Rajesh
Thomas
and
G.
Mohan Rao
*
Indian Institute of Science, Dept. of Instrumentation and Applied Physics, Malleshwaram PO IISc Bangalore, Bangalore, Karnataka 560012, India. E-mail: thomasphy@gmail.com
First published on 21st October 2014
All solid state batteries are essential candidate for miniaturizing the portable electronics devices. Thin film batteries are constructed by layer by layer deposition of electrode materials by physical vapour deposition method. We propose a promising novel method and unique architecture, in which highly porous graphene sheet embedded with SnO2 nanowire could be employed as the anode electrode in lithium ion thin film battery. The vertically standing graphene flakes were synthesized by microwave plasma CVD and SnO2 nanowires based on a vapour–liquid–solid (VLS) mechanism via thermal evaporation at low synthesis temperature (620 °C). The graphene sheet/SnO2 nanowire composite electrode demonstrated stable cycling behaviours and delivered a initial high specific discharge capacity of 1335 mAh g−1 and 900 mAh g−1 after the 50th cycle. Furthermore, the SnO2 nanowire electrode displayed superior rate capabilities with various current densities.
The developments of all solid state thin film batteries (TFB) are essential for the miniaturization of portable electronic devices and to avoid the problems caused by the liquid electrolyte. A thin film battery is composed of several electrochemical cells that are connected in series and/or in parallel to provide the required voltage and capacity. The solid state (micro batteries) is essential for state of the art microelectronic and portable electronic devices. The energy and power requirements for these devices are ever increased and lead to the research for novel battery structure and electrode materials with increased volume energy and power density.4,5 Due to their thinner dimension, TFBs have greater applications in making thinner electronic devices, RFID tags, wireless sensors and implantable medical devices.6,7 These batteries exhibit the same voltage and current as their bulky counterparts. The advancement of nanotechnology and cell engineering caused TFBs to achieve high energy density and cycle life.
Thin film batteries are built layer by layer by vapour deposition. There are various methods being used to deposit thin film electrode materials onto the current collector such as, sputtering, CVD and sol gel methods.8 Plasma based material deposition techniques are well suited for industrial applications. Multiple thin film cells can reduce the footprint of a large planar battery area; this will be a solution to attain large energy and power requirements. The first solid state thin film battery was developed by Hitachi Co., Japan in 1982.9 A thin film battery, with more stable LIPON as the solid electrolyte, was developed by a group at Oak ridge national laboratory.10 There are several companies and research institutions actively involved to develop better thin film battery systems.8 J. H. Pikul et al.11 have demonstrated that 3-D micro batteries from bicontinuous nanoporous electrodes gave the power density of 7.4 mW cm−2 μm−1. Graphene (having a high theoretical specific surface area of 2630 m2 g−1) or carbon nanostructures can enhance the storage capacity by several orders.12 Graphene can accommodate Li ions on both sides. This makes its theoretical capacity two times larger than (∼740 mAh g−1) that of other carbon materials.
In order to increase the battery's capability, various elements or compounds (e.g. Si, Sn, Ge, SnO2etc) are alloyed with lithium to get larger specific capacities than commercial graphite. Among the various metal and metal oxide based electrodes, both tin (Sn) and tin oxide (SnO2)13,14 have been discussed as important anode materials for Li-ion batteries, because of their semiconducting properties combined with high capacity (Sn, 994 mAh g−1 and SnO2 781 mAh g−1) higher than that of graphite.15 However, significant capacity fading with cycling is a problem specifically with metal oxide based materials due to large volume changes during Li alloying and dealloying, which leads to metal segregation, crystallographic deformation16 and agglomeration of active materials. In the case of Sn, the volume changes are as high as 259%.17
The huge volume changes associated with lithiation/delithiation process are a major drawback for SnO2 based electrodes, hindering it from real applications.18 In order to solve the issues associated with metal oxide materials, various nanostructured SnO2 and SnO2 composites have been proposed. These can be accommodating for volume changes during the cyclic process. Nanostructured electrode materials possess many advantages, such as increased number of electrochemical active sites and better control over stress due to the lithiation/delithiation process.19,20 Nanostructured metal oxide particles with homogeneous carbon coating have already been reported to improve the mechanical and electrochemical stability.21 The carbon support on nanoparticles enables a better accommodation of the large volume change and improves the electron conductivity of the electrode.22 Nanocarbon materials often gave high coulombic efficiencies and cycle life, but the volumetric energy density and rate performance were poor due to the formation of a large SEI film.23 By using nanostructured inorganic materials, the rates of the electron and counter ion transport are increased. Transport properties have been improved somewhat by using nano-architectured electrodes with large surface areas.24
