<p> Self-assembly of soft materials represents a compelling approach to realize a wide variety of useful nanostructured materials. In particular, self-assembly of block copolymers by microphase separation results thermodynamically in the formation of a range of nanostructures including lamellae, cylinders, gyroids and spheres. There is significant potential to use these structures in applications ranging from energy generation to water purification. Despite their significant potential however, the use of block copolymers in the aforementioned areas has been critically limited by general inability to precisely direct their self-assembly, i.e. to control the orientational and positional order of their self-assembled structures over device or application relevant length scales and geometries.</p><p> In this context, we explore two distinct approaches to attain advanced ability to control the block copolymer microphase. First, this dissertation explores the self-assembly and directed self-assembly of novel liquid crystalline block copolymers. Result are presented from a systematic series of experimental investigations of the phase behavior and directed self-assembly of rationally designed liquid crystalline block copolymers (LC BCPs) under magnetic fields and in the presence of engineered surfaces. We specifically designed a block copolymer platform comprising etchable poly(D,L-lactide) (PLA) with brush architecture and side chain cyanobiphenyl LC block that imparts magnetic anisotropy on the system. Interestingly, this class of brush-like block copolymers behave in accordance with the canonical phase behavior of the conventional linear coil-coil block copolymers. With inclusion of labile mesogen, the magnetic field response of the system was significantly enhanced due to the increased grain size and faster mobility. By adopting cross-linkable mesogen, the LC phase can be readily polymerized and subsequent etching of the PLA produces well-defined nanopores with controlled orientation. At higher blending stoichiometric ratio, the system transforms its morphology from hexagonal cylinders to face-centered cubic (FCC) spheres and, strikingly, we observe the alignment of FCC spheres regardless of the 3 dimensional symmetry of the cubic structure.</p><p> In the second part, we adopt the use of electrospray deposition and soft-shear laser zone annealing process as tools to direct the self-assembly of structurally complex thin films of block copolymers. Conventionally, block copolymers confined in thin film were examined based on the equilibrium structure as a result of a single annealing process. Here we propose non-equilibrium processing methods that enable us to achieve non-conventional morphologies. Sequential electrospray deposition (ESD) was adopted to form multi-layered BCP thin films which exhibit heterolattice structure that can be precisely tuned by kinetic parameters. We also examine pathway-engineered two-step processing, shear aligning followed by thermal annealing on a neutral substrate, to achieve biaxial alignment of the BCP cylinders array with minimum defect density. </p><p> Overall, this dissertation provides new insight regarding the self-assembly of LC brush block copolymers and their orientation in the presence of magnetic fields. Further, it establishes a new mechanism for controlling the orientation of these materials in thin films. The results of the research presented here are relevant for the use of block copolymers in lithography and membrane fabrication, among other areas.</p><p>

This thesis investigates fabrication methods for Nanoporous Gold (NPG), the complex nature of the film stress evolution in thin films during their fabrication and during surface charging. Fabrication of microstructures comprised of NPG requires a precise control over selective corrosion of the precursor alloy. In many designs, the precursor alloy is constrained to a substrate, and complex surface reconstruction during dealloying potentially leads to a high overall stress in the newly forming NPG. Hence, constrained NPG thin films or suspended NPG structures often develop cracks in fabrication. Similarly, nanovoids in thin film precursor alloys or low Au content in the precursor alloys may lead to fractures in the NPG thin films, which compromise integrity and functionality of the resulting architecture. Recently developed scalable electrochemical methods, for which the rate of removal of the less noble elements in the precursor alloy can be precisely controlled, produce crack-free blanket films constrained to a substrate. In this study, conventional as well as newly developed NPG fabrication techniques are assessed from a microscale fabrication perspective, with regard to the NPG product quality, means of tailoring the final porous structure and their compatibility with the standard microdevice fabrication techniques. A galvanostatic dealloying method is introduced, and shown to be effective in fabrication of constrained, crack-free, blanket NPG films on stiff substrates. Surface stress evolution during NPG fabrication is investigated using an optical multi-beam stress sensor (MOSS). The characteristic stress variation, range of the film stress, and effects of fabrication parameters are presented. The findings suggest that the film stress increases fast in the earlier stages of dealloying, proportional to the Ag dissolution rate. Film stress relief dominates during fabrication, with enhanced Au diffusion rate and increase of surface area. Upon surface charging NPG can expand significantly due to the capillary forces created at the double layer interface. A NPG based actuation device is designed to demonstrate its mechanical response with regard to surface charging. The performance of NPG proves to be sufficient for use in MEMS actuators.

