in the design of the nanowire diameter and density. Another very interesting approach was adopted by Cho and co-workers, where they fabricated 3D porous bulk Si particles (Figure 4a) with a pore wall size of 40 nm.[26] These nanosilicon sponges were able to accommodate large strains without pulverization even after 100 cycles and maintained a charge capacity of greater than 2800 mAh g 1 throughout at a cycling rate of 1 C (Figure 4b). The unique shape and structure of these particles not only allows for faster transport of the Li ions through the electrolyte and the electrode but also guarantees faster intercalation reactions of the Li ions, thus resulting in a large specific capacity even when operated at high charge–discharge currents. It should be noted that most studies with nanstructured Si electrodes in Li ion batteries have been limited to fewer than 100 charge–discharge cycles. In order to demonstrate commercial viability, testing with varying current rates, estimation of irreversible loss and coulombic efficiency, and evaluation of half cells and full cells using known cathode material systems over significantly larger number of cycles is required. Moreover, it is also important to investigate the dependence of the capacity on the charge/discharge rate, since the stress build up and its relaxation during the rest period and charging and discharging processes are known to be influenced by the charge–discharge rates. This has already been reported by Kumta and co-workers in their publications related to amorphous Si nanolayers.[10,35] Furthermore, variation in electronic and ionic conductivities of the active material plays
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2009, x, No. x, 1–7
Final page numbers not assigned
Figure 4. Testing of porous Si particles with nanometer-scale wall thickness. a) SEM images of the porous Si particles indicating a pore wall size of 40 nm. b) These nanoporous Si structures maintained a charge capacity greater than 2800 mAh g 1 for more than a hundred cycles. Reprinted from Reference [26].
in Figure 5a. This results in the formation of isolated nanorods by the self-shadowing effect during growth. The shadowing effect is a characteristic of oblique-angle deposition and occurs when regions of relatively greater initial growth rate obstruct the flow of incident flux to other areas, causing the morphology to deviate from the smooth and regular surface that results from depositions performed at normal incidence. The shadowing effect can be controlled by adjusting the deposition rate, incidence angle, and the substrate rotation speed and in this way three-dimensional nanostructures with very large aspect ratio and controllable porosity, shape, and symmetry can be deposited using OAD without the need for expensive lithography and multistep processes. OAD allows for a convenient, inexpensive, and scalable way to fabricate such Si nanorod arrays that can accommodate the volume change during Li cycling. Compared to solutionbased growth techniques, PVD methods such as sputtering and e-beam evaporation are clean, repeatable, have high deposition rates, and can be grown over large areas with good adhesion to the substrate. They also allow for greater control over the porosity, length, and dimensions of the nanorod arrays. There is also interest in using doped silicon as anode materials due to their improved electronic conductivity and OAD provides a convenient method to grow doped silicon nanostructures. Recently, Brett and co-workers used the above technique of OAD (also called glancing-angle deposition (GLAD)) to grow vertical Si nanocolumns and tested them as anodes for Li ion batteries.[31] Their anodes showed capacities of around 3600 mAh g 1 and good capacity retention (83%) after
an extremely important role in stabilizing the capacities of the electrode during the different C rates employed. The nanostructure directly impacts the transport characteristics of electrons and lithium ions together affecting the charge-transfer impedance. The charge-transfer resistance in turn influences the kinetics of lithium ion extraction and alloying during insertion, which has a strong effect on the first cycle irreversible loss and those accompanying subsequent cycles, the volumetric stress, relaxation, and eventual cracking of the nanostructures. Studying the impact of C rates on the structure and microstructure of the electrodes is therefore extremely important.
