Three-Dimensionally Engineered Porous Silicon Electrodes for Li Ion

Oct 31, 2012 - ABSTRACT: The ultimate goal of Li ion battery design should consist of fully accessible metallic current collectors, possibly of nanosc...
0 downloads 0 Views 4MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Three-Dimensionally Engineered Porous Silicon Electrodes for Li Ion Batteries Sanketh R. Gowda,† Victor Pushparaj,∥ Subramanya Herle,*,∥ G. Girishkumar,∥ Joseph G. Gordon,∥ Hemtej Gullapalli,‡ Xiaobo Zhan,‡ Pulickel M. Ajayan,*,‡,§ and Arava Leela Mohana Reddy*,‡ †

Department of Chemical and Biomolecular Engineering, ‡Department of Mechanical Engineering and Materials Science, and Department of Chemistry Rice University, Houston, Texas 77005, United States ∥ Advanced Technology Group, Applied Materials Inc., Santa Clara, California 94085, United States §

S Supporting Information *

ABSTRACT: The ultimate goal of Li ion battery design should consist of fully accessible metallic current collectors, possibly of nanoscale dimensions, intimately in contact with high capacity stable electrode materials. Here we engineer three-dimensional porous nickel based current collector coated conformally with layers of silicon, which typically suffers from poor cycle life, to form high-capacity electrodes. These binder/conductive additive free silicon electrodes show excellent electrode adhesion resulting in superior cyclic stability and rate capability. The nickel current collector design also allows for an increase in silicon loading per unit area leading to high areal discharge capacities of up to 0.8 mAh/cm2 without significant loss in rate capability. An excellent electrode utilization (∼85%) and improved cyclic stability for the metal/silicon system is attributed to reduced internal stresses/fracture upon electrode expansion during cycling and shorter ionic/ electronic diffusion pathways that help in improving the rate capability of thicker silicon layers. KEYWORDS: silicon, anode, Li Ion Battery, 3D, porous

O

morphologies for the Si anode and the use of novel polymeric binders to accommodate volume expansion and also improve adhesion of Si to the current collector.15−27 Nanostructured morphologies for silicon have been fabricated, and half cell Li ion battery measurements have indicated some improvement over the bulk Si electrodes.8,21−25 Recent interest has been generated by the nanowire morphology due to its resistance to severe structural fracture/ deformation upon cycling. Nanosized silicon particle composites with polymer binder/carbon additives have received considerable attention to mitigate electrode decripitation from the current collector.19−22 But the use of such composites (polymer/carbon based composites) could increase the complexity of electrode processing and increase electrode volume (low density additives/current collectors results in increased electrode thickness), thereby reducing both volumetric and gravimetric energy density. On the other hand, binder-free nanostructured Si electrodes are susceptible to decripitation from current collector upon cycling and hence show limited cycle life. There have been several approaches used in prior literature to overcome the severe capacity fade resulting from large volume expansion, for binder free Si

ver the past two decades, Li ion battery has emerged as the primary energy storage device for a wide range of portable electronic devices due to its high gravimetric and volumetric energy density.1−6 Recently the Li ion battery has gained renewed interest as a potential candidate to replace gasoline in vehicular applications.7 Electrodes for current Li ion batteries comprise of intercalation-based electrode systems like graphitic carbon anode and layered lithium metal oxide/olivine phosphate cathode.8−12 But the energy demands for the operation of vehicles far surpass the energy densities obtained from commercially existing Li ion battery electrode chemistries.7 There is also a need for the increase in energy delivered by thin film lithium ion batteries, which serve as energy sources for micro- and nanoelectronic devices.1,3 Hence extensive research efforts are underway in search of new battery electrodes to improve the energy density of the Li ion cell. Silicon is an attractive candidate as a negative electrode material for the Li ion battery due to its high theoretical capacity of 4200 mAh/g (Li22Si5 composition), over 10 times higher than existing carbon anode.8 An increase in capacity of negative electrode increases the overall energy density of the Li ion battery. But Si electrodes suffer from severe capacity fade upon cycling.13,14 The capacity fade is attributed to large volume change (∼400%; Li22Si5) associated with the reversible lithium insertion resulting in pulverizing/decrepitating Si from the current collector. Hence recent research efforts have focused on addressing these drawbacks by developing nanostructured © 2012 American Chemical Society

