A STUDY ON SYNTHESIS AND CHARACTERIZATION OF THIN POLYMER/CERAMIC HYBRID FILM SEPARATORS AND METAL OXIDE COATED ELECTRODES FOR LITHIUM ION BATTERIES A Thesis Presented to the Faculty of the Graduate School of Cornell University In Partial Fulfillment of the Requirements for the Degree of Master of Science by Sri Dithya Atluri May 2016 © 2016 Sri Dithya Atluri ABSTRACT The search for new electrode and membrane materials for lithium-ion batteries (LIBs) has been under investigation to satisfy the ever-growing demands for better performance with higher energy density, improved safety and longer cycle life. In this study, electrospraying has been used to produce mesoporous thin films for the application as Li-ion battery separators. Electrospraying is a film formation technique that utilizes electrical rather than mechanical forces to form uniformly sprayed films. Polyacrylonitrile (PAN) was used to produce these thin membranes of thickness ranging between 20 and 25 microns. In this system, Polyethylene Oxide was incorporated as a sacrificial polymer. An ideal separator for LIB must be permeable and must have pore sizes ranging from 30 to 100 nm to facilitate good ion transport. In addition, a low thickness is required for high energy and power densities. Using this approach, we were able to achieve thinner and more porous membranes with pore sizes ranging from 0.1 microns to 0.3 microns. Silica precursors like PSSQ(Poly(silsesquioxane)) and OPSZ (Organopolysilazane) were incorporated into the film to increase the ionic conductivity of the membranes and thermal stability thereby increasing the battery performance. Results from SEM, BET, DSC, FTIR, Impedance Spectroscopy, Capillary Flow Analysis, Dynamic Mechanical Analysis of resulting mesoporous polymer/ceramic will be discussed. The battery tests reveal that mesoporous polymeric/ceramic film separators exhibit higher capacity and better capacity retention than polymeric/ceramic nanofiber separators. Meanwhile, metal oxides can prevent the corrosion of the electrode under harsh electrochemical conditions and thus they are regarded as promising electrode coating materials for highperformance Lithium Ion Batteries (LIBs). Zirconium metal oxide was studied as a potential anode coating material to further improve the cycle stability and performance of the LIBs. The Zirconium metal oxide was electrosprayed onto the silicon (Si)/reduced graphene ocide (RGO) anodes. Si/RGO anodes have been prepared by gas-assisted electrospraying the mixture of Si and Graphene Oxide (GO), followed by thermal treatment. Results from SEM, Impedance Spectroscopy, battery testing will be discussed. BIOGRAPHICAL SKETCH Sri Dithya Atluri was born in Vijayawada, India on 10th of Decemeber 1991. She grew up in Hyderabad, India. In August 2010 she started her undergraduate education at National Institute of Technology Warangal, India. After her sophomore year she got the opportunity to pursue an internship at Indian Institute of Technology Madras (IITM) in Chennai. She worked on the project “Electrochemical Studies of Metal Ions In Room Temperature Ionic Liquid Medium” under the able guidance of Prof. G. Ranga Rao in the Department of Physical Chemistry at IITM. During her undergraduate studies she also interned at Technical University of Darmstadt, Germany. She worked under Prof. Florian Mueller Plathe on Molecular Dynamics simulation of glycan monomers to evaluate the self-diffusion coefficients for both the water and sugar molecule. In the winter of 2012, she worked at Coca Cola Pvt. Ltd. as an industrial trainee. She received her Bachelor of Technology in Chemical Engineering from National Institute of Technology Warangal. She was awarded a gold medal for being the best outgoing student. In August of the same year she began her graduate studies in the School of Chemical and Biomolecular Engineering at Cornell University in Ithaca, NY. In May 2016 she will be receiving her Master of Science degree in Chemical Engineering from Cornell University. iii To my Parents, Rajendra Prasad Atluri and Bhavani Atluri and my brother, Raviteja Atluri and sister-in-law, Sujitha Atluri iv ACKNOWLEDGMENTS I express my gratitude to my advisor Prof. Yong Lak Joo, School of Chemical and Biomolecular Engineering, Cornell University for his invaluable ideas and valuable guidance throughout my project. I thank Prof. Margaret Frey for allowing the use of materials of her lab and also for her role on my special committee. I would like to thank EMD Inc., Axium Nanofiber LLC., LG Chemicals, MTI Inc. for providing funds to support this work. I would also like to thank Cornell Center for Materials Research (CCMR) for their facilities for all the material characterizations. I also want to thank all the members of the Joo research Group for their support and helpful advice. I especially want to express my gratitude to Dr. Ling Fei, Brian P. Williams, Joseph Michael Carlin, Soshana Smith for their help in the lab and valuable advice. Finally, I would like to thank my family and friends for their continuous support and guidance. v Biographical Sketch Dedication Acknowledgements Table of contents List of Figures List of tables TABLE OF CONTENTS iv v vi vii ix xi 1 Introduction 1.1  Electrospinning Process and Formation of the Fibers 1.2  Literature survey 1.3  Lithium Ion batteries 1.4  Post Processing and Functionalization 1.5  Inclusion of Silica Precursor into the Fiber Mat 1 1 5 9 13 14 2   Experimental Setup, Materials and Methods 17 2.1  Chemicals 17 2.2  Methods 2.2.1   Synthesis of Polyacrylonitrile (PAN) and Polyethylene Oxide (PEO) polymeric solutions 2.2.2   Preparation of PAN/PEO films form the polymeric solution 2.2.3   Preparation of Metal Oxide Coated Si/RGO anodes 2.2.4   Scanning Electron microscope (SEM) 2.2.5   Brunauer-Emmett-Teller (BET) Theory 2.2.6   Capillary flow Analysis 2.2.7   Fourier transform infrared spectroscopy (FTIR) 2.2.8   Battery Testing and Impedance Spectroscopy 2.2.9   Dynamic Mechanical Analysis 2.2.10   Differential Scanning Calorimetry and Thermal Properties 2.2.11   Thermogravimetric Analysis 2.2.12   Cyclic Voltammetry 17 18 20 22 23 26 28 29 30 31 33 34 3   Results And Discussion: PAN and PEO Films As Membranes For Li-Ion Batteries 3.1  Surface Morphology: Scanning Electron Microscopy 3.2  Brunauer-Emmett-Teller (BET) Theory 3.3  Fourier transform infrared spectroscopy (FTIR) 3.4  Capillary Flow Analysis 3.5  Battery Testing and Impedance Spectroscopy of film separators 3.6  Dynamic Mechanical Analysis (DMA) of the Film Separators 3.7  Differential Scanning Calorimetry (DSC) of PAN/PEO film 3.8  Thermogravimetric Analysis (TGA) of Film Separator 36 39 40 42 43 48 50 52 4   Results And Discussion: Metal Oxide Coated Anodes For Lithium Ion Batteries 4.1  Surface Morphology: Scanning Electron Microscopy 4.2  Battery Testing of Zirconium Oxide Coated Anodes 4.3 Cyclic Voltammetry 54 54 63 66 vi 5   Future Work 6   Conclusion 7   References 68 72 74 vii LIST OF FIGURES Figure 1.1: A typical electrospinning setup Figure 1.2: Electrospinning setup inside a Glove box Figure 1.3: Electrospraying Setup Figure 1.4: A schematic of Li-ion battery Figure 1.5: Structure of Poly(silsesquioxane) (PSSQ) Figure 1.6: Structure of Organopolysilazane (OPSZ) Figure 1.7: Comparison of DMA of PAN membrane and PAN/PSSQ membrane 1 3 4 7 11 11 12 Figure 2.1: Experimental setup used for electrospraying Zirconium Oxide Figure 2.2: Synthesis of PAN/PEO Films Figure 2.3: Synthesis of Metal Oxide Coated Anodes Figure 2.4: Scanning Electron Microscope Figure 2.5: Wet run and dry run of PAN/PEO film Figure 2.6: FTIR peaks Figure 2.7: A typical DSC curve Figure 2.8: A typical cyclic voltammogram of a redox couple 16 16 18 19 22 23 25 27 Figure 3.1: SEM image of 5 wt% 6:4 PAN/PEO as-spun SEM 28 Figure 3.2: SEM image of 3 wt % 7:3 PAN:PEO Film 29 Figure 3.3: SEM image of 3 wt% 6:4 PAN:PEO Film 29 Figure 3.4: SEM image of 3 wt% PAN/PEO/OPSZ 30 Figure 3.5: SEM image of PAN/PEO/PSSQ Figure 3.6: Incremental Pore Area vs Pore Width(Ao) 30 32 Figure 3.7: FTIR peaks of PAN/PEO Film 33 Figure 3.8: Pore Size Distribution of PAN/PEO Film 34 Figure 3.9: Pore Size Distribution of PAN Fiber 34 Figure 3.10: Comparison of the cycle performance of 6:4 PAN/PEO Film with that of PAN fiber with a C-rate of C/2 and a current density of 100mA/gm 36 Figure 3.11: Comparison of Cyclic Performance of PAN/PEO film with that of PAN/PEO/PSSQ films. The performance tests were carried out at a charge rate of C/2 37 Figure 3.12: Comparison of the cycle performance of PAN/PEO/PSSQ Film with that of PAN/PSSQ fiber. The performance tests were carried out at a charge rate of C/2 38 Figure 3.13: Comparison of the cycle performance of PAN/PEO/OPSZ Film with that of PAN/OPSZ fiber.The performance tests were carried out at a charge rate of C/2 39 Figure 3.14: Comparison of the Impedance Spectroscopy data of PAN/PEO/PSSQ Film with that of PAN/PSSQ fiber 40 Figure 3.15: Comparison of the Impedance Spectroscopy data of PAN/PEO/PSSQ Film with that of PAN/PSSQ fiber 41 Figure 3.16: Impedance Spectroscopy data of PAN/PEO/PSSQ Film of 25 microns thick 42 Figure 3.17: Impedance Spectroscopy data of PAN/PEO Film of 23 microns thick 42 Figure 3.18: Comparison of the DMA data of 6:4 PAN/PEO Film with that of PAN fiber 43 Figure 3.19: Comparison of the DMA data of PAN/PEO/PSSQ Film with that of PAN/PSSQ fiber 44 Figure 3.20: Comparison of the DMA data of PAN/PEO/PSSQ Film with that of PAN/PEO film viii to show the increased mechanical stiffness of the film with the addition of the precursor PSSQ 45 Figure 3.21: DSC curve of PAN/PEO film and PAN fiber. The broader peak corresponds to PAN/PEO film and the narrow peak to the fiber 47 Figure 3.22: TGA curves of PAN fiber and PAN/PEO film 48 Figure 4.1: Si/RGO Annealed Anode 49 Figure 4.2: Zirconium Coated Si/RGO Anode 50 Figure 4.3: SEM of Zirconium Oxide Sprayed anode before heat Treatment 50 Figure 4.4: SEM of Zirconium Oxide Coated Anodes after heat treatment at 3500C for 60 seconds 51 Figure 4.5: SEM of 50 wt% of Zirconium Oxide Sprayed anode before heat Treatment 53 Figure 4.6: SEM of 50 wt% Zirconium Oxide Coated