4. Si Nanorods by Oblique-Angle Deposition Another direct approach for patterning the nanostructured surfaces is based on oblique-angle deposition (OAD)[27–30] with substrate rotation. OAD is a physical vapor deposition (VPD) technique in which flux arrives at a large oblique incidence angle (>808) from the substrate normal (while the substrate is rotating), as shown schematically small 2009, x, No. x, 1–7
Final page numbers not assigned
Figure 5. Si nanorod anodes developed by OAD. a) Schematic diagram showing the OAD process with substrate rotation. b) Testing of Si nanorods grown by OAD; the nanorods display a stable capacity of 1600 mAh g 1, which is more than four times greater than graphite electrodes. There is an initial loss in the charge and discharge capacities for the first few cycles, which is indicative of probable wettability issues between the electrode and the electrolyte. c) SEM image of an amorphous Si nanospring array grown by OAD. The nanosprings are highly compliant along the axial direction.
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
5
concepts
N. Koratkar et al.
70 charge–discharge cycles. Experiments performed by our group with Si nanorods deposited by OAD on Cu foils also show a highly stable capacity that is more than fourfold better than graphite anodes (Figure 5b). The initial few cycles show some loss in charge and discharge capacities, probably due to the formation of a solid electrolyte interphase that has a high electronic resistivity as well as wettability issues between the electrode and the electrolyte and possible oxidation of the underlying Si. With continued cycling, these nanorods display stable capacities of 1600 mAh g 1. The OAD technique can also be used to fabricate Si nanospring architectures (Figure 5c), which cannot be achieved using VLS growth methodologies. Our group has demonstrated that such Si nanosprings are highly compliant[32] and are therefore expected to offer enhanced stress resilience.
5. Nano-Compliant Supports Previous studies utilizing bulk Si and micrometer-sized Si particles as electrodes in Li ion batteries have shown capacity fading and short battery lifetime due to pulverization and loss of electrical contact between the active material and the current collector. Anodes made of very thin films (few tens of nanometers in thickness) of amorphous Si display a stable capacity over many cycles but have insufficient material for a viable battery and for proper capacity matching with the cathode. Therefore, if thicker Si films can be developed and their stress resistance increased, the resulting anodes may potentially exhibit stable capacity without pulverization. One approach to forming stress-resistant Si thin films is by using a nanostructured compliant layer (NCL).[33,34] The NCL consists of slanted nanorods grown by OAD and it is sandwiched between the film and the substrate (Figure 6a). The technique is all in situ, does not require any lithography steps, and the nanostructured layer can be made from the same material as the deposited film (e.g., Si as in Figure 6b). By using this approach our group has shown[33] that stress values can be reduced by approximately one order of magnitude in sputterdeposited tungsten films (Figure 6c). The mechanisms[34] that are responsible for stress alleviation are 1) to delay the inception of stress buildup and 2) to toughen the interface such that delamination buckling occurs at higher levels of strain energy stored in the film (larger film thickness). In this way, the NCL relieves the stress in the continuous film and effectively improves the adhesion, which results in a larger critical thickness for delamination buckling and higher-quality films. The NCL would therefore allow the growth of much thicker Si films than normally possible, and would also increase the structural stability of the films. In fact, the process can be repeated to form multilayer (or stacked) films with very large thickness.[33,34] Such thicker Si films supported on the NCL will likely provide a greater amount of material and also better absorb the strain caused by Li insertion/extraction. One of the key issues concerning Si films is substrate adhesion. Kumta and co-workers have shown that even though amorphous Si films exhibit near theoretical capacity for a limited number of cycles, they eventually delaminate.[9] They suggested that formation of amorphous Cu–Si phases leading to
6 www.small-journal.com
Figure 6. Nanocompliant support structures for Si film anodes. a) Schematic image of the NCL concept showing a film supported by an array of inclined nanorods. b) Fabrication of a Si film supported on an NCL of inclined Si nanorods by OAD with substrate rotation. c) Reduction in film stress caused by the NCL. Data shows up to an order of magnitude reduction in stress depending on the film thickness. The data is for a tungsten film supported on a tungsten NCL. Figure 6c reprinted with permission from Reference [33]. Copyright 2005, AVS.
a weakening of the interfacial adhesion between Si and Cu was partly responsible for the observed delamination behavior. They proposed that an interfacial layer is required (between the Si film and Cu substrate) with a very low modulus to mitigate the volumetric-change-induced stresses in the Si layer, and this interfacial layer should also act as an effective diffusion barrier to prevent the formation of Cu–Si phases. The NCL proposed here can serve these functions. It provides a highly compliant support structure, which relieves the stress. Moreover, by constructing the NCL from a material such as chromium, one could significantly enhance adhesion between Si and Cu and also prevent the diffusion of Si into the underlying Cu substrate.