Nanowires and nanoparticles of alloy materials are a better choice for advanced Li batteries. The transport of Li ions is one of the main issues that give rise to current limitations. Engineering the active materials into exceptional nano-architectures will enhance the efficiency of the electrochemical performance such as specific capacity, rate capabilities and cycling stability.25 Several studies have shown that SnO2 nanowire/heterostructures gave a capacity of ∼700 mAh g−1 up to 15 cycles26 then faded to ∼300 mAh g−1 after 50 cycles. Kim et al. have demonstrated stable cycling behaviour for SnO2 nanowire electrode which delivered a high specific discharge capacity of 510 mAh g−1.27 Graphene nanosheet (GNS) matrix having a flexible two-dimensional structure, high surface area (over 2600 m2 g−1), excellent electrical conductivity and the ability to diminish the stress of the electrode upon cycling will support SnO2 based nanostructures.28,29
In the present study, we are aiming for a higher electrochemical performance of the SnO2NW@GNS anode by adding the advantage of both materials vs. graphene nanosheets and SnO2 nanowires. We have modified the carbon into a graphene nanosheet (GNS) matrix and nanowires of SnO2 were embedded in it (SnO2NW@GNS). GNS consists of a few layers of graphene grown vertically over the substrate and provides very high porosity. The high surface area of GNS is able to attach more SnO2 nanowires on to the wall of the graphene sheet, thus easing the Li ion migration into the active materials. Hence, tailoring the structure of SnO2 materials or embedding them on carbon matrices should enhance the SnO2NW@GNS anode performance. It is well known that the following equations are involved in the SnO2 based anodes.
SnO2 + 4Li+ + 4e− ↔ Sn + 2Li2O | (1) |
Sn + xLi+ + xe− ↔ LixSn (0 ≤ x ≤ 4.4) | (2) |
The first reaction is considered irreversible and is responsible for the irreversible capacity loss from the first to the second discharge cycles.30,31 However, this irreversible capacity could also become reversible for nanometer-sized SnO2 nanoparticles or nanowires.32 The second equation is widely known for the reversible capacity, in which Li is continuously alloying with Sn.33 These kinds of nano-architectures of carbon are very suitable for energy storage applications.
Fig. 2a shows a typical field emission scanning electron microscopy (FESEM) image of the as-synthesized SnO2 nanowire grown over the GNS matrix with Au catalyst at a substrate temperature of 620 °C. The total time taken for the deposition of SnO2 nanowires was about 30 minutes. The long nanowires of SnO2 are seen in the porous structure of the GNS matrix. The wires are uniformly distributed over the GNS matrix. A cross section SEM image of SnO2NW@GNS is shown in Fig. 2b. All the nanowires are seen with a tip on the top (white circles on the cross section image) and were assumed to be an eutectic metal alloy, indicating that the nanowires were grown via tip-led growth according to the VLS mechanism.35
Fig. 2 (a) FESEM image of SnO2 nanowires (b) cross section SEM image of SnO2 nanowires on GNS, (c) TEM image of SnO2NW@GNS nanowires and (d) HRTEM of single nanowire (inset) SAED pattern. |
The XRD pattern of the SnO2NW@GNS electrode is shown in the ESI (Fig. S1†). The peaks marked in the spectra are well matched with the rutile-tetragonal SnO2 phase (space group P42/mnm). A broad peak at 2θ = 21° was assigned to the graphitic peak coming from the GNS matrix.
Fig. 2c shows the transmission electron microscopy image of SnO2NW@GNS. The wires seem to arise from the graphene sheet (darker side of the image) with a eutectic metal alloy at the tip of each wire. An energy-dispersive (EDS) line scan of an individual nanowire from the tip along the nanowire stem indicated that the tip of the metal alloy was abundant of catalyst, gold, whereas the nanowire stem consisted of pure SnO2 without any presence of Au on it (ESI Fig. S2†). The HRTEM of a single nanowire of SnO2 confirmed that the as-synthesized wires are highly-ordered consisting of a single crystal phase with sharp periodic lattice fringe. Each nanowire is very thin with a diameter of ∼10 to 5 nm. The SAED pattern (inset of Fig. 2d) shows that the nanowires have the preferential growth direction of [101]. The inter-atomic plane distance calculated to the (101) lattice was 2.56 Å.