Received: June 4, 2012 Revised: September 24, 2012 Published: October 31, 2012 6060

dx.doi.org/10.1021/nl302114j | Nano Lett. 2012, 12, 6060−6065

Nano Letters

Letter

Figure 1. Structural and electrochemical characterization of type 1 700 nm 3D(Si,Ni) electrode. (a) Scanning electron microscope (SEM) image of type 1 porous nickel showing interconnected tubular porous network of nickel with an average pore diameter of ∼150 nm. (b) SEM image of 700 nm 3D(Ni,Si) obtained after deposition of 700 nm PET of Si on type 1 porous nickel film. The SEM image shows partial filling of pores. (c) Voltage profile of the 700 nm 3D(Si,Ni) vs Li/Li+ showing flat first discharge plateau and sloping voltage plateaus for subsequent cycles. (d) Cycling characteristics of 700 nm 3D(Si,Ni) at 1C showing a reversible specific capacity of 1650 mAh/g after 120 cycles of charge/discharge.

fabricate scalable tubular nickel current collectors with variable pore dimension and thickness suitable for thin film battery applications. Conformal amorphous silicon films of different mass loadings have been coated onto the tubular nickel substrates (3D(Si,Ni)) and tested in lithium half cells. The binder free 3D(Si,Ni) electrode has shown improved capacity per unit area and cyclic stability due to the porous structure and excellent electrode adhesion. A 3D(Si,Ni) negative electrode for Li ion batteries has been fabricated by (i) deposition of porous nickel films on stainless steel substrate by coelectrodepositing Cu−Ni film, (ii) selective etching of copper from microstructure of Cu−Ni films, and (iii) conformal coating of silicon films within pores of nickel by chemical vapor deposition process (Figure S1). Two types of porous Ni films (types 1 and 2) with different aspect ratios were fabricated by selective etching of the copper from coelectrodeposited Ni−Cu films (see Supporting Information, Table S1). For the fabrication of type 1 porous nickel, the Cu:Ni ion ratio in the electrolytic bath was maintained at 1:20, and a potential of −0.8 V was chosen to ensure biphasic codeposition of Ni and Cu with a core−shell microstructure.34−36 The pH of the electrolytic bath was maintained at ∼4.1 to preclude the competing hydrogen evolution reaction at the surface of the stainless steel substrate. After Cu−Ni film deposition, the copper was selectively etched from the Cu−Ni film, to obtain a highly porous Ni film of thickness ∼1−2 μm (Figure 1a). Our experiments showed that increasing Cu−Ni deposition times did not increase the thickness of porous nickel

negative electrodes, by (i) spacing the nano-Si structures and (ii) controlling the surface morphology of the current collector/substrate for better Si adhesion.13 Research efforts involving thin film Si electrodes have been conducted to improve cycle life and Si electrode adhesion to current collector substrate. The Si films have been deposited using chemical vapor deposition (CVD) or physical vapor deposition (PVD) techniques with typical silicon film thickness ≤500 nm on various textured current collectors.13,28−30 Porous Ni was also employed as a current collector reported earlier, for example, amorphous Si (1.2 μm) deposited on porous Ni showed capacity up to 1000 mAh/g,30 but the capacity retention was rather poor. Recently Wang et al. reported Si−Ni nanorod structures and 300 nm Si was deposited by plasma-enhanced vapor deposition on well-separated Ni nanocones with a mean bottom size of ∼500 nm and the heights of around 600 nm.31 Although good cycle performance and rate capability were reported for the Si thin films, these films have low Si loading per unit area and hence the areal capacity is relatively low. There is a lot of interest in studying thicker Si films (>500 nm) to increase the capacity of the Si electrode and maintain good cycle life. To increase Si film thickness beyond 500 nm a few reports have looked at roughening the current collector substrates.32,33 Hence there exists a need for the clear understanding and development of stable binder-free Si electrodes that could lead to significant improvements in energy densities and cycling characteristics of silicon based lithium ion batteries. Herein, we demonstrate the ability to 6061