Layer by Layer (LbL) deposition using primarily inorganic silica nanoparticles is employed for the modification of polymeric micro and ultrafiltration (MF/UF) membranes to produce thin film composites (TFC) with potential nanofiltration (NF) and reverse osmosis (RO) capabilities.. A variety of porous substrate membranes with different membrane surface characteristics are employed, but exhibiting in common that wicking of water does not readily occur into the pore structure, including polycarbonate track etched (PCTE), polyethersulfone (PES) and sulfonated PES (SPEES) MF/UF membranes. Both spherical (cationic/anionic) and eccentric elongated (anionic) silica nanoparticles are deposited using conditions similar to those reported by Lee et al.1 Appropriate selection of the pH's for anionic and cationic particle deposition enables the construction of nanoparticle only layers 100 -1200 nm in thickness atop the original membrane substrates. The surface layer thickness varies monotonically with the number of bilayers (anionic/cationic deposition cycles) as expected. The deposition process is optimized to eliminate drying induced cracking and to improve mechanical durability via thickness control and post-deposition hydro-thermal treatment. The hydrodynamic permeability of these TFC membranes is measured to evaluate their performance under typical NF operating conditions using dead-end permeation experiments and their performance compared quantitatively with realistic hydrodynamic models, with favorable results. For track etched polycarbonate MF substrates, surface modification causes a permeability reduction of approximately two orders of magnitude with respect to the bare substrates, to values comparable to those for typical commercial NF membranes. Good quantitative agreement with hydrodynamic models with no adjustable parameters was also established for this case, providing indirect confirmation that the LbL deposited surface layers are largely defect (crack) free. Imaging of our TFC membranes after permeation tests confirmed that no significant mechanical damage resulted, indicating integrity and robustness of the LbL deposited surface layers in typical applications. The selectivity of these novel TFC membranes was also tested using standard "rejection" tests normally used to characterize NF and RO membranes for their capabilities in typical applications, such as water softening or desalination. We report the dextran standards molecular weight "cut-off" (MWCO) using mixed dextrans from 1.5 to 500 KDa in dead-end stir cells, and the percentage of rejection of standard bivalent and monovalent salt solutions using steady cross flow permeation experiments. The results confirm rejection of at least 60% of even the smallest dextrans, an estimated dextran MWCO of 20 KDa, and rejection of 10% and 20% for monovalent (NaCl) and bivalent (MgSO4) salts, respectively, for all the TFC membranes studied, while the unmodified membranes showed no rejection capability at all. The work supports that nanoparticle based LbL surface modification of MF/UF membranes can produce filtration quality media for important water purification applications, such as nanofiltration (NF) softening processes, natural organic matter (NOM) elimination and possibly reverse osmosis (RO) desalination.