Anodes after heat treatment at 3500C for 60 second 53 Figure 4.7: SEM of 50 wt% Zirconium Oxide gas-assisted sprayed anode before heat Treatment 54 Figure 4.8: SEM of 50 wt% Zirconium Oxide gas-assisted coated anodes after heat treatment at 3500C for 60 second 55 Figure 4.9: SEM of 70 wt% Zirconium Oxide sprayed anode before heat Treatment 56 Figure 4.10: SEM of 70 wt% Zirconium Oxide sprayed anode before heat Treatment 56 Figure 4.11: Cycle Performance of Si/RGO anodes and metal oxide coated anodes. The first 5 cycles tested at 100 mA/g and after 5 cycles the current density was maintained at 1 A/gm 57 Figure 4.12: Cycle Performance of Electrosprayed and spin coated Zirconium Oxide and Si/RGO anodes 58 Figure 4.13: Cycle performance of electrosprayed Zirconium Oxide onto Si/RGO anode until 20 cycles 59 Figure 4.14: Cyclic Voltammetry Curve of Metal Oxide Coated Anode 60 Figure 4.15: Cyclic Voltammetry Curve of Metal Oxide Coated Si/GO and Si/GO 61 Figure 5.1: SEM image of Zirconium Oxide coated Si/RGO anode synthesized using co-solvents NMP and DMF Figure 5.2: SEM image of Zirconium Oxide coated Si/RGO anode after 200 cycles Figure 5.3: SEM image of Zirconium Oxide coated Si/RGO anode after 20 cycles 62 63 63 ix LIST OF TABLES Table 2.1: PAN/PEO Films composition Table 2.2: Parameters for elctrospraying of PAN/PEO films Table 3.1: BET data of the PAN/PEO Films Table 3.2: DSC parameters of PAN/PEO film and PAN fiber Table 4.1: Parameters for electrospraying Zirconium Oxide onto Si/RGO 18 19 40 51 58 x CHAPTER 1 INTRODUCTION 1.1 Electrospinning Process and Formation of the Fibers Electrospinning is a fiber production method which uses electric force to draw charged threads of polymer solutions or polymer melts up to fiber diameters in the order of ten nanometers. Electrospinning shares characteristics of both electrospraying a conventional solution and dry spinning of fibers. Electrospinning produces fiber mats with a large surface area to mass ratio. The basic setup of electrospinning is composed of a high-voltage power supply, a spinneret and an electrically conductive collector. In a typical electrospinning process, a polymer or precursors/polymer solution is pumped into a syringe through a thin nozzle. The distance between the nozzle and the collector is usually 20-30 cm. The collector is typically contacted with counter electrode to accommodate as-spun finer mats or sprayed films in this system. A schematic of the electrospinning setup to synthesize fiber mats or films is shown in figure 1.1. Figure 1.1: A typical electrospinning setup 1 There are three stages in the electrospinning process: •   Jet initiation and the extension of the jet along a straight line •   Whippping Instability: The growth of a bending instability and the further elongation of the jet. •   Solidification of the jet as nano-fibers. The fiber morphology and the diameter are affected by spinning solution properties including viscosity, elasticity, electrical conductivity and surface tension. Operational conditions such as electric field strength and flow rate of the solution affect fiber morphology and diameter, while the temperature and humidity of the surroundings also play a role. To control the effect of the humidity on the morphology of the fibers, we use glove box for the electrospinning process. The control of humidity to increase the porosity of electrospun fibres has also been proposed by other researchers. It was found that humidity does alter the fiber morphology by altering the rate at which the fibers dry and that there is an optimal humidity for the electrospinning of fibers. A fiber morphology consisting of beaded fibers is undesirable in the case of the membrane application and anode coating processes. The three main factors that have been found to influence fiber beading are: Viscosity, surface tension and the density of the net charges carried by the liquid jet. To reduce the beading either the viscosity is increased, the surface tension decreased or the net charge density increased. On the other hand, fiber diameter is mainly affected by the viscosity of the solution, electrical conductivity of the solution, the feeding rate and the electric field strength. Increasing the viscosity leads to larger fiber diameters, increased electrical conductivity reduces the fiber diameter and a faster feeding rate results in larger fibers. 2 Figure 1.2: Electrospinning setup inside a Glove box We extended electrospinning to electrospraying process to produce thinner films as separators in the Li-ion batteries. The thin films in comparison to thicker commercial membranes will aid in improving the energy density of the batteries. Using thinner films increase the cycle life of the battery and also it has been shown in the later studies that film separators showed higher cyclic retention in comparison to the polymeric nanofiber separator. To give a brief overview of the process in usage; Electrospraying is a process similar to electrospinning but is used when the viscosity of the liquid is sufficiently low. The electric charge draws the liquid from the capillary nozzle in the form of a fine jet, which eventually disperses into droplets. The droplets produced by electrospraying are highly charged, usually close to one-half of the Rayleigh limit, and can be smaller than 1 mm. The size distribution of the droplets is usually narrow, with low standard deviation. Electrospraying can be used for the production of small, nearly monodisperse particles when a colloidal suspension of solid 3 nanoparticles or a solution of a material is sprayed [9]. In electrospraying the size of the droplets can be controlled mainly by the liquid flow rate, and the droplet charge by adjusting the voltage applied to the nozzle. The charged aerosol is self dispersing, which prevents the droplets from coagulation. Figure 1.3: Electrospraying Setup 1.2 Literature Survey The term “Electrospinning” derives from electrostatic spinning and was first widely used around 1994. Electrospinning roots from the earliest electrospray used for painting in early nineteenth century. The milestone for this technique should be marked by a series of patents filed by Formhals during 1934 and 1944 [2-6]. He designed an electrospinning setup to produce polymer fibers which were much larger than so called nanofibers. His contributions other than this setup 4 also included producing fibers of polymer blend system and aligning fibers in a parallel way. Since then, electrospinning has seemed to be dormant until 1966 when a new apparatus was designed by Simons [7]. In his patent, he demonstrated that production of ultra thin and light non-woven fabrics were feasible through electrospinning and found that less viscous solutions could produce shorter and finer fibers while more viscous solutions produce thicker and relatively continuous fibers. During the same period, Taylor [8-10] studied the shape of initial charged droplet in an electrospinning. In his studies, he found that the solution droplet starts to become cone-shaped when the needle potential increases. The critical angle for the stable cone is at φ=49.3o. Beyond this angle, a jet will break through which is due to the maximum instability of the fluid surface induced by the electrical field. The deformed droplet at the tip of the needle was later called “Taylor Cone”. Taylor’s theory later was proved by the production of sub-micron fibers by Baumgarten in 1971 [11]. When the critical voltage was reached, the relative viscous solution (1.7-215 Poise in his experiments) did not breaks into droplet instead a fluid stream broke out. He found that the diameter of the resulting fibers was determined by the solution viscosity by a power law of ~0.5. In the 1980’s Larrondo and Manley devised the first melt electrospinning setup [12-14]. They studied the jet formation from polyethylene/paraffin solution and more importantly molten polyethylene. His studies greatly extended the application of electrospinning. However, due to the lack of temperature control in spinning region, only 50 µm fibers were able to be produced. Nevertheless, he demonstrated that controlling the temperature of the spinneret and spinning 5 voltage could affect the fiber diameters. Reneker and coworkers re-explored this processes and hence contributed greatly to the development of this field [15-20]. Numerous experiments and a number of theoretical works have been carried out to better understand the process and search for more applications. Most of current studies deal with organic polymers or biopolymers. On the other hand, due to some special requirements in applications such as high temperature or strong mechanical strength, inorganic filaments have also been attracting huge focus. Some of them are quite interesting, for example, carbon nanotubes on carbon fibers [20] and Vanadium Oxide whiskers on Titanium oxide fibers [21]. Polymeric fibers containing strengthening agents such as carbon nanotubes [22-24], silica nanoclays [25-27] and graphite [28] et al. are also hot topics in the field of electrospinning. Coaxial electrospinning is another branch of electrospinning which can produce fibers with multiple layers [29-31] or hollow fibers [32]. Current researches on electrospinning can be categorized into four categories: electrospinning of different materials; functionalization of as-spun fibers; various applications of electrospun fibers; structural studies of electrospun fibers; Simulation of the electrospinning process. 1.3 Lithium Ion Batteries (LIBs) Lithium ion batteries have proven themselves the main choice of power sources for portable electronics. Besides consumer electronics, lithium ion batteries are also growing in popularity for military, electric vehicle, and aerospace applications. 