dx.doi.org/10.1021/nl302114j | Nano Lett. 2012, 12, 6060−6065

Nano Letters

Letter

Figure 2. Capacity per unit area as a function of C-rate and mass loading for type 1 3D(Si,Ni) electrodes. (a) Cycling characteristics of 500 nm 3D(Si,Ni) at 0.8C, showing a reversible capacity of 0.18 mAh/cm2 after 75 cycles of charge/discharge. (b) Discharge capacity after 25 cycles vs Crate for 500 nm 3D(Si,Ni) electrode. (c) Cycling characteristics of 1 μm 3D(Si,Ni) at 0.9C showing a reversible capacity of 0.42 mAh/cm2 after 70 cycles of charge/discharge. (d) Discharge capacity after 60 cycles vs planar equivalent thickness of a-Si plot for type 1 3D(Si,Ni) electrode.

been plotted as a function of cycle number (Supporting Information, Figures S4 and S5). On observation of the cycling data it can be seen that there is a large irreversible loss observed after the first cycle in the discharge capacity as expected. The irreversible loss in the first cycle can be attributed to the SEI formation due electrolyte decomposition on the surface of the silicon electrode. Stable reversible discharge capacities of 7, 30, and 50 μAh/cm2 were observed for the 10, 40, and 100 nm cases, respectively. A capacity loss of around 30% was observed in the discharge capacity from the second cycle until the 100th cycle for 10, 40, and 100 nm Si thicknesses (Supporting Information, Figure S5). Yet the discharge capacity per unit area observed for the cases of 10, 40, and 100 nm thickness is very low due to the low Si mass loading. Hence we fabricated 3D(Si,Ni) electrodes with Si PET greater than 500 nm to increase the capacity per unit area. Figure 1b shows the SEM image of Si coated type 1 porous nickel with silicon PET of 700 nm. The 700 nm type 1 3D(Si,Ni) was tested in lithium half cells for its electrochemical performance. Figure 1c and d show the results of galvanostatic charge/discharge measurements conducted at a current rate of 1C (C-rate is defined as the time of discharge for the stable reversible capacity) in CR2032 coin cells. From the voltage profile of the 700 nm 3D(Si,Ni) a flat discharge plateau can be observed for the first discharge whereas the subsequent charge/discharge cycles show sloping charge and discharge profiles. Hence we can conclude that the 700 nm 3D(Si,Ni) is nanocrystalline and becomes amorphous upon Li insertion and extraction typical of silicon electrodes.13 Figure 1d shows the cycling characteristics of the 700 nm 3D(Si,Ni). A first cycle capacity of 0.36 mAh/cm2 is observed which is about 85% of the theoretical capacity expected for the corresponding Si loading [We have not employed trickle dischargeconstant voltage after constant current at 0.05 V conditions to obtain full Li insertion (Li3.75Si composition)]. Hence it can be concluded that excellent electrode utilization

due to preferential deposition of the mass transfer controlled componentcopperin the Cu−Ni codeposition process beyond a certain deposition time.37 The SEM image in Figure 1a shows the tubular morphology of the type 1 porous nickel. Pore diameters ranging from 100 to 300 nm are observed over the sample size and the pores are interconnected forming a tubular nickel network. Qualitative peel tests were conducted (using 3 M Scotch tape) on the nanoporous Ni substrate to ensure that the adhesion between the stainless steel substrate and the Ni film was strong. It was found that less than 1% of the nickel mass was removed from the peel test and hence the Ni film was proved to have good adhesion properties. To demonstrate scalability of the process, we have also deposited large area (25 cm2) type 1 porous nickel films onto stainless steel substrates using the same electrolyte under galvanostatic conditions (see Supporting Information, Table 1 and Materials and Methods). Irrespective of potentiostatic or galvanostatic deposition conditions, the SEM image showed no change in Ni nanotube morphology (Supporting Information, Figure S2). To study the performance of silicon electrode on the porous nickel current collector for Li ion battery applications, we deposited silicon of different mass loading by a low-pressure chemical vapor deposition. The typical deposition rate for Si was around 350 Å per minute. Table S2 (Supporting Information) shows different Si planar equivalent thicknesses (PET) on porous Ni films and their corresponding mass per unit area and theoretical capacity expected. The theoretical capacity calculations in Table S2 (Supporting Information) were conducted using Li3.75Si as the fully lithiated composition. First, we deposited very thin films (10, 40, and 100 nm) of amorphous Si onto type 1 porous Ni film. The morphology of the 3D(Si,Ni) electrode was observed using SEM (Supporting Information, Figure S3). Extended cycling studies were also conducted where the capacity and Coulombic efficiency have 6062