Carbon nanotube (CNT) is a hollow structure consisted by one-atom-thick sheet of carbon atoms, which can be considered as a rolled-up graphene sheet. The diameter and rolling angle (chirality) uniquely determines its electronic structure. Over two decades of study, due to the difficulty of synthesizing clean individual CNTs and the limitation of accurate chirality characterization, there are still unveiled questions towards the intrinsic properties of this 1-D material at single molecular level. In this thesis, I will discuss the approaches of fabricating chirality assigned CNT device and the experimental results of its optical and electrical properties. In the first part, I describe using 'fast heating' chemical vapor deposition (CVD) method to achieve the high quality suspended CNT growth. Combining Rayleigh and Raman spectroscopy, I demonstrate the accurate assignment of chirality for each suspended individual CNT. With the ability of chirality identification, a series of optical and electrical experiments were conducted on the selected CNTs of interest. In the following part, I first discuss the probe of many-body effect in a semiconducting CNT by observing the elastic scattering (Rayleigh spectra) with electrostatic gating. We found the dominant short-range interaction is reduced to 85% of its intrinsic strength for doping level of ρ=0.4e/nm, demonstrating the possible control of sub-band exciton resonance frequency without rely on Pauli-blocking effect in CNTs. In order to study the substrate effect in electrical transport of CNTs, I improved the transfer technique to accurately place individual CNT on a specific substrate. With this technique, I've achieved transferring individual CNT on 20µmx20µm thin layer of hexagonal-boron nitride (h-BN) substrate with a ± 5µm error. The low field electrical transport studies were conducted on both metallic and semiconducting CNTs with known chiralities on h-BN. Temperature dependent measurement shows the resistivity becomes super-linear around 250K, consistent with the prediction that the surface polar phonon of h-BN couples with electrons in CNT at higher phonon energy than SiO₂. Moreover, the FET devices of CNT on h-BN with graphite local back gate show hysteresis free feature in vacuum, and the subthreshold swing of 118mV/dec is comparable to high κ dielectric HfO₂ based device.

Graphene is an ideal candidate for use in flexible field-effect transistors (FETs) which require both high flexibility and high operating frequencies, because it offers exceptional electronic properties (room temperature mobility in excess of 10,000 cm² Vⁱ s⁻¹ and high saturation velocity of 3-7x10⁷ cm s⁻¹) as well as outstanding mechanical performance (strain limits up to 25%). Indeed, graphene FETs (GFETs) fabricated on rigid substrates from single crystals of mechanically exfoliated graphene have demonstrated unity power gain cut-off frequencies, fmax, up to 34 GHz, even at modestly scaled channel lengths of 600 nm. However, in order to realize commercial production of graphene-based technologies, it is essential to integrate large-area graphene produced by scalable synthesis methods into device fabrication. Chemical vapor deposition (CVD) offers a promising method to produce low-cost, large-area films of graphene, crucial for the commercial realization of graphene-based technologies. However, the electronic performance of CVD-grown graphene has remained problematic. Compared to exfoliated graphene, CVD graphene exhibits lower mobility, greater impurity doping, and higher asymmetry between electron and hole conduction, indicative of disorder and scattering processes that are not present in exfoliated samples. In order to achieve commercial scalability of high-performance graphene-based technologies, it is prerequisite to minimize disorder present in CVD graphene and achieve equivalent electronic properties to exfoliated graphene. In this work, I present a detailed study of the electronic transport behavior of CVD graphene in which the predominant sources of intrinsic disorder, grain-boundary scattering, is eliminated and extrinsic disorder, transfer induced contamination and substrate-induced scattering, are minimized. Grain boundaries within fabricated devices are eliminated by varying the CVD synthesis conditions to yield CVD graphene with large grain sizes, up to 250 μm in dimension. Process-related contamination is minimized by employing a novel dry-transfer technique that greatly reduces the extrinsic doping in CVD graphene devices, and samples are transferred onto hexagonal boron nitride (hBN), a dielectric which minimizes substrate-induced scattering and permits for the most precise assessment of the intrinsic performance of graphene. By minimizing the presence of these three predominant sources of disorder in CVD graphene, measurements presented in this work are the first demonstration that large-area graphene can not only be synthesized but also transferred onto arbitrary substrates while reproducibly achieving electrical performance comparable to that of high-quality exfoliated graphene. Related research demonstrates that the CVD graphene synthesized in this work additionally demonstrates equivalent mechanical properties to exfoliated graphene. After demonstrating that CVD graphene films can achieve both exceptional electronic