6 Figure 1.4: A schematic of Li-ion battery LIBs are characterized by high specific energy and high specific power, which are the advantages that most other electrochemical energy storage technologies cannot offer. In addition, some other advantages such as high efficiency, long life cycle and low self discharge rate make lithium-ion batteries well suited for applications such as energy storage grid and electric transportation. In Li-ion batteries, the separator is placed between two electrodes, the anode and the cathode. It prevents the physical contact of electrodes while serving as the electrolyte reservoir to enable ionic transport. Although the separator does not directly participate in electrode reactions, its structure and properties affect battery performance, including cycle life, safety, energy density, and power density by regulating the cell resistance and kinetics. Microporous polyolefin membranes are widely used in Li-ion batteries since they have good chemical stability and 7 mechanical strength. However, low porosity and poor wettability of these membranes affect the cell resistance and kinetics negatively and restrict cell performance, including energy density and rate capability. Over the past 10 years, electrospun nanofiber membranes have been extensively studied as alternative separators for Li-ion batteries due to their large porosity and unique pore structure. In electrospinning technique, continuous nanosized polymer fibers are produced through the action of an external electric field imposed on a polymer solution. Enhanced electrochemical properties such as higher C-rate capability, better cycling performance and lower cell resistance have been reported for Li-ion cells using electrospun nanofiber-based separators. Many polymeric fiber mats have been investigated as separators for Li-ion batteries. Polyacrylonitrile (PAN) has been studied as a separator material and PAN-based separators show promising properties, including high ionic conductivity, good thermal stability, high electrolyte uptake and good compatibility, with Li metal. Separator materials in lithium ion batteries must have the ability to transport ions through their porous membranes while maintaining a physical separation between the anode and cathode materials in order to prevent short-circuiting. Furthermore, the separator must be resistant to degradation during the battery’s operation. In a Li-ion battery, the separator must be a thin and flexible solid, and hence the main aim of the present study would be able to produce thinner membranes with higher porosity and comparable thermal and mechanical stability. Electrospraying is a promising technology for producing polymeric films. Narrow size distributions can be obtained by the control of electrospraying flow rate and polymer concentration, with average particle sizes ranging from 10 to 20 µm. Electrospraying is a onestep technique which has a potential to generate narrow size distributions of submicrometric particles, with limited agglomeration of particles and high yields. The principles of 8 electrospraying are based on the ability of an electric field to deform the interface of a liquid drop. The theory of charged droplets states that if an electrified field is applied to any droplet, the electric charge generates an electrostatic force inside the droplet, known as the Coulomb force, which competes with the cohesive force intrinsic to the droplet. When the applied Coulomb force is able to overcome the cohesive force of the droplet manifested in the surface tension, the droplet will undergo breakup into smaller droplets in the micro- to nano-scales. This phenomenon begins at the Taylor Cone, referring to the progressive shrinkage of the unstable, charged macro-droplet into a cone from which the smaller charged droplets will be ejected as soon as the surface tension is overcome by the Coulomb force. 1.4 Post Processing and Functionalization Post processing is another important method towards the functionality of fiber mat. Surface coating, calcinations, etching and hydrolysis et al. are widely implemented in achieving additional functionalities. Such post processing generally does not change the morphology of the fiber mat and nano-sized fibers are preserved. Calcination usually creates porous structures into fibers and produce high surface area materials. Etching is an effective way to produce hollow fibers from two component coaxial fibers and it is also an alternative to remove one component to create porous structures into fibers made of multi-component polymer blend. Hydrolysis as seen in literature is mostly used to introduce inorganic particles. Surface coating is one of the most straightforward ways to introduce more functionality to the fibers. Drew and co-workers have shown that the surface of electrospun nanofibers can be modified by via liquid-phase deposition to titania nano- particles [33]. Gold [34] or silver films [35] have been deposited onto the electrospun fiber surface via chemical vapor deposition or sputtering coating. Calcination is a well established method for producing carbon fibers from 9 PAN fiber precursor [36, 37]. Fibers composed of other inorganic materials such as silica [38], alumina-borate oxide [39] and titanium oxide [40-42] et al. have been successfully prepared by calcinations of their electrospun composite fiber precursors. For polymer blend fibers, selective removal of certain component can form highly structured morphologies such hollow fiber from co-axial electrospinning or porous fibers. 1.5 Inclusion of Silica Precursor into the Fiber Mat Inclusion of the Silica precursor into the membranes increases the thermal conductivity as well as the thermal stability of the polymer membrane. The Silica precursors used in our studies are Organopolysilazane(OPSZ), otherwise called 1500 Rapid Cure and Poly(silsesquioxane)(PSSQ). Addition of these Silica precursors also helps in increasing the mechanical strength of the membranes. From the DMA (Dynamic Mechanical Analysis) results (Figure 1.7) it can be seen that the PAN (Polyacrylonitrile) membranes with Silica precursor incorporated into them have much higher tensile strength than the PAN membranes. 10 Figure 1.5: Structure of Poly(silsesquioxane) (PSSQ) Figure 1.6: Structure of Organopolysilazane (OPSZ) 11 Stress(MPa) DMA   14 PAN  Membrane 12 10 Inclusion o  f  Silica   Precursor 8 6 4 2 0 0 5 10 15 20 25 30 Strain(%) Figure 1.7: Comparison of DMA of PAN membrane and PAN/PSSQ (60:40) membrane 1.5 Metal Oxide Coated Anodes In a conventional LIB cell, lithium metal oxide (e.g., LiCoOx) and graphite are used as cathode (positive electrode) and anode (negative electrode) materials, respectively. Two electrodes are separated by a porous membrane separator and soaked in a non-aqueous liquid electrolyte. During discharging, Li+ ions de-intercalate from the anode, pass through the electrolyte and insert into the cathode, while electrons flow from the anode to the cathode through the external circuit and power the electronic devices. The process is reversed during charging when an external voltage is applied to the battery. Since LIBs function based on the reversible shuttling of Li+ ions and the difference of electrochemical properties between the cathode and the anode, the inherent properties of the electrode materials would be a crucial factor that largely determines the overall performance of batteries. 12 Metal oxides can prevent the corrosion of the electrode under harsh electrochemical conditions and thus they are regarded as promising electrode coating materials for high-performance Lithium Ion Batteries (LIBs). In this study Zirconium Oxide has been investigated as a potential electrode coating material for LIBs. 13 CHAPTER 2 EXPERIMENTAL SETUP, MATERIALS AND METHODS 2.1 Chemicals All the chemicals used in this study are of analytical reagent grade. Polyacrylonitrile(PAN) of two different molecular weights: 150,000 gm/mole and 200,000 gm/mole, Polyethylene Oxide(PEO) of molecular weight 10,000 gm/mole were produced from M/s. Sigma Aldrich, Solvents Dimethyl Formamide (DMF), N-Methyl Pyrrolidine (NMP) were purchased from M/s. Sigma Aldrich, commercial silicon nanoparticle (70-100nm) form M/s. MTI., Zirconium Oxide was provided by M/s. EMD Performance Matrials , Graphene Oxide aqueous suspension(2wt% GO sheets, 98wt% H2O) was from M/s. Dongjin Semichem, Ltd. 2.2 Methods 2.2.1 Synthesis of Polyacrylonitrile (PAN) and Polyethylene Oxide (PEO) polymeric solutions PAN and PEO films have been synthesized from 3 wt. % solution of the polymers varying both PAN and PEO weight ratios in the polymer matrix. The PAN/PEO films were synthesized using electrospinning method. Since the viscosity of the solution should be lower to get a uniform thinner film, the concentration of the solution must be much lower than that required for fiber formation. After conducting a viscosity test, we observed that the ideal concentration of the polymeric solution would be 3 wt % for the production of thin polymeric films. 