dx.doi.org/10.1021/nl302114j | Nano Lett. 2012, 12, 6060−6065

Nano Letters

Letter

Figure 3. Structural and electrochemical characterization of type 2, 2 μm 3D(Si,Ni) electrode. (a) SEM image of type 2 porous nickel showing large pore sizes of 3−5 μm. (b) SEM image of top view of type 2, 2 μm 3D(Si,Ni) electrode obtained after deposition of 2 μm PET of Si on type 2 porous nickel film. The SEM image shows conformal Si deposition around pores of nickel. (c) SEM image of cross section of type 2, 2 μm 3D(Si,Ni) electrode obtained after deposition of 2 μm PET of Si on type 2 porous nickel film. The SEM image shows the individual channels of the tubular nickel matrix coated with a-Si along the length of the channels. (d) Cycling characteristics of type 2, 2 μm 3D(Si,Ni) electrode at 0.05 C and 0.2 C showing a reversible capacity of ∼0.8 mAh/cm2 and 0.6 mAh/cm2.

of reversible capacity upon doubling the thickness of silicon (∼0.2 mAh/cm2 for Si PET 500 nm). The 1 μm 3D(Si,Ni) electrode was also cycled at higher current rate of ∼2 C to study its rate capability. It has been observed that there is a 30% capacity loss over the first 10 cycles and a stable reversible capacity of ∼0.3 mAh/cm2 after 200 cycles of charge/discharge (Supporting Information, Figure S7). To summarize the electrochemical performance of the type 1 3D(Si,Ni) electrode we have plotted the reversible capacity vs the Si PET and observe a linear increase in capacity per unit area upon increase in Si loading up to a Si PET of 1 μm (Figure 2d). Hence it can be concluded that type 1 porous nickel serves as a good current collector substrate for up to relatively thick (1 μm) PETs of silicon. To further increase capacity obtained from the 3D(Si,Ni) electrode, we deposited 2 μm Si PET onto the type 1 porous nickel. On observation of SEM images of 2 μm 3D(Si,Ni) it is clear that the pores get completely filled, thereby resulting in large microsized silicon particle overgrowth (Supporting Information, Figure S8). Due to larger particle sizes, galvanostatic charge/discharge measurements were conducted at a very low current rate (C/20) for the 2 μm 3D(Si,Ni). From the transient voltage increase observed due to polarization within the voltage profile of the first cycle discharge, it can be inferred that the crystallinity of the deposited silicon electrode increases with the increase in thickness (Supporting Information, Figure S9). The 2 μm 3D(Si,Ni) shows poor cycling characteristics due to increased internal stress within the microparticles of silicon and resultant decripitation from current collector substrate (Supporting Information, Figure S10). Hence it has been observed that the increase of Si PET beyond 1 μm leads to complete filling of pores (type 1 porous nickel), large volume of expansion and cracking upon Li