and mechanical properties, the synthesis and transfer methods developed are subsequently applied to the fabrication of high-performance, flexible, radio-frequency FETs (RF-FETs), an application demanding both high-frequency operation and high mechanical flexibility. Methods to fabricate RF-FETs on flexible substrates using CVD graphene as the active channel material are presented. Devices fabricated with channel lengths of 500 nm show extrinsic values of unity current gain cut-off frequency, fT, and unity power gain cut-off frequency, fmax, up to 10.7 GHz and 3.7 GHz, respectively, and strain limits of 1.75%. By reducing the channel length to 260 nm, extrinsic values of fT and fmax increase to 23.6 GHz and 6.5 GHz, respectively, with intrinsic fmax = 28.2 GHz and strain limits of 2% attainable. Flexible graphene RF-FETs fabricated with channel lengths of 260 nm not only represent the highest values of fmax achieved in any flexible technology to date, but they also show an order of magnitude improvement in strain limit over flexible technologies demonstrating the next highest reported value of fmax. The structure of flexible GFETs is further improved by encapsulating the graphene channel in hBN dielectric layers and by implementing a self aligned fabrication scheme. RF-FETs fabricated with channel lengths of 375 nm demonstrate extrinsic cut-off frequencies fT and fmax of 12.0 GHz and 10.6 GHz, respectively, and intrinsic fT and fmax of 29.7 GHz and 15.7 GHz, respectively. The improved extrinsic cut-off frequencies indicate that using both a self-aligned fabrication scheme and hBN encapsulation are paramount to improving RF performance in flexible GFETs. Collectively, this work demonstrates that CVD graphene can achieve both outstanding electronic and mechanical performance and establishes CVD graphene as a competitive semiconductor technology for use in flexible RF-FETs. As such, it reveals the potential of CVD graphene as a material to enable a wide-range of flexible technologies requiring both high frequency operation and high mechanical flexibility.

Graphene, two dimensional lattice of covalent bonds of carbon atoms, has been studied as a prospective new material for the next generation. Pristine graphene, mechanically exfoliated graphene from graphite, has gained much attention due to its outstanding properties: conductivity, permeability, transparency, and mechanical stability. While pristine graphene has shown great promise as an innovative new material, the limitations from the randomness of sizes and domains are challenging for uniform mass production. In this dissertation, we present graphene produced by chemical vapor deposition (CVD) synthesis for producing designated sizes and domains. In order to prospect the utilization, the mechanical stability of CVD graphene should be determined. We first present mechanical properties of CVD graphene. Introducing transfer method, we present how to minimize damages on graphene during the fabrication. For the measurement of mechanical properties of CVD graphene, we introduce nanoindentation test with AFM and nanoindenter. Experimental results are demonstrated by the results of FEA analysis on the basis of nonlinear elastic behaviors. Through the experiment and simulation, we verify the ultra-high mechanical strength of CVD graphene. We also present defect-engineered graphene for the utilization. To determine the change of the status of defects on pristine graphene, we employed plasma etching to induce defects gradually. Through the observation of change of defects from sp3 type to sp2 type on pristine graphene, we understand how the phase changes depending on defects. Using nanoindentation, the mechanical strength of defective graphene is determined and we discuss its utilization based on the mechanical stability. We next exploit grains and grain boundaries of polycrystalline graphene. Transmission electron microscope (TEM) is used for precise observation of suspended membrane with grains and grain boundaries. Applying the same nanoindentation test, we compare the values of grain boundaries to pristine lattice in order to determine how grains and grain boundaries affect the ultra-high mechanical properties of graphene as defects. We finally present angular dependence of the mechanical properties of grains and grain boundaries. Although previous research reported the angular dependence of graphene regarding its mechanical strength, it was questionable that tilt angles among grains could not affect mechanical strength based on our previous experimental data. Therefore, here we reveal that how tilt angles among grains affect the mechanical properties. Furthermore, we investigate the crack propagation at rupture of graphene in both nanoindentation and e-beam exposure. Hence, we conclude the dissertation by a discussion of directions for future work, proposing well-stitched condition of graphene, and HR TEM for the verification of real structure of grain boundaries to apply into simulation. Therefore, this thesis is an arrangement of the outstanding mechanical properties of graphene from pristine graphene to CVD graphene in both small grain and large grain type, and from macroscopic region of interests over suspended membrane to microscopic observation such as the mechanical behaviors of grains and grain boundaries.

