Processing of Carbon–Silicon Carbide Hybrid Fibers

Home , Carbide, Carbon

PROCESSING OF CARBON–SILICON CARBIDE HYBRID FIBERS

Dissertation

Submitted to

The School of Engineering of the

UNIVERSITY OF DAYTON

In Partial Fulfillment of the Requirements for

The Degree of

Doctor of Philosophy in Engineering

Saja M. Nabat Al-Ajrash

Dayton, Ohio

December 2019

PROCESSING OF CARBON–SILICON CARBIDE HYBRID FIBERS

Name: Al-Ajrash, Saja M. Nabat

APPROVED BY:

______

Khalid Lafdi, Ph.D. Donald A. Klosterman, Ph.D. Advisory Committee Chairman Committee Member Professor, Wright Brothers Endowed Associate Professor, Torley Endowed Chair in Nanomaterials Chair in Composite Materials Department of Chemical and Materials Department of Chemical and Materials Engineering Engineering

______

Erick S. Vasquez, Ph.D. Youssef Raffoul, Ph.D. Committee Member Committee Member Assistant Professor Professor Department of Chemical and Materials Department of Mathematics Engineering

______Robert J. Wilkens, Ph.D., P.E. Eddy M. Rojas, Ph.D., M.A., P.E. Associate Dean for Research and Dean, School of Engineering Innovation Professor School of Engineering

Saja M. Nabat Al-Ajrash

2019

iii

ABSTRACT

PROCESSING OF CARBON–SILICON CARBIDE HYBRID FIBERS

Name: Al-Ajrash, Saja M. Nabat University of Dayton

Advisor: Dr. Khalid Lafdi

Two processing methods were used to fabricate carbon-SiC hybrid fibers. The first approach utilized various ratios of a silicone polymer (polydimethylsiloxane) as a silicon source and polyacrylonitrile (PAN) as a carbon source. The second approach used a mixture of silicon nanoparticles at various concentrations in PAN. The two formulations were independently converted into fiber form using an electrospinning process. Nanofibers with several hundred of nanometer diameters were successfully fabricated and subsequently stabilized and carbonized at 1000 oC. In the first approach, three phases were found to be present: nanocrystalline SiC, turbostratic carbon and SiOC. The resulting fibers showed a core-skin structure with the skin rich in carbon and a core dominated by silicon- based phases in the form of SiC or SiOC phases. A significant improvement in both tensile strength and elastic modulus was observed for C-SiC hybrid fibers as compared to SiC- free carbon fibers produced in this study.

In the second approach, one of the key issues identified for the study was particle distribution during the electrospinning. The analysis revealed that large silicon particles were located in the skin and the smaller ones were located at the core. The migration rate from the core was the fastest for large particles and was diminished as the particles became smaller in size due to inertial effects. The threshold for the Stokes number was found to be around 2.2 x 10-4 with a critical particle size of 1.0 x 10-7m in diameter. The current results

are very promising, as it demonstrated a novel way for the fabrication of PAN/SiC ceramic nanofiber with a gradient of particle size and properties from the skin to the core.

In addition, the ratio of Si to carbon precursor and heat treatment procedure were optimized to process hybrid nanofibers with high oxidation resistance. After pyrolysis to

1250 oC, the nanofibers showed two-dimensional ordered carbon and SiC nano phases.

Samples with 90 wt%PAN/10 wt% Si showed approximately four-time improvement in char yield as compared with 100 wt% PAN. It was concluded that the SiC phase played a major role in ordering the carbon phase. The carbon and SiC crystallinities had a great impact on improvements in the mechanical properties and the oxidation resistance, respectively. The SiC grain growth was predicted using Scherrer formula, and its exponent was found to be around n = 4 with activation energy around 35 KJ/mol.K. For such growth, the dominant grain growth mechanism was concluded to be grain boundary diffusion.

Furthermore, a complete transformation of Si to SiC occurred at 1250 °C. However, for heat treatments below 1000 °C, three phases, including Si, C, and SiC were present.

The effect of microstructural changes due to the heat treatment on oxidation resistance was determined using thermogravimetric analysis. The char yield showed linear increasing growth as the carbonization temperature ranged from 850 °C to 1250 °C. Increasing the holding times at higher temperatures produced a significant increase in thermal stability because of SiC grain growth. At long holding times, the SiC phase was observed to function as both an antioxidation coating and a mechanical reinforcing phase. Such structural changes lead to changes in fiber mechanical properties. The tensile strength was the highest for fibers carbonized at 850 °C, while the modulus increased monotonically with increasing carbonization temperature.