14 Polymer Concentration(Weight %) PAN to PEO weight ratio Results 3 8:2 Less pores 3 7:3 Cell dies 3 6:4 Better battery test results 3 5:5 Cell dies Table 2.1: PAN/PEO Films composition 2.2.2 Preparation of PAN/PEO films from the polymeric solution As discussed in chapter 1, for electrospinning and electrospraying the basic components required are: syringe pump, high voltage supply and collector. We use electrospraying technique to produce the PAN/PEO films on an electrospun PVA nanofiber mat. During the electrospraying process, there are several parameters which all have an inter-dependent influence on viscosity, electrical conductivity, particle size, distribution and loading capacities. These parameters include voltage, distance to collector, needle gauge, flow rate, polymer, solvent, surfactant and aqueous/organic phase. The various parameters used for the PAN/PEO film are: PARAMETERS VALUES AND RANGES Voltage 20 kV Distance between the syringe and collector 20 cm Flow rate 0.008 ml/min Needle Gauge 17 Table 2.2: Parameters for elctrospraying of PAN/PEO films 15 The electrosprayed samples were immersed in heated water bath for approximately 6 hours for the removal of the sacrificial polymer PEO. Figure 2.1: Experimental setup used for electrospraying Zirconium Oxide Figure 2.2: Synthesis of PAN/PEO Films 16 2.2.3 Preparation of Metal Oxide Coated Si/RGO anodes In a typical preparation procedure of active materials/GO electrodes, 3.0 g of GO aqueous suspension (Dongjin Chemical, 2wt% GO sheets, 98wt% H2O) was diluted in 5.0 g of DI water. After sonicating the suspension for 1 hr, 120mg active materials (2:1 weight ratio with graphene oxide) were added. The main material of interest in this study is commercial silicon nanoparticle (70-100nm, MTI.), other materials include commercial sulfur (Spectrum chemical MFG, Corp), homemade MoO3 nanobelts and Sn/carbon naofibers (Sn/CNF). Detailed preparation methods for homemade materals are available in the following paragraphs. The mixture was then sonicated for another hour and stirred overnight before spraying. Gas-assisted electrospray was applied for directly depositing binder-free electrodes. The electrospray was carried out under ambient condition using a Harvard Apparatus PHD 2000 Infusion syringe pump with a coaxial needle set. Solution was supplied through the inner 17G needle and gas through outer 12G needle. The working voltage was set at 20kV, working distance at 20 cm, solution feeding rate between 0.05mL min-1 – 0.1 mL min-1 , and gas pressure at 28 psi. To obtain active materials/RGO electrodes, the as sprayed active materials/GO were annealed at 400 °C in N2 atmosphere (MTI Tube Furnace) for 1 hr to reduce GO, ramp 5ºC/min. All anode materials were deposited on copper foil, and sulfur on porous carbon nanofiber substrate. The annealed samples were then coated with Zirconium metal oxide using conventional electrospraying technique. The conditions used for the coating process are: voltage is 20 kV; distance between the collector and the spinneret being 20 cm; 17G needle and flowrate being 0.01 ml/min. The coated samples were then heat treated at 3500C for 60 seconds to loose the organic tail associated with them. 17 Figure 2.3: Synthesis of Metal Oxide Coated Anodes 2.2.4 Scanning Electron microscope (SEM) The scanning electron microscope (SEM) uses a focused beam of high-energy electrons to generate a variety of signals at the surface of solid specimens. The signals that derive from electron-sample interactions reveal information about the sample including external morphology (texture), chemical composition, and crystalline structure and orientation of materials making up the sample. These signals include secondary electrons (that produce SEM images), backscattered electrons (BSE), diffracted backscattered electrons (EBSD that are used to determine crystal structures and orientations of minerals), photons (characteristic X-rays that are used for elemental analysis and continuum X-rays), visible light (cathodoluminescence--CL), and heat. Secondary electrons and backscattered electrons are commonly used for imaging samples: secondary electrons are most valuable for showing morphology and topography on samples and backscattered electrons are 18 most valuable for illustrating contrasts in composition in multiphase samples (i.e. for rapid phase discrimination). SEM analysis is considered to be "non-destructive"; that is, x-rays generated by electron interactions do not lead to volume loss of the sample, so it is possible to analyze the same materials repeatedly. The morphologies of all the samples were characterized using fieldemission scanning electron microscopy (TESCAN-MIRA3-FESEM) microscope. Figure 2.4 Scanning Electron Microscope 2.2.5 Brunauer-Emmett-Teller (BET) Theory The Brunauer-Emmett-Teller (BET) theory has its basis in statistical thermodynamics. Although dependent on many assumptions, it is still a widely accepted theory of physical adsorption Two basic assumptions for the derivation of the equation are: (i) the surface of the adsorbent is homogeneous, and 19 (ii) there is no lateral interaction between the adsorbed molecules. The rate of condensation of molecules from the gas phase is equated to the rate of molecules evaporated from the second layer of adsorbed molecules. The rate of condensation from the second layer is equated to the rate of evaporation from the third layer, and so on. Below the saturation water pressure, the surface fraction of adsorbed molecules will increase with the increasing partial pressure of the substrate. This means that the thickness of the adsorbed layer will not be a constant through the sorption process, but will change depending on the values that are specific for each layer. The heat of adsorption of the molecular layers is assumed to be equal to the heat of liquefaction after the formation of first monolayer. Also, the evaporationcondensation conditions are assumed to be identical, except for the first layer. The last assumption is that, at the saturation vapor pressure, the number of adsorbed layers is infinite. The BET equation (Equation 1) is then derived by summing the fractional coverage at each layer. Thus, the theory is also referred as the kinetic theory of multimolecular adsorption. P/X(PO-P) = 1/XmC + (C-1/XmC).P/PO Equation(1) where X is the mass of adsorbate at vapor pressure P and temperature T, P0 is the corresponding saturated vapor pressure, and Xm is the quantity adsorbed (g/g) at monolayer coverage. The BET constant C is an energetic term defined as C = (exp(E1-EL)/RT) Equation(2) where is the heat of adsorption of the first molecular layer of adsorbate, is the heat of condensation of the adsorbate, R is the molar gas constant and T is the absolute temperature (K). The difference (E1 – EL) is a measure of the energy of interaction (J/mol) between the first monolayer of adsorbed molecules and the surface. Large values of (E1 – EL) indicate large forces 20 of hydration between the surface and the monolayer. Similar to Langmuir Theory, Xm and C can be calculated by the slope and intercept of the plot P/P0 vs P/(X·(P0-P))X or by choosing a constant value for C and solving the equation for Xm. Then, specific surface area is calculated by using Equation (3). SSA= "#∗%∗  ' ( [m2/g] Equation(3) 2.2.6 Capillary flow Analysis Capillary flow porometry analysis gives the smallest, largest and mean-flow diameter of through pores. It is a characterization technique based on the displacement of a wetting liquid from the sample pores by applying a gas at increasing pressure. Capillary flow porometry permits obtaining several parameters and information in one individual and fast measurement. In general, measurements with the wet sample (impregnated with wetting liquid) is carried out first. It is normally known as the "wet run" and the representation of the gas flow vs. the applied pressure is the so-called "wet curve". After the wet run the measurement of the same sample in dry state is carried out in order to register and analogous "dry curve". The half-dry curve is calculated and represented by dividing the flow values with respect to the applied pressure by 2 and it is also represented in the same graphic. From the representation of the three curves it is possible to identify relevant information about the sample: the maximum pore size (or first bubble point) is recorded when gas flow through the sample is detected (please see explanation about FBP above), the mean flow pore size corresponds to the pore size calculated at the pressure where the wet curve and the half dry curve meet (it corresponds at the pore size at which 50% of the total gas flow can be accounted), and the minimum pore size results from the pressure at which the wet and the dry curve meet (from this point onwards the 21 flow will be the same because all the pores have been emptied). The capillary flow analysis for the samples were carried out at Analytical Services Division, Porous Materials Inc. FLOW  RATE  VS  PRESSURE 60 POREDIST:  Sample  A  -­ Wet 50 40 FLOW RATE L/MIN 30 20 10 0 0 10 20 30PRESSUR4E0  (PSI) 50 60 70 Figure 2.5: Wet run and dry run of PAN/PEO film 2.2.7 Fourier transform infrared (FTIR) spectroscopy Fourier transform infrared spectroscopy measures the intensity of infrared beams after passing through the sample or being reflected by the sample. This technique works because the change of molecular dipoles when they move in a certain way: symmetrical and asymmetrical stretching, scissoring, rocking, wagging and twisting. The resonant frequencies are affected not only by the motion the molecule undergoes but also by the other conditions such as the strength of molecular bond and the surrounding influence by other molecular groups. Typically for the same bond, stretching corresponds to higher frequency compared to other motions and for the same motion, stronger bonds have higher frequency (triple bond > double bond> single bond). In this study, solid film was directly mounted on a sample holder with a transmission window. 22 Sample holder with the sample then was loaded into Mattson Model 5020 FTIR Spectrometer. -1 -1 -1 This instrument has a resolution of 0.4cm and covers the range from 400cm to 6000cm . Figure 2.6: FTIR peaks 2.2.8 Battery Testing and Impedance Spectroscopy Electrochemical measurement was conducted using CR-2032 coin cell. The electrodes prepared were directly used without any further treatment. A lithium foil was used as the counter electrode for testing the metal oxide coated anodes, and a mixture of 1 M LiPF6 in fluoroethylene carbonate/dimethyl carbonate (1:1 in mass) was used as the electrolyte. Cell assembly was carried out in an argon-filled glove-box. For testing the membranes, cells were prepared using the same electrolyte. The anode used in this cell was graphite whereas Lithium Cobalt Oxide cathode (MTI Corp) was used as the cathode. The galvanostatic charge/discharge measurements 23 were performed using a Land battery testing system in the voltage cutoff window of 0.01-1.5 V (vs. Li+ /Li). The current density and specific capacity are based on the total mass of electrode materials. 2.2.9 Dynamic Mechanical Analysis Dynamic Mechanical Analysis (DMA) is a technique used to measure the mechanical properties of a wide range of materials. Many materials, including polymers, behave both like an elastic solid and a viscous fluid, thus the term viscoelastic. DMA differs from other mechanical testing devices in two important ways. First, typical tensile test devices focus only on the elastic component. In many applications, the inelastic, or viscous component, is critical. It is the viscous component that determines properties such as impact resistance. Second, tensile test devices work primarily outside the linear viscoelastic range. DMA works primarily in the linear viscoelastic range and is therefore more sensitive to structure. The sample was characterized using DMA Q800 V7.5. 