6. Conclusions Si anodes offer a more than 10-fold increase in theoretical charge capacity compared to graphite anodes but their full potential has so far not been realized commercially due to pulverization and fatigue of the electrode arising from stresses generated by the huge volume change associated with the insertion and extraction of lithium. Several published papers utilizing nanometer-sized silicon as well as nanocomposites of nanostructured Si and carbon including carbon nanotubes have demonstrated the ability to achieve capacities in excess of
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2009, x, No. x, 1–7
Final page numbers not assigned
1000 mAh g 1 while microcrystalline Si is known to fail within a few cycles. These concepts clearly have promise; however, there is still substantial research to be conducted to obtain a full understanding of their impact. Another approach to enhancing the stress resiliency of Si anodes is by the use of porous nanowire/nanorod architectures. These can range from Si nanowires grown by the VLS process to Si nanorods grown by OAD to 3D porous Si particles with nanometer-scale wall thickness. Another concept is to support thicker Si films on nanocompliant supporting structures that delay the inception of stress buildup, and toughen the interface such that delamination buckling is prevented. While all of the above approaches show potential to significantly enhance the charge capacity and cycle life of Si anodes in Li ion batteries, there are several critical scientific issues that must be addressed before this technology can be developed for use in commercial applications. First, an in-depth understanding of the mechanism of stress build-up in the nanostructure during lithiation and its relaxation during the rest period and the correlation of the stress generated with the accompanying phase transitions during the alloying and reverse de-alloying processes are essential to predict how the fracture of the active material can be minimized and completely avoided in the nanostructures. Second, it is important to establish the dependence of the capacity, and the electronic and lithium ion transport on the charge and discharge rates. Finally, extended cycling of the electrode under realistic conditions is required to demonstrate commercial viability. In addition, estimation of irreversible loss and coulombic efficiency, and evaluation of half cells and full cells using known cathode material systems over significantly larger number of cycles is necessary.
Acknowledgements Funding support from the United States National Science Foundation (NIRT Award 0506738) to N.K. and T.M.L. is acknowledged. P.N.K. acknowledges the support of the DOEBATT program and the Edward R. Weidlein Chair Professorship funds for partial support of this research.
[1] J.-M. Tarascon, M. Armand, Nature 2001, 414, 359. [2] M. Broussely, P. Biensan, B. Simon, Electrochim. Acta 1999, 45, 3. [3] M. Mohri, N. Yanagisawa, Y. Tajima, H. Tanaka, T. Mitate, S. Nakajima, M. Yoshida, Y. Yoshimoto, T. Suzuki, H. Wada, J. Power Sources 1989, 26, 545. [4] T. Nagaura, K. Towaza, Prog. Batt. Solar Cells 1990, 9, 209. [5] Y.-G. Guo, J.-S. Hu, L.-J. Wan, Adv. Mater. 2008, 20, 2878.