Dedicated to my

Parents

Modher Nabat & Kawakib Hameed

Brothers

Ahmed and Manar

Sisters

Zahraa and Reyam

Lovely Daughters

Sarah and Sally

ACKNOWLEDGEMENTS

My special thanks to Dr. Khalid Lafdi, my advisor, for providing the time, material and equipment necessaries, for guiding me to be a competent researcher, for directing this dissertation and bringing it to an accomplishment patiently and professionally, for being an inspiring mentor.

I would also express my appreciation to everyone who helped me. This includes but not limited to my parents, my committee, Faculty including: Dr. Charles Browning,

Dr. Donald A. Klosterman, Dr. Erick S. Vasquez, Dr. Youssef Raffoul, Dr. Cao Li, Dr.

Francisco Chinesta, Dr. Abdulaziz Baçaoui, and Friends including: Robyn Braford,

Qichen Fang, Yuhan Liao, Jean-Baptiste Dumuids, Nuha Al Habis, Chang Liu, Robert

Busch, Rose Eckerle and Shuangshan Li.

Last but foremost, I would like to thank my family, especially my parents (Dr.

Modher Nabat) and Kawakib. No words expressing my love and gratitude to the best, first and wisest man I have ever seen, my father. His hard working, ethics, and courage shaped me into what I am today. The sacrifices from my mother, including quitting her job to support us are never limited. Finally, I would like to thanks my six-year-old daughter,

Sarah, and three-year-old daughter, Sally. Just looking to their innocent face inspires and motivates me to work hard and being a successful woman who they admire. I also genuinely apologize for spending less time with them and, hopefully, when they understand what life is, will help them to appreciate the struggle I went through.

vii

TABLE OF CONTENTS

ABSTRACT ...... iv

DEDICATION ………………………………..…………………………..…………………vi

ACKNOWLEDGEMENTS ...... vii

LIST OF FIGURES ...... xi

LIST OF TABLES ...... xv

LIST OF ABBREVIATIONS AND NOTATIONS ...... xvi

CHAPTER I INTRODUCTION ...... 1

1.1 Introduction ...... 1 CHAPTER II MATERIALS AND BACKGROUND ...... 5

2.1 Carbon-Based Material ...... 5 2.1.1 Carbon Fibers ...... 7 2.2 Hybrid Carbon- Preceramic Polymer Materials ...... 16 2.3 Hybrid Carbon Materials with Solid Particles ...... 18 CHAPTER III FABRICATION OF HYBRID NANOFIBER USING POLYDIMETHYLSILOXANE AND POLYACRYLONITRILE POLYMER BLENDS ...... 20 3.1 Introduction ...... 20 3.2 Materials and Method...... 24 3.2.1 Electrospinning of PDMS/PAN Nanofiber ...... 24 3.2.2 Stabilization and Pyrolysis of PDMS/PAN Nanofiber ...... 25 3.2.3 Characterizations ...... 26 3.3 Results and Discussion ...... 27 3.3.1 Nanofiber Morphologies ...... 27 3.3.2 Energy Dispersive Spectroscopy ...... 29 3.3.3 Fourier Transform Infrared Spectroscopy ...... 30 3.3.4 Raman Spectroscopy ...... 32

viii

3.3.5 X-Ray Diffraction ...... 33 3.3.6 Transmission Electron Microscopy ...... 34 3.3.7 Tensile Test ...... 36 3.4 Conclusions ...... 38 CHAPTER IV EXPERIMENTAL AND NUMERICAL INVESTIGATION OF THE SILICON PARTICLE DISTRIBUTION IN ELECTROSPUN NANOFIBER ...40

4.1 Introduction ...... 40 4.2 Experimental Procedure ...... 43 4.3 Numerical Modeling ...... 45 4.3.1 Meshing ...... 45 4.3.2 Governing Equations ...... 46 4.4 Results and Discussions ...... 48 4.4.1 SEM Characterization ...... 48 4.4.2 TEM Characterization ...... 50 4.4.4 Modeling Results...... 50 4.5 Conclusions ...... 55 CHAPTER V THE ROLE OF CARBON AND SIC CRYSTALLINITIES IN THE OXIDATION AND MECHANICAL PROPERTY IMPROVEMENT OF HYBRID NANOFIBER ...... 56

5.1 Introduction ...... 56 5.2 Experimental Setup ...... 59 5.2.1 Materials ...... 59 5.3 Results and Discussion ...... 61 5.3.1 Scanning Electron Microscopy ...... 61 5.3.2 Mechanical Test ...... 64 5.3.3 Thermogravimetric Analysis ...... 65 5.3.4 Raman Spectroscopy ...... 67 5.3.5 Transmission Electron Microscopy ...... 70 5.3.6 X-Ray Diffraction ...... 71 5.4 Conclusions ...... 76