XPS measurements were performed for the electrodes before and after the electrochemical studies. Fig. 3a and b show the survey spectrum and high-resolution spectrum of Sn 3d5/2 before performing the electrochemical studies. The spectrum of Sn 3d5/2 was deconvoluted into two peaks at 486.7 eV and 484.8 eV for SnO2 and metallic tin, respectively. A small percentage of metallic tin might have accumulated during deposition.
Fig. 3 XPS spectra of SnO2NW@GNS electrodes before electrochemical cycles (a) survey scan of electrode, (b) high resolution spectrum of Sn 3d5/2. |
Fig. 4b and c show galvanostatic charge–discharge cycles and the rate capability of the SnO2NW@GNS electrode. The first discharge and charge capacities are 1335 mAh g−1 and 4930 mAh g−1, respectively at a current density of 23 μA cm−2. The total capacity results from the contributions of SnO2 nanowires and graphene nanosheets. The much higher charge capacity is due to the GNS matrix. The high degree of defects in GNS may act as Li ion sites that contributed to a greater capacity. A large irreversible capacity loss of about 3600 mAh g−1 (73%) was observed for the sample at the first cycle. This huge irreversible capacity loss is due the irreversible reduction of SnO2 to Sn as described in eqn (1) and other irreversible processes such as decomposition of the electrolyte.39,40 During the second cycle, 64% of the charge capacity was retained for the discharge capacity, which was 1240 mAh g−1. Subsequent discharge cycles obtained 67% and 75% of the charge capacities. On the first charge, a flat plateau appeared around 0.8 V, which was attributed to the SEI formation. The charge–discharge profiles of SnO2 nanowires on copper substrates are given in S4 (ESI†). The first charge and discharge capacities were obtained as 2169 mAh g−1 and 828 mAh g−1, respectively, at a current density of 23 μA cm−2. About 62% of the irreversible capacity loss was observed in the first cycle, which was less than that of the SnO2NW@GNS electrode.
During the second cycle, about 89% of the charge capacity was retained for the discharge capacity. Only first cycles of charge–discharge capacities of various current densities were given in Fig. S5 of the ESI.† The irreversible decomposition of SnO2 to metallic tin surrounded by an amorphous, inactive Li2O matrix was ascribed to the SEI.41,42 The amorphous Li2O matrix allows Li ions to diffuse through and prevent the agglomeration of Sn atoms or LixSn alloy regions during volume changes.43 The rate performance of the electrode was evaluated at various current densities from 23 μA cm−2 to 177 μA cm−2. A coulombic efficiency of ∼87% was obtained for the SnO2NW@GNS electrode after 50 cycles at a current density of 23 μA cm−2 (Fig. 4d).
A broad peak of Sn 3d5/2 is observed after the first discharge cycle (Fig. 5c). The peaks observed at 485.3 eV and 486.7 eV are assigned to Sn and Sn4+, respectively. The corresponding Li 1s peak also deconvoluted to two peaks at 54.6 and 56.3 eV (Fig. 5d) attributed to metallic lithium and the Li2O phase, respectively. The Li2O is an amorphous phase, which allows Li ions to pass through it and it is electrically insulating. XPS spectra of the electrode after the 50th cycle are presented in the ESI (S6†).
In order to understand the modifications that occur in the electrode during the charge/discharge process, an additional TEM study was carried out for the electrode after the charge cycle (after the first lithiation) and after the discharge cycle (after the first delithiation). Fig. 6a and b show TEM images of electrodes after the first charge state. The nanowire morphology of SnO2 was largely distorted and even a mild structure of nanowire can be seen in the TEM image (Fig. 6a). During charging, the formation of Li2O results in a large volume expansion in the nanowire.44 The high resolution TEM image of the electrode shows clear nanowire morphology and the SAED pattern (inset of Fig. 6b) shows retention of some degree of crystallinity of the nanowire during the first charged state. The lattice fringes have been changed when compared to the original SnO2 nanowire shown in Fig. 2d. After the discharge cycle, the complete morphology of SnO2 nanowires undergoes significant change and nanoparticles of Tin are formed (Fig. 6c). The HRTEM of tin particles shows crystalline nature and sizes of 5–10 nm. In the discharge process, the Li2O did not take part in electrochemical lithiation process and only LixSn nanoprecipitates were active in the lithiation process.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ta04836a |
This journal is © The Royal Society of Chemistry 2015 |