was observed even at a relatively high current rate of nearly 1 C. An irreversible capacity loss of about ∼28% was observed after the first discharge, due to secondary surface reactions, and a capacity loss of only ∼5% was observed between the second and 120th cycle. A stable reversible capacity of ∼0.22 mAh/cm2 was observed for the 700 nm 3D(Si,Ni). The stable cycling characteristics could be attributed to improved adhesion of active silicon electrode to nickel current collector and the porous electrode structure which allows for volume expansion of electrode during lithiation. We have also studied the rate capability of the 3D(Si,Ni) electrode films. Figure 2a shows the cycling characteristics of 500 nm 3D(Si,Ni) electrode at a current rate of ∼0.8C. A stable reversible capacity of 0.18 mAh/cm 2 and Coulombic efficiencies greater than 99.2% was observed after 75 cycles of charge/discharge. Figure 2b shows a capacity (mAh/cm2) vs Crate plot for the 500 nm 3D(Si,Ni) electrode, where discharge capacities after 25 cycles are plotted at different C-rates. From Figure 2b it can be observed that the 3D silicon electrode shows a capacity retention of ∼60% of nominal capacity (capacity at C-rate C/10) at a current rate of 1.1 C. The 3D structure of the porous nickel allows for improved silicon mass accommodation resulting in thinner layers of Si electrode. Thinner Si electrode layers lead to faster ion transport paths, hence improving charge/discharge capability of the electrode. Excellent adhesion of Si on Ni substrate31,33 leads to better electronic pathway also affecting the rate capability of the electrode. To improve the overall capacity of the electrode per unit area, we further increased the Si loading per unit area on the porous nickel sample. For the 1 μm 3D(Si,Ni) we observe conformal coating of silicon with partial filling of pores in the nickel structure (Supporting Information, Figure S6). Figure 3c shows a stable reversible discharge capacity of ∼0.42 mAh/cm2 at 0.9 C after 70 cycles of charge/discharge, showing a doubling 6063

dx.doi.org/10.1021/nl302114j | Nano Lett. 2012, 12, 6060−6065

Nano Letters

Letter

accommodate increased Si electrode loading. The 3D(Si,Ni) was fabricated by Si deposition on porous Ni by the thermal CVD method. Type 2 3D (Si,Ni) showed a reversible capacity of ∼0.8 mAh/cm2 at a current rate of 0.05 C with 80% of the reversible capacity retained at four times larger current density of 0.2 C. The porous Ni substrates help in excellent electrode utilization for high silicon PET due to shorter diffusion distances and reduced internal stresses upon volume expansion. 3D(Si,Ni) electrodes exhibit shorter ionic and electronic diffusion pathways, thereby showing good charge/discharge rate capability. We believe that the internal pore engineering around the Si electrode helps to accommodate the large volume changes during Li insertion/extraction, thereby mitigating of internal stress build up. The formation of nickel silicide at the interface of electrode and current collector plays an important role in improving electrode adhesion, thereby improving capacity retention upon extended cycling.38,39 The electrochemical method based on pore size and film thickness engineering is scalable and has the potential to control the aspect ratio of the pores which could lead to the development of higher capacity Si anode. The binder free approach of developing scalable, stable Si anodes, could have a huge impact in realizing high energy density lithium ion batteries. Materials and Methods. Fabrication and Characterization of Porous Type 1 and 2 Nickel Films. The Ni−Cu films were deposited by the electrochemical method on surface roughened stainless steel (SS) foils (Type 304, 0.1 mm thick, Alfa Aesar). The stainless steel foil was polished with fine sand paper and cleaned thoroughly with deionized water before electrodeposition. The electrodepositions were carried out in a three-electrode cell consisting of a Pt wire/gauge (CH Instruments) or stainless steel foil (Alfa Aesar) counter electrode, Ag/AgCl reference electrode (CH Instruments), and the stainless steel foil as working electrode using a potentiostat/galvanostat (NADA scientific). For type 1 porous nickel film SS foils of area 4 and 25 cm2 was chosen. The counter electrode area was matched with the working electrode area to ensure uniform current distribution. An electrochemical bath consisted of 5 or 200 mL aqueous solution of NiSO4 (1 M), CuSO4 (0.05 M), and H3BO3 (0.5 M) for the type 1a or type 1b porous nickel film, respectively. Electrochemical deposition was conducted under potentiostatic (type 1a) or galvanostatic (type 1b) conditions of −1.1 V with respect to Ag/AgCl reference and −4 mA/cm2, respectively. Deposition was conducted at room temperature and at a pH of 4.1. For type 2 porous nickel film SS foils of area 25 cm2 was chosen. The counter electrode area was matched with the working electrode area to ensure uniform current distribution. The electrochemical bath consisted of 200 mL of aqueous solution of NiSO4·6H2O (184 g/L), CuSO4·5H2O (6.24 g/L), citric acid monohydrate (62.2 g/L), sodium dodecyl sulfate (0.2 g/ L), and saccharin (0.5 g/L) for type 2 nickel. Electrochemical deposition was typically conducted under galvanostatic conditions of −10 mA/cm2 at room temperature with magnetic stirring (200 rpm). After the electrochemical deposition, the stainless steel foil supported Ni−Cu film was removed from the cell and cleaned with large amounts of deionized water, followed by 80 °C drying in vacuum for 4 h. The Ni−Cu film was then treated with copper etchant (Transene Company) for 6 h to remove the copper component from the film. After etching the copper, the nanoporous nickel film was washed with copious amounts of deionized water to remove contaminants within the pores