small 2009, x, No. x, 1–7
Final page numbers not assigned
[6] M. Endo, C. Kim, K. Nishimura, T. Fujino, K. Miyashita, Carbon 2000, 38, 183. [7] M. S. Whittingham, Chem. Rev. 2004, 104, 4271. [8] C. K. Chan, H. Peng, G. Lin, K. McIlwrath, X. F. Zhang, R. A. Huggins, Y. Cui, Nat. Nanotechnol. 2008, 3, 31. [9] W. Wang, M. K. Datta, P. N. Kumta, J. Mater. Chem. 2007, 17, 3229. [10] J. P. Maranchi, A. F. Hepps, A. G. Evans, N. T. Nuhfer, P. N. Kumta, J. Electrochem. Soc. 2006, 153, A1246. [11] S. Ohara, J. Suzuki, K. Sekine, T. Takamura, J. Power Sources 2003, 119, 591. [12] U. Kasavajjula, C. Wang, A. J. Appleby, J. Power Sources 2007, 163, 1003. [13] M. Datta, P. N. Kumta, J. Power Sources 2006, 158, 557. [14] M. K. Datta, P. N. Kumta, J. Power Sources 2007, 165, 378. [15] W. Liu, J. Wang, H. Wu, D. Shieh, M. Yang, N. Wu, J. Electrochem. Soc. 2005, 152, A1719. [16] S. H. Ng, J. Wang, D. Wexler, K. Konstantinov, Z. P. Guo, H. K. Liu, Angew. Chem. Int. Ed. 2006, 45, 6896. [17] W. Wang, P. N. Kumta, J. Power Sources 2007, 172, 650. [18] H. Li, X. Huang, L. Chen, G. Zhou, Z. Zhang, D. Yu, Y. J. Mo, N. Pei, Solid-State Ionics 2000, 135, 181. [19] M. Green, E. Fielder, B. Scrosati, M. Wachtler, J. S. Moreno, Electrochem. Solid-State Lett. 2003, 6, A75. [20] J. Yin, M. Wada, K. Yamamoto, Y. Kitano, S. Tanase, T. Sakai, J. Electrochem. Soc. 2006, 153, A472. [21] T. Shimizu, T. Xie, J. Nishikawa, S. Shingubara, S. Senz, U. Go¨sele, Adv. Mater. 2007, 19, 917. [22] J. Nanda, M. Datta, J. T. Remillard, A. O’Neill, P. N. Kumta, Electrochem. Commun. 2009, 11, 235. [23] C. K. Chan, X. F. Zhang, Y. Cui, Nano Lett. 2008, 8, 307. [24] D. K. Kim, P. Muralidharan, H.-W. Lee, R. Ruffo, Y. Yang, C. K. Chan, H. Peng, R. A. Huggins, Y. Cui, Nano Lett. 2008, 8, 3948. [25] L.-F. Cui, R. Ruffo, C. K. Chan, H. Peng, Y. Cui, Nano Lett. 2009, 9, 491. [26] H. Kim, B. Han, J. Choo, J. Cho, Angew. Chem. 2008, 47, 1. [27] K. Robbie, J. C. Sit, M. J. Brett, J. Vac. Sci. Technol. B. 1998, 16, 1115. [28] Y.-P. Zhao, D.-X. Ye, G.-C. Wang, T.-M. Lu, Nano Lett. 2002, 2, 351. [29] C. Li, Z. Wang, P.-I. Wang, Y. Peles, N. Koratkar, G. P. Peterson, Small 2008, 4, 1084. [30] R. Teki, T. Karabacak, T.-M. Lu, N. Koratkar, Appl. Phys. Lett. 2006, 89, 193116. [31] M. D. Fleischauer, J. Li, M. J. Brett, J. Electrochem. Soc. 2009, 156, A33. [32] D. L. Liu, D. X. Ye, F. Khan, F. Tang, B. K. Lim, R. C. Picu, G.-C. Wang, T.-M. Lu, J. Nanosci. Nanotechnol. 2003, 3, 492. [33] T. Karabacak, J. J. Senkevich, G.-C. Wang, T.-M. Lu, J. Vac. Sci. Technol. A 2005, 23, 986. [34] T. Karabacak, R. C. Picu, J. J. Senkevich, G.-C. Wang, T.-M. Lu, J. Appl. Phys. 2004, 96, 5740. [35] J. P. Maranchi, A. F. Hepp, P. N. Kumta, Electrochem. Solid-State Lett. 2003, 6, A198–A201.
ß 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Received: March 3, 2009 Revised: July 16, 2009 Published online:
www.small-journal.com
7