2.2.10 Differential Scanning Calorimetry and Thermal Properties Differential scanning calorimetry (DSC) is a thermal-analytic technique that measures the difference of heat required to increase the temperatures of reference pan and sample pan. The principle underlying this technique is that different physical transitions and sometimes chemical transitions will give off or absorb heat. These physical transitions usually include crystallization, melting and phase transition. Figure 2.7 shows a typical DSC curve under heating. 24 Figure 2.7: A typical DSC curve Analysis of DSC curve can provide us great information on transition glass temperature, melting temperature and crystallinity et al. These data are often influenced by the thermal history of the samples. For instance, two samples with same chemical composition can have different crystallinities if one was produced using quenching method and the other using natural cooling. Glass transition temperature and melting temperature are determined by the energy state of the samples which is often affected by different processing history such as thermal history or mechanical history. Programming the heating and cooling stage in a manipulated way is able to generate other useful information on the behavior of nucleation and crystallization. In most of the cases of our study, 1-3 mg of fiber samples were loaded into DSC pan and characterized in a DSC instrument (DSC Q2000 V24.9)Heating rate was 100C/min. 25 2.2.11 Thermogravimetric Analysis To quantitatively monitor the weight loss during polymer to ceramic conversion thermogravimetric analysis (TGA) is the method of choice. The weight loss is predominantly caused by the evolution of oligomers at lower temperatures followed by the release of by-product gases at higher temperatures. For example, for polysilazanes, a three-stage weight loss was ° observed by Choong et.al. First, a weight loss occurs at about 400 C due to the evolution of low molecular weight oligomers. The second weight loss occurs in the temperature range of 400 to 700 °C due to the loss of hydrocarbons, such as CH4, C2H6 and others, and the third and final weight loss was due to the loss of hydrogen evolution. Thus TGA was found to be an efficient tool to investigate and also to optimize the polymer-to-ceramic transformation during the thermal treatment. The other aspect of pyrolysis is the change in the chemical composition from polymer to ceramic, which depends on number of factors, such as the starting polymer precursor, pyrolysis atmosphere and also the degree of crosslinking. The final amorphous ceramic composition, for example, in case of polysilazanes is the stable ternary phases of silicon nitride (Si3N4), silicon carbide (SiC) and the carbon (C). The thermogravimetric analysis was ° performed on Q500 V20.10 Thermogravimetric Analyzer at a heating rate of 10 C/min to 1000 °C in an inert atmosphere of N2. 2.2.12 Cyclic Voltammetry Cyclic voltammetry (CV) is the most commonly used voltammetric technique for determining the behaviour of electroactive species and the mechanism of the electrochemical reaction. 26 In a typical CV experiment, the potential of the working electrode is continuously scanned from initial value to final value at a specified scan rate. The scan is commenced at a potential where no electrochemical reaction takes place (no current) and scanned to a region where electrochemical oxidation or reduction takes place. The response of the system is a plot between current and potential called cyclic voltammogram[44]. In these experiments, Cyclic voltammetry was measured using a PARSTAT 4000(Princeton Applied Research) electrochemical work station. Figure 2.8: A typical cyclic voltammogram of a redox couple 27 CHAPTER 3 RESULTS AND DISCUSSION: PAN AND PEO FILMS AS MEMBRANES FOR LI-ION BATTERIES 3.1 Surface Morphology: Scanning Electron Microscopy: SEM was conducted on PAN/PEO samples with various composition ratios. We can observe from the SEM of 5 wt% as-spun sample of PAN/PEO (6:4) in figure 3.1 that there is beading on the fibers and there is no film formation. From the SEM image of 3 wt% PAN/PEO (7:3) sample in figure 3.2 we can observe that the film formed is non-uniform and also agglomeration of PAN/PEO beads was observed. Figure 3.1: SEM image of 5 wt% 6:4 PAN/PEO as-spun SEM 28 Figure 3.2: SEM image of 3 wt % 7:3 PAN:PEO Film The SEM image (Figure 3.3) of 3 wt% of electrosprayed sample of PAN/PEO (6:4) solution showed a uniform polymeric film. It was also observed that as the PEO content increases more than 40 wt% with respect to PAN polymer, pores of size about 28 microns were formed. 29 Figure 3.3: SEM image of 3 wt% 6:4 PAN:PEO Film The inclusion of Silica precursors OPSZ (organopolysilazane) and PSSQ (Poly(silsesquioxane) increases the conductivity of the membranes which would further improve the battery performance. The SEM images of the films containing 30 wt % of the Silica precursor with respect to the polymeric content showed films with uniform pore distribution. Figure 3.4: SEM image of 3 wt% PAN/PEO/OPSZ 30 Figure 3.5: SEM image of PAN/PEO/PSSQ 3.2 Brunauer-Emmett-Teller (BET) Analysis The BET isotherms were obtained using the Gemini VII, Micrometrics Inc. BET data can be seen in Table 3.1. The BET data shows that the surface area of 6:4 weight ratio of PAN and PEO film is higher compared to the other. System Single Point Surface area (m2/gm) BET Surface Area(m2/gm) t-plot micropore Area(m2/gm) t-plot external Surface Area(m2/gm) BJH Adsorption Cummulative Surface Area(m2/gm) PAN/PEO 6:4 Film 31.83 33.74 3.46 30.66 32.68 PAN/PEO 5:5 Film 26.77 28.43 3.08 25.44 22.5 3 wt% h- PAN/PEO 5:5 Film 12.94 13.82 - 13.84 12.50 Table 3.1: BET data of the PAN/PEO Films 31 Figure 3.6: Incremental Pore Area vs Pore Width(Ao) From figure 3.6 we can observe that PAN/PEO film containing 40 wt % of PEO has pores between 100 Ao and 500 Ao. 3.3 Fourier transform infrared spectroscopy (FTIR) FTIR was conducted on the PAN/PEO film to test the removal of the water soluble Polyethylene Oxide. The FTIR peaks of the sample (Figure 3.7) are in line with the FTIR peaks of Polyacrylonitrile (PAN) only. This ensures the complete removal of the sacrificial polymer, PEO. 32 0.03 0.02 0.02 Absorbance 0.01 0.01 0.00 0 1000 2000 3000 Wavenumber 4000 Figure 3.7: FTIR peaks of PAN/PEO Film Apart from the FTIR analysis, weight loss tests were conducted on the samples to measure the removal of PEO polymer from the film. The weights measured before and after immersion of the film into the heated water bath showed that there was 95% removal of PEO from the PAN/PEO film. 3.4 Capillary Flow Analysis From the capillary flow analysis data for both PAN/PEO films and fibers, it can be observed that PAN/PEO films have higher number of smaller pores compared to PAN nanofibers. The mean flow pore diameter of PAN/PEO film was observed to be 0.1412 microns whereas that of PAN nanofiber mat was deducted to be 0.7217. It should be noted that the maximum pore size was 33 about 1.5 micron for the PAN nanofiber mat, and pores larger than 1 micron are reported to lead to short-circuit. On the contrary, the maximum pore size was less than 0.5 micron for PAN film. PORE  SIZE  DISTRIBUTION 8000 7000 6000 5000 4000 3000 2000 1000 0 0 0.1 0.2 0.3 AVERAGE  DIAMETER  (MICRONS) 0.4 Figure 3.8: Pore Size Distribution of PAN/PEO Film 0.5 PORE  SIZE  DISTRIBUTION 1400 1200 1000 800 600 400 200 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 AVERAGE  DIAMETER  (MICRONS) Figure 3.9: Pore Size Distribution of PAN Fiber 34 3.5 Battery Testing and Impedance Spectroscopy of film separators The battery testing was done using the cathodic cell as described in the Chapter 2. The battery testing for the PAN/PEO films was compared with that of the PAN fiber. Fig 3.10 compares the cycling performance of PAN fiber and PAN/PEO films as membranes in the cells via galvanic charging/discharging. PAN/PEO films show excellent retention and high specific capacity and also as it can be seen in comparison to PAN fiber membrane PAN/PEO system shows longer term cycle stability. Hence, using PAN/PEO films is a promising for application as Li-ion battery membrane. The battery tests were done with a current density of 100 mA/gm and at a C-rate of C/2 which is a measure of the rate at which a battery is discharged relative to its maximum capacity 140 C/2 120 100 Discharge  Capacity(mAh/gm) 80 60 40 20 PAN/PEO F  ilm PAN  Fiber 0 0 10 20 30 40 50 60 70 Cycle  Index 35 Figure 3.10: Comparison of the cycle performance of 6:4 PAN/PEO Film with that of PAN fiber with a C-rate of C/2 and a current density of 100mA/gm In comparison to the PAN/PEO film separator it can be observed from figure 3.11 that the PAN/PEO/PSSQ film separator has lower cyclic stability and comparatively lower discharge capacity. The addition of Silica precursor increases the ionic conductivity between the electrodes and also can help in preventing explosions of these batteries due to overcharging or rupture. The addition of PSSQ would not only aid in improving the thermal safety and prevent accelerated heat but also improve the mechanical stability and stiffness of the separator. Figure 3.11: Comparison of Cyclic Performance of PAN/PEO film with that of PAN/PEO/PSSQ films. The performance tests were carried out at a charge rate of C/2 36 Capacity  (mAh/g) When the mesoporous PAN/PSSQ film after PEO removal is compared to the PAN/PSSQ fiber it was observed that the PAN/PSSQ films showed higher discharge capacity and also longer cycle life. The addition of the Silica precursor increases the capacity of the cell, in comparison to the other system. It shows an increased capacity and cycle stability. 