insertion/deinsertion, thereby preventing the increase of capacity per unit area by increased Si thickness. The pore size and film thickness of the porous nickel have been engineered to improve the Si electrode mass distribution for higher Si loading (PET ∼ 2 μm) compared to that of the type 1 porous nickel. We have fabricated type 2 porous nickel using selective etching of copper from Cu−Ni films deposited from a modified electrochemical bath (with saccharine additive) and under galvanostatic conditions.37 Saccharine was used as a smoothening agent to reduce dendritic Cu−Ni surface37 which enhanced the thickness of Cu−Ni film without the enrichment of mass transfer controlled Cu phase. The electrochemical bath was also stirred at 200 rpm to ensure uniform distribution of ions at the surface of the working electrode (stainless steel foil). The resulting Cu phases in the Cu−Ni microstructure were larger than that of type 1 porous nickel hence increasing the average pore size of the type 2 porous nickel.34 Type 2 porous nickel also exhibited a tubular morphology (Figure 3a). It is observed that type 2 porous nickel film had an average pore diameter of ∼3 μm and an average film thickness of ∼5−7 μm (Supporting Information, Figure S11). Silicon films of PET 2 μm were deposited onto type 2 porous nickel, and the morphology of the resulting 3D(Si,Ni) electrode was observed (Figure 3b and c). From the SEM images in Figure 3b and c it can be observed that the silicon electrode is coated conformally around the pores of the type 2 nickel along the entire length of the nickel tubular morphology. We have also observed nickel silicide formation at the interface of the electrode and current collector by X-ray photoelectron spectroscopy (XPS) analysis of the electrode (Supporting Information, Figure S12). The improved adhesion of the electrode can be attributed to the silicide formation thereby mitigating capacity losses due to decripitation.38,39 The porous structure of the nickel combined with excellent electrode adhesion to the current collector results in the improved structural integrity of the silicon based electrodes upon cycling (Supporting Information, Figure S13). The internal pores in the 3D(Si,Ni) architecture help in mitigating the internal stress and subsequent fracture of silicon electrode from the current collector surface. Hence, by engineering pore size and film thickness, we have shown effective control over the electrode mass distribution for different electrode loading. The type 2, 2 μm 3D(Si,Ni) was tested for its electrochemical performance in Li half cells. Figure 3d shows the cycling characteristics of the type 2, 2 μm 3D(Si,Ni) at current rate of 0.05 and 0.2 C. From the cycling characteristics it can be seen that a stable reversible capacity of ∼0.8 mAh/cm2 is observed after 12 cycles of charge/discharge when cycled at current rate 0.05 C. Even at higher current rate of 0.2 C we observe good stable capacity retention (0.6 mAh/ cm2). Hence, by the use of type 2 porous nickel (larger pore diameter and film thickness), an overall increase in capacity per unit area has been accomplished by scalable increase in Si loading (Si PET ∼ 2 μm), which was not possible with the type 1 porous nickel. Engineering pore dimensions (pore thickness and diameter) of porous Ni current collector substrates is shown to be an efficient strategy to achieve excellent utilization of the high capacity Si electrode at relatively high current rates. The porous nickel current collector was fabricated by selective leaching of Cu from the codeposited Cu−Ni films on stainless steel substrates. Nanoporous tubular-like Ni morphology with an average diameter of 150 nm and height of 1.5 μm was obtained. With the use of an additive, the Ni pore size and porous Ni film thickness was increased to 6064