<p> UV-C LEDs in the range of 265&ndash;280 nm are needed to develop new disinfection and biotechnology applications. The market share for UV-C LED, versus UV-C lamps (Hg discharge and Xe), increased from 8% in 2008 ($240M) to 25% in 2018 ($810M). However, while low-pressure mercury lamps are ~30% energy efficient, the best commercial UV-C LEDs in the 265&ndash;280 nm range are ~2% energy efficient; InGaN blue LEDs are 80% energy efficient. Research on AlGaN LEDs has made significant progress into AlGaN material quality (including threading dislocation density and n-AlGaN electrical conductivity) but has lagged regarding light extraction efficiency. Light extraction from UV LEDs is limited by p-GaN absorption because of the lack of p-contact to p-AlGaN with AlN fraction (AlN content > 50%). Furthermore, AlGaN emitters at the 265&ndash;280 nm range emit 40&ndash;50% of their emissions as transverse magnetic (TM) waves, which are harder to extract than transverse electric (TE) waves. </p><p> SiC is an absorbing substrate that has been largely overlooked in developing UV-C LEDs, even though it has a small lattice mismatch with AlN (~1%) and a similar Wurtzite crystal structure and is more chemically stable. We demonstrate the first lateral thin-film flip-chip (TFFC) ultraviolet (UV) light-emitting diodes grown on SiC. UV LEDs were made at 310 nm, 298 nm, 278 nm, and 265 nm. </p><p> In this dissertation, we discuss the design, epi development, and fabrication of TFFC AlGaN LEDs with reflective p-contacts. The AlGaN:Mg growth temperature and the Mg doping profile in AlGaN:Mg were found to significantly impact the electroluminescence (EL) efficiency of the AlGaN MQWs. KOH roughening enhanced the light-extraction efficiency (LEE) by 100% and by ~180&ndash;200% for UV LEDs with 10 nm p-GaN and 5 nm p-GaN, respectively, without affecting the devices&rsquo; IV characteristics. The thin-film architecture led to a high LEE of about ~28&ndash;30% without LED encapsulation when used with LEDs with 5 nm p-GaN. The best light extraction efficiency in the literature is ~24% (without LED encapsulation) for a 275 nm flip-chip LED grown on PSS sapphire substrate. KOH roughening of AlN is discussed and is compared to KOH roughening of N-Face GaN. To advance LEE further, we attempted to develop LEDs with transparent current n-AlGaN spreading layers as well as highly doped n⁺-AlGaN tunnel junctions on top of UV-C LEDs. Reflective and ohmic n-contacts with low resistivities were developed for the n-Al_.58Ga_.42N regrown by MBE. Furthermore, a highly reflective MgF₂/Al omnidirectional mirror was developed, which can be used with n-contact microgrid to further enhance the LEE in UV-C LEDs with a transparent tunnel junction. </p><p>

Orientador: Débora Barros Barbosa / Banca: Wirley Gonçalves Assunção / Banca: Emerson Rodrigues de Camargo / Resumo: O objetivo do presente estudo foi avaliar a incorporação de nanopartículas de prata a uma resina acrílica para base protética por meio de testes de liberação em água deionizada por diferentes períodos e através de análises da distribuição e dispersão destas partículas na massa polimérica. Utilizou-se a resina acrílica termopolimerizável Lucitone 550 e as nanopartículas de prata foram sintetizadas por meio da redução dos íons prata do nitrato de prata pelo citrato de sódio. A forma e o tamanho das partículas foram confirmados por microscopia eletrônica de varredura (MEV) e de transmissão (MET), tendo-se obtido partículas na forma esférica e com tamanho médio de 60 nm. A resina acrílica foi proporcionada de acordo com as instruções do fabricante e a solução coloidal de nanopartículas de prata foi adicionada ao componente líquido da resina acrílica nas concentrações de 0,05%, 0,5% e 5% baseadas na massa do polímero. Após o processamento laboratorial, os espécimes foram armazenados em água deionizada à 37ºC por 7, 15, 30, 60 e 120 dias. As amostras de cada solução foram analisadas por espectroscopia de absorção atômica. Espécimes antes e após 120 dias de imersão em água foram analisados por MEV para caracterização morfológica do nanocompósito. Não houve liberação de prata detectável pelo aparelho, independentemente da concentração de colóide adicionada ao polímero e do tempo de imersão em água deionizada. As microscopias mostraram que, de uma forma geral, quanto menor a concentração de colóide de prata adicionada, menor a distribuição e maior a dipersão das partículas no polímero. Também ocorreu uma tendência das nanopartículas localizarem-se, principalmente, na superfície externa dos espécimes após 120 dias de armazenamento. Concluiu-se que houve incorporação das nanopartículas de prata ao polímero da resina... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The aim of this study was to evaluate the incorporation of silver nanoparticles to an acrylic denture base resin by testing the silver release in deionized water at different times, and through morphological analysis to check the distribution and dispersion of these particles in the polymer. The heat-polymerised acrylic resin Lucitone 550 was used and the silver nanoparticles were synthesized by reduction of silver nitrate with sodium citrate. The form and size of the particles were confirmed by scanning and transmission electron microscopy. Most of the particles showed a diameter of about 60 nm and spherical form. The acrylic resin was prepared in accordance with the manufacturers' instructions and silver nanoparticles solution was added to the monomer of the acrylic resin in the concentrations of 0.05wt%, 0.5wt% and 5wt%. The specimens were stored in deionized water at 37ºC for 7, 15, 30, 60 and 120 days, and each solution was analyzed by atomic absorption spectroscopic. The specimens were characterized by scanning electron microscopy before and after their immersion in water. Silver was not detected in deionized water regardless of the silver nanoparticles added to the polymer and of the storage period. The micrographs usually showed that when lower concentrations of silver nanoparticles were added, the particles distribution was reduced whereas their dipersion was improved into the polymer. Moreover, nanoparticles were mainly located at the surface of the specimens after 120 days of storage. The results showed that the silver nanoparticles were incorporated in the denture base resin polymer, and these nanoparticles were not detected in deionized water for up to 120 days. Moreover, the distribution and dispersion of the particles in the polymer changed with the silver concentration added and the period of storage. / Mestre

<p> The use of spectrofluorometers to examine nanomaterials is quite popular using either fluorescence or synchronous measurements. However, understanding how a material&rsquo;s optical properties can influence spectral acquisition are of great importance to accurately characterize nanomaterials. This dissertation presents a series of computational and experimental studies aimed at enhancing the quantitative understanding of nanoparticle interactions with matter and photons. This allows for more reliable spectrofluorometer based acquisition of nanoparticle containing solutions. </p><p> Chapter I presents a background overview of the works described in this dissertation. Correction of the gold nanoparticle (AuNP) inner filter effect (IFE) on fluorophore fluorescence using PEGylated AuNPs as an external reference method is demonstrated in Chapter II. The AuNP IFE is corrected to quantify tryptophan fluorescence for surface adsorbed proteins. We demonstrate that protein adsorption onto AuNPs will only induce ~ 20% tryptophan fluorescence reduction instead of the commonly assumed 100% reduction. </p><p> Using water Raman intensities to determine the effective path lengths of a spectrofluorometer for correction of fluorophore fluorescence is discussed in Chapter III. Using Ni(NO3)2 and K2Cr2O7 as Raman IFE references, the excitation and emission path lengths are found to exhibit chromophore and fluorophore independence, however path lengths are spectrofluorometer dependent. </p><p> Finally, ratiometric resonance synchronous spectroscopy (R2S2) is discussed in Chapter IV. Using a combination of UV-vis and R2S2 spectroscopy, the optical cross sections of a wide range of nanomaterials were determined. Also on-resonance fluorescence in solution is demonstrated for the first time. The nanoparticles discussed range from photon absorbers, scatterers, simultaneous photon absorbers and scatterers, all the way to simultaneous photon absorbers, scatterers, and emitters.</p>