180 160 140 120 100 80 60 40 20 0 0 5 10 15 20 25 30 35 No.  of  Cycles Figure 3.12: Comparison of the cycle performance of PAN/PEO/PSSQ Film with that of PAN/PSSQ fiber. The performance tests were carried out at a charge rate of C/2 37 140 120 Discharge  Capacity(mAh/gm) 100 80 60 40 20 PAN/PEO/OPSZ F  ilm PAN/OPSZ F  iber 0 0 10 20 30 40 50 60 70 Cycle  Index Figure 3.13: Comparison of the cycle performance of PAN/PEO/OPSZ Film with that of PAN/OPSZ fiber. The performance tests were carried out at a charge rate of C/2 The mesoporous PAN film with Silica precursor OPSZ shows great capacity retention although the specific charge of the cell utilizing the film as the membrane is less in comparison to the PAN nano-fiber membrane cell. Impedance Spectroscopy We can derive the reason for better cycle stability of the films than fiber separators from the impedance spectroscopy data. For the impedance measurements the thickness of the films and fiber membranes without the Silica precursor were kept at 25 microns. The thickness of the polymer/ceramic membranes in the case of film was 23 microns and that of the fiber was kept at 25 microns. We can see from the impedance graphs (figure 3.14) that the resistance offered by 38 the films is much lower than compared to the fiber membranes. In the case of the addition of the Silica precursor (figure 3.15) we can see that the resistance of the PAN and PSSQ membrane is almost twice that of PAN/PEO and PSSQ film. Zim(Ohms) 450 400 350 300 250 200 150 100 50 0 0 'PAN/PEO' 'PurePAN' 200 400 600 800 1000 1200 1400 Zre(Ohms) Figure 3.14: Comparison of the Impedance Spectroscopy data of 6:4 PAN/PEO Film with that of PAN fiber 39 800 700 600 500 Zim(Ohms) 400 300 200 100 0 0 500 1000 1500 2000 2500 Zre(Ohms) Figure 3.15: Comparison of the Impedance Spectroscopy data of PAN/PEO/PSSQ Film with that of PAN/PSSQ fiber The addition of the Silica precursor for the PAN fibers increases the discharge capacity of the batteries and also results in the decrease of the resistance offered by the membranes which aid the increase in the ionic transport of the lithium ions. However, in the case of the films we observed that resistance offered by the polymer/ceramic films was higher compared to the polymer films. This might be attributed to the decreased porosity of the ceramic included polymeric membranes. From the impedance graph figures 3.16 and figure 3.17, we can see that the resistance offered by the PAN/PEO/PSSQ films is higher in comparison to the PAN/PEO films. The thickness of the PAN/PEO/PSSQ separator used in this study was 25 microns and that of the PAN/PEO film was kept at 23 microns. 40 Zim(Ohms) 800 700 600 500 400 300 200 100 0 0 500 1000 1500 2000 2500 Zre(Ohms) Figure 3.16: Impedance Spectroscopy data of PAN/PEO/PSSQ Film of 25 microns thick 300 250 200 150 100 50 0 0 100 200 300 400 500 600 700 800 900 Zre(Ohms) Figure 3.17: Impedance Spectroscopy data of PAN/PEO Film of 23 microns thick Zim(Ohms) 41 3.6 Dynamic Mechanical Analysis (DMA) of the Film Separators DMA tests were conducted for both the PAN fibers and PAN/PEO films to test the mechanical stability of the membranes. From the figure 3.18, it can be noted that the that the mechanical stability of the PAN fibers is more than that of the films. This can be accounted to the fact that the PAN/PEO films have been prepared using PAN of molecular weight 150,000 gm/mole whereas the PAN fibers have been produced using polymer of molecular weight 200,000 gm/mol. Though the mechanical strength of the film is not equivalent to that of the fiber, this technique of using electrospraying for producing films for the application of LIB membranes is valuable in terms of decreasing the thickness of the membranes and also improving the specific capacity and thereby the performance of the overall cell. A low thickness is required for high energy and power densities. We can also improve the mechanical strength by using higher molecular weight PAN for the film approach. 5 4.5 4 3.5 Strain  (MPa) 3 2.5 2 PAN  Fiber PAN/PEO F  ilm 1.5 1 0.5 0 0 1 2 3 Strain  (%) 4 5 6 7 Figure 3.18: Comparison of the DMA data of 6:4 PAN/PEO Film with that of PAN fiber 42 12 PAN/PSSQ F  iber PAN/PEO/PSSQ F  ilm 10 8 Stress  (MPa) 6 4 2 0 0123456 Strain  (%) Figure 3.19: Comparison of the DMA data of PAN/PEO/PSSQ Film with that of PAN/PSSQ fiber Addition of Silica precursors improves the not only the thermal stability but also the mechanical stability of the membrane to be used as the separator. We have to ensure a stress level of greater than 5 MPa to reduce shrinkage, tearing or pinhole formation. The Silica precursor PSSQ addition has improved the mechanical stability of the film separator to a great extent. From the stress-strain graph it can be shown that the addition of the precursor stiffened the film. In comparison to the PAN/PEO film the precursor added film appears to be less ductile. 43 3.5 3 2.5 Stress(MPa) 2 1.5 1 PAN/PEO/PSSQ F  ilm PAN/PEO F  ilm 0.5 0 01234567 Strain(%) Figure 3.20: Comparison of the DMA data of PAN/PEO/PSSQ Film with that of PAN/PEO film to show the increased mechanical stiffness of the film with the addition of the precursor PSSQ 44 3.7 Differential Scanning Calorimetry (DSC) of PAN/PEO film Differential Scanning Calorimetry was performed on the samples. The system was ramped at a heating rate 100C/ minute to 3500C in nitrogen atmosphere. Type Initial exothermic temp./ 0C End exothermic Peak Temp./ temp./ 0C 0C Difference between end and initial exothermic temp./ 0C Fiber 283.5 304 296 20.5 38.38 Film 303 329 312 26 16.45 Table 3.2: DSC parameters of PAN/PEO film and PAN fiber It can be noted that there were not apparent changes in the shape and heat flow energy of DSC curve of a film, in comparison with that of its corresponding precursor fiber. However, it should be noted that there is a substantial increase in the initial and peak exothermic temperature for the film which may be attributed to the fact that molecular weight of PAN used in film was much lower than that in fiber and there is residual PEO present in the film. In previous studies, it has been found that with the increase in content of some co-monomers, incorporated in the PAN fibers or films, the initiation temperature of cyclization reaction decreases, so the exothermic regime becomes broader and broader. The enthalpy calculated for the fiber was found to be 1150 J/gm whereas that for the film it was observed 458 J/gm. The lower enthalpy of crystallization indicates a less dense crystalline structure of the film in comparison to the fiber. 45 Figure 3.21: DSC curve of PAN/PEO film and PAN fiber. The broader peak corresponds to PAN/PEO film and the narrow peak to the fiber 3.8 Thermogravimetric Analysis (TGA) of Film Separator From the TGA thermogram in figure 3.22, it was observed that, weight loss significantly occurred around 300-350°C for both fiber and film and that the film is less stable than the fiber possibly due to its lower molecular weight. The onset of thermal decomposition of PAN fiber and film took place at 279 ºC temperature region. The reason was possibly due to the glassy temperature of polyacrylonitrile at 317°C. However in the region at 400-850°C TGA 46 temperature, the loss was slower for PAN/PEO film than the fiber comparatively. This indicated that the PAN film structure stabilized sooner. It is also shown in the figure 3.22 that after for about 50 % weight loss the losing rate was slower for both PAN films and fibers. It is perhaps due to the molecular cross linking that occurred in PAN structure which could inhibit the weight loss of PAN fiber. It is commonly believed that a cross-linked polymer would naturally be more thermally stable than the corresponding thermoplastic polymer. It is believed that molecular crosslink in the PAN polymer could crystallize the PAN structure. 120 PAN  Fiber   PAN/PEO F  ilm 100 80 Weight  Percentage 60 40 20 0 0 200 400 600 800 1000 Temperature (  0C) Figure 3.22: TGA curves of PAN fiber and PAN/PEO film 1200 47 CHAPTER 4 RESULTS AND DISCUSSION: METAL OXIDE COATED ANODES FOR LITHIUM ION BATTERIES 4.1 Surface Morphology: Scanning Electron Microscopy The annealed samples as of Si/RGO prepared as discussed in Chapter 2 were coated with various metal oxides and tested using the land battery setup also described in Chapter 2. Zirconium Oxide and Aluminum Oxide coated anodes were tested. Various annealing temperatures were considered for Zirconium Oxide coated anodes. After an initial battery testing with Zirconium Oxide coated anodes at different annealing temperatures as shown in figure 4.11, it was observed that coated anodes annealed at 3500C for 60 seconds showed higher specific charge and the electrodes annealed at 4000C for 60 seconds showed lower capacity but higher retention. The SEM images (figure 4.3) of the coated electrodes before and after annealing showed that the reason for lower capacity was mainly due to the non uniformity of the coated surfaces. We observed cracks on the surface of the coated electrodes before and after annealing. Figure 4.1: Si/RGO Annealed Anode 48 Figure 4.2: Zirconium Coated Si/RGO Anode Figure 4.3: SEM of Zirconium Oxide Sprayed anode before heat Treatment 49 Figure 4.4: SEM of Zirconium Oxide Coated Anodes after heat treatment at 3500C for 60 seconds 50 To improve the surface morphology of the electrosprayed metal oxide coating on Si/RGO anodes, the solution spraying was modified. Dimethyl formamide solvent was added as a co- solvent to ensure wet spray coating on the electrode. S. No. Solvent to Oxide ratio Electrospraying Conditions 50:50 Zirconium Oxide and co1 solvent DMF Feedrate= 0.02 ml/min Voltage=20 kV Gas Assisted Electrospraying 50:50 Zirconium Oxide and co2 Feedrate= 0.05 ml/min solvent DMF Voltage=20 kV Gas Pressure= 28 psi 70:30 Zirconium Oxide and co3 solvent DMF Feedrate= 0.02 ml/min Voltage=20 kV Gas Assisted Electrospraying 70:30 Zirconium Oxide and co4 Feedrate= 0.05 ml/min solvent DMF Voltage=20 kV Gas Pressure= 28 psi Table 4.1: Parameters for electrospraying Zirconium Oxide onto Si/RGO From the SEM images (figure 4.5 and figure 4.6) of 50 weight percent co-solvent electrosprayed annealed anodes, we can observe that the coating consists of distinct units of metal oxide but not a uniform film of Zirconium Oxide on the anode. 