dx.doi.org/10.1021/nl302114j | Nano Lett. 2012, 12, 6060−6065

Nano Letters

Letter

and dried at 120 °C in vacuum for 24 h. The mass of the electro-deposited Cu−Ni on SS before and after selective Cu etching was measured using a microbalance (Mettler Toledo). The average thickness of the films was measured using a digital micrometer. To determine complete Cu removal after etching of nanoporous Ni on SS, powder X-ray diffraction studies were carried out (Rigaku benchtop powder X-ray diffractometer with Cu Kα radiation). Fabrication and Structural Characterization of 3D(Si,Ni) Electrode. The Si deposition process was carried out using a thermal chemical vapor deposition process (Applied Material, LPCVD) which has a capability of remotely cracking Si precursor. The silane precursor was used to deposit Si thin films. To achieve conformal Si coating on nanoporous Ni tubes and to increase the Si loading, experimental parameters such as carrier/reacting gas mixture flow, substrate temperature and chamber pressure, and deposition time was varied. Here onward we refer to Si coated on nanoporous tubular Ni as 3D (Si,Ni) electrode. The surface morphology of the nanostructured tubular Ni before and after the Si deposition was performed using a scanning electron microscope (FEI, Environmental SEM). Electrochemical Characterization of 3D(Si,Ni) Electrode. All galvanostatic charge/discharge measurements were conducted using an Arbin battery analyzer. The electrochemical measurements of the 3D (Si,Ni) were performed in a CR2032 coin cell (Pred Materials). The Li metal foil was used as a counter/reference electrode and 1 M solution of LiPF6 in 3:7 (v/v) mixture of ethylene carbonate (EC) and diethyl carbonate (DEC) with vinylene carbonate (2%) additive as an electrolyte. Celgard 2500 polyethylene−polypropylene separator was used in this study. The cell charge/discharge was performed between 1.5 and 0.05 V vs Li/Li+ at different current rates. Here the discharge means going negative in potential (V) with respect to open circuit potential (OCV) to 0.05 V before next charging step to 1.5 V. Coin cells fabricated with Li metal as anode in this study are referred as coin half cells.

