Letter pubs.acs.org/NanoLett
Tissue-like Silicon Nanowires-Based Three-Dimensional Anodes for High-Capacity Lithium Ion Batteries Emanuel Peled,*,† Fernando Patolsky,*,† Diana Golodnitsky,†,‡ Kathrin Freedman,† Guy Davidi,† and Dan Schneier† †
School of Chemistry, Faculty of Exact Sciences and ‡Applied Materials Research Center, Tel Aviv University, Tel Aviv, 69978, Israel ABSTRACT: Here, we report on the scalable synthesis and characterization of novel architecture three-dimensional (3D) high-capacity amorphous silicon nanowires (SiNWs)-based anodes with focus on studying their electrochemical degradation mechanisms. We achieved an unprecedented combination of remarkable performance characteristics, high loadings of 3− 15 mAh/cm2, a very low irreversible capacity (10% for the 3−4 mAh/cm2 anodes), current efficiency greater than 99.5%, cycle stability (both in half cells and a LiFePO4 battery), a total capacity of 457 mAh/cm2 over 204 cycles and fast charge−discharge rates (up to 2.7C at 20 mA/cm2). These SiNWs-based binder-free 3D anodes have been cycled for over 200 cycles, exhibiting a stable cycle life. Notably, it was found that the growth of the continuous SEI layer thickness, and its concomitant increase in resistivity, represents the major reason for the observed capacity loss of the SiNWs-based anodes. Importantly, these NWs-based anodes of novel architecture meet the requirements of lithium batteries for future portable, and electric-vehicle, applications. KEYWORDS: Silicon, batteries, nanowires, anode, composite materials, energy storage size of ∼150 nm, below which cracking does not occur, and above which surface cracking and particle fracture takes place. Silicon nanowire arrays provide a highly porous medium, which allows easy expansion of silicon during lithium insertion. Furthermore, nanowires may be directly connected to the current collector, and this is expected to improve the electron conductivity of the whole anode material. There are two main approaches for the preparation of silicon nanowires: direct growth methods and etching methods.11−18The vapor−liquid−solid (VLS) mechanism discovered about 50 years ago by Wagner and Ellis19 is the most popular of the growth methods. VLS growth is usually performed in a chemical vapor deposition (CVD) reactor by decomposition of silicon-bearing gas precursors, like silane (SiH4) or silicon tetrachloride (SiCl4), over a temperature range of ∼300−1000 °C, depending on the gas precursor and the type of metal catalyst employed.20 Silicon NWs can be grown on different types of metal catalysts, like Au, Cu, Ag, In, Ga, Zn and others. The first papers dealing with the lithiation of silicon nanowire (SiNW) structures were published 15 years ago by Zhou and co-workers.21,22 The authors investigated the amorphization process of crystalline SiNWs upon lithium insertion. In most cases, SiNWs were grown on the surface of a flat solid substrate, mainly stainless steel, as a “forest” structure23,24 The main drawbacks of this approach are (i) relatively too-low surface capacity (typically less than 1 mAh/ cm2), (ii) very high irreversible capacity (about 30%), which is
T
he increasing demand in energy storage has stimulated great interest in lithium-battery research. Most commercially available lithium-ion batteries have, as the anode, graphite with a theoretical capacity of 372 mAh/g. In order to increase the energy density of the lithium battery, better anodes and cathodes are still required. Silicon has attracted much attention because its theoretical capacity is 4200 mAhg−1, an order of magnitude greater than that of graphite. Si exhibits a low delithiation potential against Li/Li+, thus high battery voltages can be reached. Furthermore, silicon is a low-cost and environmentally friendly material, and it is the second most abundant element on Earth. Nevertheless, the main disadvantage of high-capacity anode materials, such as Sn, Sb, Si and Al, is their very large volume expansion and contraction during Li insertion/deinsertion, followed by cracking and pulverization of the anode material.1 For instance, silicon exhibits up to ∼320% volume expansion2 upon complete alloying with lithium, thus inducing a rapid degradation of Si-based anodes.3,4 One plausible way to deal with the Si detrimental pulverization is to reduce the size, and/ or thickness, of the anode down to the nanoscale. Several approaches have been reported, including the use of nanospheres, nanotubes, nanowires and porous structures.5−8 Si nanostructures have the advantage of a shorter diffusion distance for lithium species, which can improve the power performance of the battery. It has been shown that the high surface-to-volume ratio of nanoparticles helps to better withstand stress, and substantially limit the cracking extent.9 The existence of a strong particle-size-dependent fracture behavior of Si nanoparticles during the first lithiation cycle was shown experimentally;10 that is, there exists a critical particle © XXXX American Chemical Society
Received: February 24, 2015 Revised: May 7, 2015
A
DOI: 10.1021/acs.nanolett.5b00744 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 1. SEM micrographs of 3D anodes before cycling. (a) Pristine paper, (b) 0.36 mg/cm2 SiNW loading, (c) 1.92 mg/cm2 SiNW loading, and (d) cross section of 3D carbon matrix with a 1.63 mg/cm2 SiNW loading. Scale bars are (a) 200, (b) 5, (c) 20, and (d) 50 μm.