51 Figure 4.5: SEM of 50 wt% of Zirconium Oxide Sprayed anode before heat Treatment Figure 4.6: SEM of 50 wt% Zirconium Oxide Coated Anodes after heat treatment at 3500C for 60 second 52 The gas-assisted electrospray uses a coaxial needle with the inner needle supplied with solution and outer shell with compressed air. The air pressure is ecternally controlled. The air pressure used for all our experiments was 28 psi. The air helps in atomizing the larger droplets into smaller ones. The electrostatic force assembles these smaller droplets onto the collector. But the gas-assisted electrospray of the metal oxide was not successful as we can observe from the SEM images (figure 4.7 and figure 4.8), the coating lacks uniformity and we can also see agglomoration of the droplets both before and after heat treatment. Figure 4.7: SEM of 50 wt% Zirconium Oxide gas-assisted sprayed anode before heat Treatment 53 Figure 4.8: SEM of 50 wt% Zirconium Oxide gas-assisted coated anodes after heat treatment at 3500C for 60 second It can be noted form the SEM images (figure 4.9 and figure 4.10) of 70 wt% Zirconium Oxide coating on the Si/RGO anodes, there was an improvement in the surface morphology and we were able to form a better anodic coating and also were able to achieve a wet spray. 54 Figure 4.9: SEM of 70 wt% Zirconium Oxide sprayed anode before heat Treatment Figure 4.10: SEM of 70 wt% Zirconium Oxide sprayed anode before heat Treatment 55 Specific/capacity/(mAh/g) / In order to achieve a wet spray coating an aprotic polar solvent N-Methyl Pyrrolidine was used along with DMF as a co-solvent. The co-solvents NMP and DMF were in the ratio of 1:9 respectively. Using NMP as a co-solvent helped in improving the battery performance of the coated anode. 4.2 Battery Testing of Zirconium Oxide Coated Anodes As discussed in section 4.1, initial land battery tests were performed using Lithium half cells to determine the temperature and the duration of the heat treatment for the Zirconium coated anodes. The initial tests were conducted with spin coated samples. From the battery tests shown in figure 4.11, we can observe that the anodes undergoing heat treatment at 3500C show higher specific capacity. This duration of the heat treatment wa/s 60 seconds. 2500 2000 1500 /New/Silicon /250/C/AlxOy/1.26mg/Si/RGO /350/C/ZrxO/1.35mg/Si_RGO /400/C/ZrxO/1.09mg/Si/RGO /400/C/ZrxO/1.33mg/Si/GO 1000 500 0 0 5 10 15 20 25 30 Cycle Figure 4.11: Cycle Performance of Si/RGO anodes and metal oxide coated anodes. The first 5 cycles tested at 100 mA/g and after 5 cycles the current density was maintained at 1 A/gm 56 Fig. 4.11 compares the cycling performance of coated electrodes heat treated at different temperatures via galvanic charging/discharging at a current density of 1A/gm. It must be noted that all the current density and specific capacity are calculated on the basis of total electrode materials rather than Si only and the first five cycles were tested under a low current density of 0.1A/gm as activation cycles. The land battery test results of both the electroprayed and spin coated 70 wt % Zirconium Oxide in DMF and NMP coated anode have been shown in figure 4.12. They exhibit lower specific charge than the Si/RGO anodes but exhibit good cyclic retention during the first 20 cycles. It can be noted that the electrosprayed coated anode samples have much higher specific charge capacity than the spin coated anodes. There is a dip in the specific charge after the 20th cycle both for the electrosprayed as well as the spin coated anode samples. 1200 1000 Electrospray C  oating Spin  Coating SPECIFIC  CHARGE(MAH/GM) 800 600 400 200 0 0 50 100 150 200 250 CYCLE  INDEX 57 Specific  Charge(mAh/gm) Figure 4.12: Cycle Performance of Electrosprayed and spin coated Zirconium Oxide and Si/RGO anodes the current density was maintained at 1 A/gm 1200 1000 800 600 400 Electrospayed Z  irconium O  xide  Coating 200 0 0 5 10 15 20 25 Cycle  Index Figure 4.13: Cycle performance of electrosprayed Zirconium Oxide onto Si/RGO anode until 20 cycles 4.3 Cyclic Voltammetry The cyclic voltammetric (CV) curves of Zirconium Oxide coated anodes at the scan rate of 0.5mV/s in the voltage window of 0.01-1.5V vs. Li+/Li can be seen in the figure 4.14. Catholic peak around 0.2V is observed, corresponding to the conversion of Si to LixSi alloy. The two anodic peaks around 0.38V and 0.58V are associated with the delithiation reaction with the phase transition from LixSi alloy to Si. From figure 4.15 we can see the comparison between the Si/GO 58 anode and Zirconium Oxide coated Si/GO anode. In the 1st cycle CV we can observe a peak at 1.1 V which corresponds to the Zirconium Oxide. 0.2 0 -­‐0.2 Current  (mA) -­‐0.4 -­‐0.6 -­‐0.8 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 Voltage (  V  vs  Li+/Li) 2 Figure 4.14: Cyclic Voltammetry(1st cycle) Curve of Metal Oxide Coated Anode We can observe that the peak current density of coated layer is substantially higher than that of Si/GO sample. This indicates that the conductivity of the coated metal oxide layer on Si/GO is higher than the uncoated layer, leading to much higher electrochemical activity. 59 0.4 0.2 0.0 Current  (mA) -­‐0.2 -­‐0.4 ZrO2  coated  Si/GO Si/GO   -­‐0.6 -­‐0.8 0 0.2 0.4 0.6 0.8 1 Volatge (  V  vs  Li+/Li) 1.2 1.4 1.6 Figure 4.15: Cyclic Voltammetry(2nd Cycle) Curve of Metal Oxide Coated Si/GO and Si/GO 60 CHAPTER 5 FUTURE WORK 5.1 Fiber Morphology post the half cell testing From the SEM images (figure 5.1), we can observe that after 200 cycles the coated layer gets cracked up and it can be noted that the coating becomes non-uniform. It can be noted that the coating is not adhering to the Si/RGO layer beneath it after 20 cycles. This accounts for the dip in the specific capacity after 20 cycles of cycle performance. We have seen that the electrosprayed coating shows good cyclic retention until 20 cycles. The morphology of the coated anode after 20 cycles is shown in figure 5.2. We can see that there are no cracks on the surface of the coated anode. Figure 5.1: SEM image of Zirconium Oxide coated Si/RGO anode synthesized using co-solvents NMP and DMF 61 Figure 5.2: SEM image of Zirconium Oxide coated Si/RGO anode after 200 cycles Figure 5.3: SEM image of Zirconium Oxide coated Si/RGO anode after 20 cycles 62 From the cyclic voltammograms we can observe that the coated layer on Si/GO has better electrochemical activity than the Si/GO sample. This suggests that we can in the future try coating the Si/GO samples with the metal oxide and anneal the whole coated sample. In the future we can also try coating commercial graphite anodes with Zirconium Oxide to further improve the stability. 63 CHAPTER 6 CONCLUSION In conclusion, a thinner and porous membrane system was developed using electrospraying technique and Polyethylene Oxide as a sacrificial polymer. This system was compared to the fiber system of Polyacrylonitrile (PAN). Silica precursors OPSZ and PSSQ were added to these polymer membranes to increase their ionic conductivity. SEM images show the uniformity of the formed film. Battery testing performed using coin cells of these films on the land battery show excellent retention and high specific capacity. In addition, it was observed that in comparison to the nanofiber system of PAN, PAN films as membranes for LIBs exhibit longer term cyclic stability. Performing Impedance spectroscopy studies on these samples showed us that they offer less resistance to the electrolyte flow compared to the PAN fibers. Dynamic Mechanical Analysis (DMA) results show that the PAN/PEO films exhibit less mechanical stability compared to the fibers possibly due to low molecular weight but by using this approach we will be able to reduce the thickness of the membranes and increase the cyclic retention. Metal Oxide have also been studied as potential anode coating materials. The transition metal oxide used in the study was Zirconium Oxide. The SEM images of the metal oxide coated anodes revealed a non-uniform coating on Si/RGO which was responsible for the low specific capacity of the half cells tested on the land battery. To improve the battery performance cosolvents Dimethyl formamide and N-Methyl Pyrollidine (NMP) were added. Using NMP facilitated us in attaining a wet spray. The battery test results with the half cells of the coated anodes showed good cyclic retention for the first 20 cycles. Electrosprayed samples showed higher specific capacity than the spin coated anodes. 64 REFERENCES 1. Huang ZM, Zhang YZ, Kotaki M, Ramakrishna S. Compos. Sci. Technol. 2003, 63, 22232253 2.   Formhals A. US patent 1,975,504, 1934 
 3.   Formhals A. US patent 2,160,962, 1939 
 4.   Formhals A. US patent, 2,187,306, 1940 
 5.   Formhals A. US patent, 2,323,025, 1943 
 6.   Formhals A. US patent, 2,349,950, 1944 
 7.   Simons HL. US patent 3,280,229, 1966 
 8.   Taylor GI. Proc R Soc London, Series A 1964;280:383 
 9.   Taylor GI. Proc R Soc London, Series A 1966;291:159 
 10.   Taylor GI. Proc R Soc London, Ser A 1969;313:453–475 
 11.   Baumgarten PK. J of Colloid and Interface Science 1971;36:71–79 
 12.   Larrondo L, Manley SJ. J Polym Sci: Polym Phys Ed 1981;19:909-920 
 13.   Larrondo L, Manley SJ. J Polym Sci: Polym Phys Ed 1981;19:921-932 
 14.   Larrondo L, Manley SJ. J Polym Sci: Polym Phys Ed 1981;19:933 
 15.   Srinivasan G, Reneker DH. Polym. Int. 1995 36 (2): 195-201 
 16.   Doshi J, Reneker DH. J. Electrostatics 1995;35 (2-3): 151-160 AUG 1995 
 17.   Reneker DH, Chun I. Nanotechnology 1996;7 (3): 216-223 
 18.   Fong H, Chun I, Reneker DH. Polymer 1999; 40 (16): 4585-4592 
 19.   Reneker DH, Yarin AL, Fong H, et al. J. Appl. Phys. 2000;87 (9): 4531-4547 
 65 20.   Hou HQ, Reneker DH. Adv. Mater. 2004;16 (1): 69-73 
 21.   Ostermann R, Li D, Yin YD, et al. Nano Lett. 2006;6 (6): 1297-1302 
 22.   Ko F, Gogotsi Y, Ali A, et al. Adv. Mater. 2003; 15 (14): 1161-1165 
 23.   Seoul C, Kim YT, Baek CK. J. Polym. Sci., Part B: Polym. Phys. 2003;41 (13): 1572- 1577 160 24.   Sen R, Zhao B, Perea D, et al. Nano Lett. 2004;4 (3): 459-464 
 25.   Fong H, Liu WD, Wang CS, et al. POLYMER 2002 ;43 (3): 775-780 