■

(7) Wagner, F. T.; Lakshmanan, B.; Mathias, M. F. J. Phys. Chem. Lett. 2010, 1, 2204−2219. (8) Nagaura, T.; Tozawa, K. Batteries Solar Cells 1990, 9, 209. (9) Schalkwijk, W. A.; van Scrosati, B. Advances in lithium-ion batteries; Kluwer Academic/Plenum Publishers: New York, 2002. (10) Whittingham, M. S. Chem. Rev. 2004, 104, 4271−4302. (11) Sato, K.; Noguchi, M.; Demachi, A.; Oki, N.; Endo, M. Science 1994, 264, 556−558. (12) Endo, M.; Kim, C.; Nishimura, K.; Fujino, T.; Miyashita, K. Carbon 2000, 38, 183−197. (13) Kasavajjula, U.; Wang, C.; Appleby, A. J. J. Power Sources 2007, 163, 1003−1039. (14) Ryu, J. H.; Kim, J. W.; Sung, Y.-E.; Oh, S. M. Electrochem. SolidState Lett. 2004, 7, A306−A309. (15) Chan, C. K.; Peng, H.; Liu, G.; McIlwrath, K.; Zhang, X. F.; Huggins, R. A.; Cui, Y. Nat. Nanotechnol. 2008, 3, 31−35. (16) Wu, H.; Chan, G.; Choi, J. W.; Ryu, I.; Yao, Y.; McDowell, M. T.; et al. Nat. Nanotechnol. 2012, 7, 310−315. (17) Wu, H.; Zheng, G.; Liu, N.; Carney, T. J.; Yang, Y.; Cui, Y. Nano Lett. 2012, 12, 904−909. (18) Lee, S.-J.; Lee, J.-K.; Chung, S.-H.; Lee, H.-Y.; Lee, S.-H.; et al. J. Power Sources 2001, 97−98, 191−193. (19) Kovalenko, I.; Zdyrko, B.; Magasinski, A.; Hertzberg, B.; Milicev, Z.; Burtovyy, R.; Luzinov, I.; Yushin, G. Science 2011, 334, 75−79. (20) Liu, G.; Xun, S.; Vukmirovic, N.; Song, X.; Olalde-Velasco, P.; Zheng, H.; Battaglia, V. S.; Wang, L.; Yang, W. Adv. Mater. 2011, 23, 4679−4683. (21) Hu.; et al. Chem. Commun. 2011, 47, 367−369. (22) Choi.; et al. J. Power Sources 2010, 195, 8311−8316. (23) Graetz, J.; Ahn, C. C.; Yazami, R.; Fultz, B. Solid-State Lett. 2003, 6, A194−A197. (24) Cui, L.-F.; Yang, Y.; Hsu, C.-M.; Cui, Y. Nano Lett. 2009, 9, 3370−3374. (25) Park, M. H.; et al. Nano Lett. 2009, 9, 3844−3847. (26) Cui, L.-F.; Ruffo, R.; Chan, C. K.; Peng, H.; Cui, Y. Nano Lett. 2009, 9, 491−495. (27) Magasinski, A.; Dixon, P.; Hertzberg, B.; Kvit, A.; Ayala, J.; Yushin, G. Nat. Mater. 2010, 9, 353−358. (28) Takamura, T.; Ohara, S.; Uehara, M.; Suzuki, J.; Sekine, K. J. Power Sources 2004, 129, 96−100. (29) Baranchugov, V.; Markevich, E.; Pollak, E.; Salitra, G.; Aurbach, D. Electrochem. Commun. 2007, 9, 796−800. (30) Bourderau, S.; Brousse, T.; Schleich, D. M. J. Power Sources 1999, 81−82, 233−236. (31) Wang, D.; Yang, Z.; Li, F.; Wang, X.; Liu, D.; et al. Mater. Lett. 2011, 65, 3227−3229. (32) Uehara, M.; Suzuki, J.; Tamura, K.; Sekine, K.; Takamura, T. J. Power Sources 2005, 146, 441−444. (33) Takamura, T.; Uehara, M.; Suzuki, J.; Sekine, K.; Tamura, T. J Power Sources 2006, 158, 1401−1404. (34) Liu, Z.; Xia, G.; Zhu, F.; Kim, S.; Markovic, N.; et al. J. Appl. Phys. 2008, 103, 064313. (35) Sun, Li.; Chien, C.-L.; Searson, P. C. Chem. Mater. 2004, 16, 3125−3129. (36) Green, T. A.; Russell, A. E.; Roy, S. J. J. Electrochem. Soc. 1998, 145, 875−881. (37) Pellicer, E.; Varea, A.; Pané, S.; Nelson, B. J.; Menéndez, E.; et al. Adv. Funct. Mater. 2010, 20, 983−991. (38) Xu, W.; Palshin, V.; Flake, J. C. J. Electrochem. Soc. 2009, 156, H544−H547. (39) Cho, J.-H.; Li, X.; Picraux, S. T. J. Power Sources 2012, 205, 467−473.

ASSOCIATED CONTENT

S Supporting Information *

Tables and figures. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: email:[email protected], [email protected], and [email protected]. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Long, J. W.; Dunn, B.; Rolison, D. R.; White, H. S. Chem. Rev. 2004, 35. (2) Zhang, H.; Yu, X.; Braun, P. V. Nat. Nanotechnol. 2011, 6, 277− 281. (3) Long, J. W.; Rolison, D. R. Acc. Chem. Res. 2007, 40, 854−862. (4) Hassoun, J.; Panero, S.; Simon, P.; Taberna, P. L.; Scrosati, B. Adv. Mater. 2007, 19, 1632−1635. (5) Armand, M.; Tarascon, J.-M. Nature 2008, 451, 652−657. (6) Taberna, P. L.; Mitra, S.; Poizot, P.; Simon, P.; Tarascon, J.-M. Nat. Mater. 2006, 5, 567−573. 6065

dx.doi.org/10.1021/nl302114j | Nano Lett. 2012, 12, 6060−6065