on SiNWs directly grown on a carbon support. The (de)alloying reactions occurring in amorphous-Si have been followed for the first time at an atomistic level for the whole electrode structure, which is important to the functioning of practical lithium-ion batteries. Importantly, in current lithium-ion-battery technology the irreversible capacity is about 10% or less, the areal capacity is about 3−4 mAh/cm2, and the current efficiency is over 99.9%. In addition, in all lithium batteries the anode is covered by a thin solid electrolyte interphase layer (SEI).25 Ideally, this SEI is permeable to lithium ions, but it is intrinsically an electronic insulator,25 and its formation prevents (or slows down) further electrolyte decomposition during the cycles to follow. Unfortunately, in the case of the silicon-based anodes, significant “breathing” of the anode material during insertion/ deinsertion of lithium causes cracks, exposing the bare silicon surface to the electrolyte, followed by the creation of a fresh SEI,28 thus losing battery capacity and increasing battery impedance. Here, in order to overcome the disadvantages of the “SiNWs forest” or binder-nanowire composite film concepts, we have developed a novel approach based on the three-dimensional (3D) growth of highly dense mostly amorphous SiNWs on
required for the formation of the SEI (solid electrolyte interphase),25 and (iii) insufficient current efficiency (typically 95 to 99.5%). Notably, most publications on SiNWs-based anodes demonstrate a single desired property (low Qir, high surface capacity, high electrode capacity (mAh/gSi), high current efficiency or high cycle number) but not all of these attributes concurrently achieved for the same electrode, and in most cases the good performances were demonstrated for only very low and impractical areal capacity. Recently, 6000 cycles at 12C-rate with 940 mAh/gSi were demonstrated26 by the use of a 30 nm thick double-walled silicon nanotube (DWSiNT) anode. In this work, the authors used a very low Si loading in the range of 0.02 to 0.1 mg DWSiNT/cm2, which consists of 60% Si and 40% SiOx and thus, a pure Si loading of 0.016 to 0.06 mgSi/cm2. As the cycle life is known to be inversely proportional to the thickness of the electrode, we can safely assume that the 6000 cycles were demonstrated for 16 μgSi/ cm2 anodes. The electrode capacity under these conditions is 0.015 mAh/cm2, which is at least 2 orders of magnitude lower than required for a practical lithium ion battery. Also, during 6000 cycles, this anode delivered 90 mAh/cm2, providing a current density of only 0.18 mA/cm2 at 12C-rate. Also recently, Gray et al.27 presented an in situ NMR study of Li−Si processes B
DOI: 10.1021/acs.nanolett.5b00744 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters carbon fibers-based open-structure conductive networks (GDL25AA by SGL group). The idea was to avoid a dense SiNWs forest, to grow SiNWs that are not parallel to each other, and to provide open space for the electrolyte to penetrate into the anode. In addition, this open space allows room for the growth of the SEI without blocking the ion path in the electrolyte, which is the case for a dense SiNWs forest. By this route, we have been able to produce remarkably high loadings of 3−33 mAh/cm2, very low irreversible capacity (of the order of only 10% for the 3−4 mAh/cm2 samples), and current efficiency greater than 99.5%; properties that meet the requirements of lithium batteries for future portable and electric vehicle applications. These SiNWs-based binder-free anodes have been cycled for over 200 cycles, exhibiting a stable cycle life. Moreover, in this work a focus was made on studying the degradation mechanism of 3D-SiNWs-GDL-based anodes, thus enabling their controlled fabrication and improved electrochemical performance. Experimental Section. SiNWs were grown by the CVDVLS method on the carbon fibers-based three-dimensional conductive networks, 160 μm thick GDL25AA carbon substrate, (Figure 1a). The thickness of individual carbon fibers is approximately 7 μm and its mass is 3.8 mg/cm2. The SiNWs were grown in a three-dimensional configuration, covering the whole accessible surfaces of the individual fibers, throughout the complete volume of the 3D carbon matrix. Shortly, carbon fiber conductive networks (i.e., SGL and Freudenberg) were first treated by oxygen plasma (400 mTorr O2, at 100 W for 10 min) in order to modify and improve their surface wetting properties, followed by the adsorption of polyL-lysine at room temperature for 60 min. The positively charged polylysine layer serves as an electrostatic adhesion agent for the subsequent deposition of gold nanoparticles (20−80 nm AuNPs diameter in water for 15 min). Silicon nanowires growth was carried out in a CVD reactor via the VLS mechanism at 460 °C and 25 Torr, using SiH4 gas as precursor (flow rate 5 sccm), B2H6 (flow rate 6.25 sccm), and diluted with Ar gas carrier (flow rate 10 sccm) for a period of 30−60 min. Then, the sheets were cut into about 1 cm2 square pieces, and a drop of an ink made of Shawnigan Black carbon and poly(styrene-co-butadiene) (Sigma- Aldrich) mixture in toluene (9:1 w/w) was applied to one side of the sheet in order to improve the electric contact to the stainless steel current collector. The carbon loading was about 0.5 mg/ cm2. The electrodes were dried in vacuum for 24 h at 50 °C and 2 h at 100 °C. CR2032 coin cells were assembled inside a glovebox (O2 and H2O