 26.   Zhou HJ, Kim KW, Giannelis E, et al. ACS SYMPOSIUM SERIES 918: 217- 
 230 2006 27.   Ji Y, Li BQ, Ge SR, et al. Langmuir 2006, 22, 1321-1328 
 28.   Mack JJ, Viculis LM, Ali A, et al. Adv. Mater. 2005;17 (1): 77-80 
 29.   Sun ZC, Zussman E, Yarin AL, et al. Adv. Mater. 2003;15 (22):1929-1932 
 30.   Zhang YZ, Huang ZM, Xu XJ, et al. Chem. Mater. 2004 ;16 (18):3406-3409 
 31.   Li D, Babel A, Jenekhe SA, et al. Adv. Mater. 2004;16 (22): 2062-2066 
 32.   Li D, Xia YN. Nano Lett. 2004;4 (5): 933-938 
 33.   Drew C, Wang XY, Bruno FF, et al. Compos. Interfaces 2005 ;11 (8-9): 711- 724 34.   Han GY, Guo B, Zhang LW, et al. Adv. Mater. 2006;18 (13): 1709-1712 
 35.   Wang YZ, Yang QB, Shan GY, et al. Mater. Lett. 2005;59 (24-25): 3046-3049 
 36.   Wang Y, Serrano S, Santiago-Aviles JJ. Synth. Met. 2003;138 (3): 423-427 
 37.   Kim C, Yang KS. Appl. Phys. Lett. 2003;83 (6): 1216-1218 
 38.   Choi SS, Chu B, Lee SG, et al. J. Sol-Gel Sci. Technol. 2004;30 (3): 215-221 
 66 39.   Dai HQ, Gong J, Kim H, et al. Nanotechnology 2002;13 (5): 674-677 
 40.   Li D, Xia YN. Nano Lett. 2003 ;3 (4): 555-560 
 41.   Son WK, Cho D, Park WH. Nanotechnology 2006;17 (2): 439-443 
 42.   Lee DY, Kim BY, Lee SJ, et al. J. Kor. Phys. Soc. 2006;48 (6): 1686-1690 
 43.   Bognitzki M, Frese T, Steinhart M, et al. Polym. Eng. Sci. 2001 ;41 (6): 982- 989 44.   F.Scholz(Ed.),Electroanalytical Methods,2nd,revised and extended edotion october 2009,Greifswald,Germany 45.   American Ceramic Society Bulletin, Vol. 86, No. 1 46.   Zong XH, Kim KS, Chu B et. al. Polymer 2002;43: 403-4412 
 47.   Shin K, Xiang HQ, Moon SI, et al. Science 2004; 306 (5693): 76-76 
 48.   Ruotsalainen T, Turku J, Heikkila P, et al. Adv. Mater. 2005;17 (8): 1048-1052 
 49.   Kalra V, Kakad PA, Mendez S, et al. Macromolecule 2006;39 (16): 5453-5457 
 50.   Huang CB, Chen SL, Reneker DH, et al. Adv. Mater. 2006;18 (5): 668-671 
 51.   Kim J-S, Reneker DH. Polymer Composites 1999;20(1):124–31 
 52.   Graham K, Ouyang M, Grafe T, et al. Fifteenth Annual Technical Conference & Expo of the American Filtration & Separations Society, Gavelston, Texas 2002
90. Shin C, Chase GG, Reneker DH. Colloids Surf., A Physicochem. Eng. Aspects 2005;262 (1-3): 211-215 53.   Zhou HJ, Green TB, Joo YL. Polymer 2006; 47 (21): 7497-7505 
 54.   Li WJ, Laurencin CT, Caterson EJ, et al. J. Biomed. Mater. Res. 2002;60 (4): 613-621 55.   Ji Y, Ghosh K, Shu XZ, et al. Biomaterials 2006 ;27 (20): 3782-3792 
 56.   Jia HF, Zhu GY, Vugrinovich B, et al. Biotechnol. Progr. 2002 ;18 (5): 1027-1032 57.   Wang XY, Lee SH, Ku BC, et al. 2002 ; J. Macromol. Sci., Pure Appl. Chem. 2002;A39 (10): 1241-1249 67 58.   Wang XY, Drew C, Lee SH, et al. Nano Lett. 2002;2 (11): 1273-1275 
 59.   Lyons J, Li C, Ko F. Polymer 2004;45(22):7597-7603 
 60.   Lyons J. PhD Dissertation, Drexel University, 2004 
 61.   Kim JS, Lee DS. Polymer. J. 2000; 32: 616-618 
 62.   Lee S, Obendorf SK. J. Appl. Polym. Sci. 2006;102 (4): 3430-3437 63.   Dalton PD, Klinkhammer K, Salber J, et al. Biomacromolecules 2006;7 (3): 686-690 64.   Zhou, X.; Yin, Y.-X.; Wan, L.-J.; Guo, Y.-G. Adv.Energy Mater. 2012, 2, (9), 1086- 1090. 65.   Chang, J.; Huang, X.; Zhou, G.; Cui, S.; Hallac, P. B.; Jiang, J.; Hurley, P. T.; Chen, J. Adv. Mater. 2014, 26, (5), 758-64. 66.   Jing, S. L.; Jiang, H.; Hu , Y. J.; Li, C. Z. J. Mater. Chem. A 2014, 2, 16360-16364. 67.   .Zhou, X.; Wan, L.-J.; Guo, Y.-G. Small 2013, 9, (16), 2684-2688. 68.   Hassan, F. M.; Chabot, V.; Elsayed, A. R.; Xiao, X.; Chen, Z. Nano Lett. 2013. 69.   Duan B, Dong CH, Yuan XY, et al. J. Biomater. Sci. - Polym. Ed. 2004;15 (6): 797-811 70.   Spasova M, Manolova N, Paneva D, et al. E-Polymers 2004;Art. No. 056 
 71.   Bhattarai N, Edmondson D, Veiseh O, et al. Biomaterials 2005;26 (31): 6176- 6184 72.   Babel A, Li D, Xia YN, et al. Macromolecules 2005 ;38 (11): 4705-4711 
 73.   Cha DI, Kim KW, Chu GH, et al. Macromol. Res. 2006;14 (3): 331-337 
 74.   Y. Li, Z.-Y. Fu, B.-L. Su, Adv. Fun. Mater. 2012, 22, 4634-4667. 75.   S. Agarwala, M. Kevin, A. S. W. Wong, C. K. N. Peh, V. Thavasi, G. W. Ho, ACS Appl. Mater. Interfaces 2010, 2, 1844-1850 76.   F. Cheng, Z. Tao, J. Liang, J. Chen, Chem. Mater. 2008, 20, 667-681 77.   C. Detavernier, J. Dendooven, S. P. Sree, K. F. Ludwig, J. A. Martens, Chem. Soc. Rev. 2011, 40, 5242-5253. 68 78.   S.–W. Kim, T. H. Han, J. Kim, H. Gwon, H.–S. Moon, S.-W. Kang, S. O. Kim, K. Kang, ACS Nano, 2009, 3, 1085-1090. 79.   Y.-S. Min, E. J. Bae, K. S. Jeong, Y. J. Cho, J.-H. Lee, W. B. Choi, G.-S. Park, Adv. Mater. 2003, 15, 1019-1022. 80.   D. J. Comstock, S. T. Christensen, J. W. Elam, M. J. Pellin, M. C. Hersam, Adv. Fun. Mater. 2010, 20, 3099-3105 81.   J. T. Korhone, P. Hiekkataipale, J. Malm, M. Karppinen, O. Ikkala, R. H. A. Ras, ACS Nano, 2011, 5, 1967-1974 82.   F. Li, X. Yao, Z. Wang, W. Xing, W. Jin, J. Huang, Y. Wang, Nano Lett. 2012, 12, 50335038 83.   S. Deng, M. Kurttepeli, S. Deheryan, D. J. Cott, P. M. Vereecken, J. A. Martens, S. Bals, G. Van Tendeloo, C. Detavernier, Nanoscale, 2014, 6, 6939. 84.   S. Deng, S. W. Verbruggen, Z. He, D. J. Cott, P. M. Vereechen, J. A. Martens, S. Bals, S. Lenaerts, C. Detavernier, RSC Advances, 2014, 4, 11648 85.   S. W. Verbruggen, S. Deng, M. Kurttepeli, D. J. Cott, P. M. Vereechen, S. Bals, J. A. Martens, C. Detavernier, S. Lenaerts, Appl. Catal. B: Environ. 2014, 160-161, 204-210 86.   J. S. Lee, B. Min, K. Cho, S. Kim, J. Park, Y. T. Lee, N. S. Kim, M. S. Lee, S. O. Park, J. T. Moon, J. Crystal Growth, 2003, 254, 443- 448. 87.   Y. Yang, L. Qu, L. Dai, T. S. Kang, M. Durstock, Adv. Mater. 2007, 19, 1239-1243 88.   Lewis, N. S. & Nocera, D. G. Powering the planet: Chemical challenges in solar energy utilization. Proc. Natl Acad. Sci. USA 103, 15729–15735 (2006) 89.   Hwang, Y. J., Boukai, A. & Yang, P. D. High density n-Si/n-TiO2core/shell nanowire arrays with enhanced photoactivity. Nano Lett. 9, 410–415 (2009). 90.   Sivula, K., Le Formal, F. & Gratzel, M. WO3–Fe2O3 photoanodes for water splitting: A host scaffold, guest absorber approach. Chem. Mater. 21, 2862–2867 (2009) 91.   Matsumura, M. & Morrison, S. R. Anodic properties of n-Si and n-Ge electrodes in HF solution under illumination and in the dark. J. Electroanal. Chem. 147, 157–166 (1983) 92.   Fan, F. R. F., Keil, R. G. & Bard, A. J. Semiconductor electrodes.48. Photo-oxidation of halides and water on N-silicon protected with silicide layers. J. Am. Chem. Soc. 105,220–224 (1983) 69 93.   Contractor, A. Q. & Bockris, J. O. M. Investigation of a protective conducting silica film on N-Silicon. Electrochim. Acta 29, 1427–1434 (1984). 94.   Siripala, W., Ivanovskaya, A., Jaramillo, T. F., Baeck, S. H. & McFarland, E. W. A CU2O/TiO2 heterojunction thin film cathode for photoelectrocatalysis. Sol. Energ. Mater. Sol. C 77, 229–237 (2003). 95.   James A. Laussert, Stephen E. Fosdick, and Richard M. Crooks, Langmuir, 2014, 30 (45), pp 13707–13715 96.   M. V. Reddy, G. V. Subba Rao, and B. V. R. Chowdari, Chem. Rev., 2013, 113 (7), pp 5364–5457 97.   Jian Liua, Xifei Li, Mei Cai, Ruying Li, Xueliang Suna, J. Liu et al. / Electrochimica Acta 93 (2013) 195–201 98.   M. Armand, J-M. Tarascon, Building better batteries, Nature 451 (2008) 652. 99.   J.-M. Tarascon, Key challenges in future Li-battery research, Philosophical Transactions of the Royal Society A 368 (2010) 3227. 100.   N.A. Kaskhedikar, J. Maier, The effect of the charging protocol on the cycle life of a Liion battery, Advanced Materials 21 (2009) 2664. 101.   E. Ferg, R.J. Gummow,A. deKock, M.M. Thackeray, Spinel anodes for lithium-ion batteries, Journal of the Electrochemical Society 141 (1994) L147. 102.   P. Liu, E. Sherman, M. Verbrugge, Electrochemical and structural characterization of lithium titanate electrodes, Journal of Solid State Electrochemistry 14 (2010) 585. 103.   C.Y. Ouyang, Z.Y. Zhong, M.S. Lei, Ab initio studies of structural and electronic properties of Li4Ti5O12 spinel, Electrochemistry Communications 9 (2007) 1107. 104.   T-F. Yi, Y. Xie, Y-R. Zhu, R-S. Zhu, H. Shen, Structural and thermodynamic stability of Li4Ti5O12 anode material for lithium-ion batteries, Journal of Power Sources 222 (2013) 448. 105.   C.H. Chen, J.T. Vaughey, A.N. Jansen, D.W. Dees, A.J. Kahaian, T. Goacher, M.M. Thackeray, Studies of Mg-substituted Li4−xMgxTi5O12 spinel electrodes (0 ≤ x ≤ 1) for lithium batteries, Journal of the Electrochemical Society 148 (2001) A102. 106.   G.N. Zhu, Y.G. Wang, Y.Y. Xia, Ti-based compounds as anode materials for Li-ion batteries, Energy & Environmental Science 5 (2012) 6652. 70 107.   T.F. Yi, L.J. Jiang, J. Shu, C.B. Yue, R.S. Zhu, H.B. Qiao, Recent development and application of Li4Ti5O12 as anode material of lithium ion battery, Journal of Physics and Chemistry of Solids 71 (2010) 1236. 108.   J. Li, Z. Tang, Z. Zhang, Controllable formation and electrochemical properties of onedimensional nanostructured spinel Li4Ti5O12, Electrochemistry Communications 7 (2005) 894. 109.   C. Lai, Y.Y. Dou, X. Li, X.P. Gao, Improvement of the high rate capability of hierarchical structured Li4Ti5O12 induced by the pseudocapacitive effect, Journal of Power Sources 195 (2010) 3676. 110.   E.M. Sorensen, S.J. Barry, H.K. Jung, J.M. Rondinelli, J.T. Vaughey, K.R. Poeppelmeier, Three-dimensionally ordered macroporous Li4Ti5O12: effect of wall structure on electrochemical properties, Chemistry of Materials 18 (2006) 482. 111.   J. Haetge, P. Hartmann, K. Brezesinski, J. Janek, T. Brezesinski, Ordered largepore mesoporous Li4Ti5O12 spinel thin film electrodes with nanocrystalline framework for high rate rechargeable lithium batteries: relationships among charge storage, electrical conductivity, and nanoscale structure, Chemistry of Materials 23 (2011) 4384. 112.   L. Cheng, J. Yan, G.N. Zhu, J.Y. Luo, C.X. Wang, Y.Y. Xia, General synthesis of carbon-coated nanostructure Li4Ti5O12 as a high rate electrode material for Li-ion intercalation, Journal of Materials Chemistry 20 (2010) 595. 113.   L. Zhao, Y.S. Hu, H. Li, Z. Wang, L. Chen, Porous Li4Ti5O12 coated with N-doped carbon from ionic liquids for Li-ion batteries, Advanced Materials 23 (2011) 1385 71