The growing environmental concern over carbon dioxide emission has driven the demand for next-generation green vehicles like electric vehicles (EVs) and hybrid electric vehicles (HEVs). This in turn calls for higher capacity and higher energy rechargeable batteries for supporting long-distance driving of EVs. In this research, freestanding flexible Si nanoparticles-carbon nanotubes (SiNPs-CNTs) composite paper anodes for lithium-ion batteries (LIBs) have been prepared by a simple, inexpensive, and scalable approach of ultrasonication and pressure filtration. No conductive additive, binder or metal current collector is used. The composite using multi-walled CNTs (MWNTs) shows electrochemical properties superior to those using single-walled CNTs (SWNTs) or vertically aligned carbon nanotubes (VACNTs). The SiNPs-MWNTs composite (Si-MW) sample achieves first cycle specific discharge and charge capacities of 2298 and 1492 mAh/g, respectively. The first cycle irreversibility is compensated for by stabilized lithium metal powder (SLMP) prelithiation, leading to reduction of initial capacity loss from 806 to 28 mAh/g and increase of initial coulombic efficiency from 65% to 98%. The relationship between different SLMP loadings and cell performance has been established to understand the prelithiation mechanism of SLMP, optimize the construction of Si-based cells, and enable the exploration of novel cathode materials. The positive effect of FEC as electrolyte additive (10%) on the cyclability is verified. Through control of Si/CNT weight ratio, the optimal combination between the high capacity of SiNPs and the high electrical conductivity and structural stabilization ability of MWNTs is found in the case of the Si-MW 3:2 composite, resulting in improved cycling stability and high rate capability. The reversible capacity can be recovered to 1866 mAh/g when the current rate returns to 100 mA/g during cycling at current rate from 100 to 1000 mA/g. After 100 cycles, the electrode retains a reversible capacity of 1170, 850, and 750 mAh/g at the current rate of 100, 280, and 500 mA/g, respectively. FEC-based electrolytes using FEC as the co-solvent (50 wt%) are compared with the one using FEC as the additive. It is found that the EC-free FEC-based electrolyte achieves higher specific capacity and better capacity retention in terms of long-term cycling. After 500 cycles, the capacity retention of the cell using DEC-FEC (1:1) is increased by 88% and 60% compared to the cells using EC-DEC-FEC (45:45:10) and EC-FEC (1:1), respectively. Through SEM-EDX and XPS analyses, a possible reaction route of formation of fluorinated semicarbonates and polyolefins from FEC is proposed. The inferior cell performance related to the EC-containing electrolytes is likely attributed to the formation of excessive polyolefins which do not favor Li ion migration. The strategy of capacity-control cycling is employed to seek extended cycle life. Stable 326 charge-discharge cycles at designated capacity of 506 mAh/g are attained for the Si-MW 1:1 cell. A self-healing phenomenon is observed by studying the specific capacities and charge/discharge end voltage, and proposed as the possible mechanism behind the improved cycling stability. Prolonged cycling of over 500 cycles under capacity control (500 mAh/g) and the interesting pattern of variation in the discharge/charge end voltage are successfully reproduced with different electrode/electrolyte and current conditions. EIS and SEM-EDX analyses suggest that by setting the capacity/voltage limits for charge-discharge cycling, the growth of SEI can be limited. We believe the 3D network of MWNTs forms a continuous conductive pathway within the composite structure, which ensures sufficient electrical conductivity, holds the Si particles together, and alleviates the volume expansion of Si. Moreover, the freestanding feature of our electrode eliminates the non-active mass, giving rise to specific energy enhanced by 27% compared to current graphite-based cells by theoretical calculation. / A Dissertation submitted to the The Program in Materials Science and Engineering in partial fulfillment of the Doctor of Philosophy. / Summer Semester 2017. / June 23, 2017. / carbon nanotubes, freestanding electrodes, high capcity, lithium ion batteries, silicon nanoparticles / Includes bibliographical references. / Jim P. Zheng, Professor Co-Directing Dissertation; Richard Liang, Professor Co-Directing Dissertation; Susan Latturner, University Representative; Albert E. Stiegman, Committee Member; Mei Zhang, Committee Member; Kenneth Hanson, Committee Member.