CHAPTER VI HYBRID CARBON-SIC NANOFIBER WITH IMPROVED OXIDATION RESISTANCE ...... 78

6.1 Introduction ...... 78 6.2 Experimental Setup ...... 80 6.3 Results and Discussion ...... 82 6.3.1 Scanning Electron Microscopy ...... 82 6.3.2 Nanofiber Structural Examinations ...... 83 6.3.3 Thermogravimetric Analysis ...... 86 6.3.4 Mechanical Test ...... 88 6.4 Conclusions ...... 89 CHAPTER VII CONCLUSIONS AND RECOMMENDATIONS ...... 90

7.1 Conclusions ...... 90 7.2 Recommendations and Future Work ...... 92 REFERENCES ...... 93

LIST OF FIGURES

Figure 1 Statistics of the Number of Publications with Inorganic-Organic Hybrid

Materials in the Last 11 Years...... 2

Figure 2 Organic-Inorganic Hybrid Materials at Different Structural Levels...... 3

Figure 3 Carbon Hybridization States...... 5

Figure 4 Graphene Sheets Morphologies...... 6

Figure 5 Macromorphology of Two Polymeric Phases (PAN:PMMA) a) 5:5, b)7:3,

and c) 9:1 . d, e,f) Micrograph Images of Pyrolyzed Fiber at 1000 oC.

PAN:PMMA = d) 5:5, e) 7:3, And f) 9:1. g) TEM Images of the Sample (d),

Showing Linearly Developed Hollow Cores Along the Fiber Length.

The Inset is a Magnified TEM Image ...... 7

Figure 6 Carbon Fiber Precursors...... 8

Figure 7 Carbon Fibers Thermal Treatment Procedure...... 9

Figure 8 CNF with Three Distinct Aspect Ratios a) Platelet, b) Tubular, c) Fishbone .... 10

Figure 9 Tensile Strength of Different Materials ...... 12

Figure 10 Graphite and Turbostratic Carbon Structure...... 13

Figure 11 TEM Images of a) PAN-5% MWCNT Nano-Fibers Carbonized at 750 oC,

in a Region without Carbon Nanotubes, b) PAN-25% MWCNT Nanofiber

Carbonized at 1100 oC ...... 14

Figure 12 a) Specific Capacitance vs Discharge Current Density and b) Ragone Plot

in Organic Electrolyte (H2SO4 And KOH). The Results Obtained with Activated

Carbon ...... 15

Figure 13 Carbon Nanofiber Anode Made from PAN/PLLA a) Charge-Discharge

Curve, b) Cycling Performance ...... 15

Figure 14 Inorganic Particles Incorporation in Polymers Structure ...... 19

Figure 15 Viscosity- Shear Rate for Different Polymers Solutions...... 25

Figure 16 SEM Micrographs of As-spun Nanofibers with Different PAN:PDMS

Ratios a) PAN10, b) PAN8:PDMS2, c) PAN7:PDMS3, d) PAN5:PDMS5...... 27

Figure 17 SEM Micrographs of PAN8:PDMS2 Nanofibers at Different Treatments

Temperatures a) 225 oC, b) 600 oC, c) 1000 oC...... 28

Figure 18 Average Diameter of Electrospun Nanofibers for Different Heat Treatment. . 29

Figure 19 Si Distribution across Fiber Diameter for PAN8:PDMS2 Sample...... 30

Figure 20 FTIR Spectrum of the As-spun Nanofibers...... 30

Figure 21 FTIR Spectrum of Nanofibers with Composition of PAN7:PDMS3...... 31

Figure 22 Raman Spectroscopy for Pyrolyzed Samples at 1000 oC...... 33

Figure 23 XRD for PAN8:PDMS2 at 1000 oC...... 34

Figure 24 TEM Images a) Bright Field Image, b) Dark Field Image, c) Selected Area

Diffraction (SAD) of the PAN8:PDMS2...... 35

Figure 25 Lattice Fringes Imaging of Skin Surface...... 35

Figure 26 Lattice Fringes of Nanofiber Core a) The Small Arrows Indicate the Carbon

Basic Units; b) Large Arrows Indicate the Existence of 111 SiC...... 36

Figure 27 The Effect of PDMS Percentage on Fiber a) Tensile Strength, b) Young’s

Modulus...... 37

Figure 28 Stress-Strain Curve at Different Heat Treatments for PAN8:PDMS2...... 38

Figure 29 Mesh Generation...... 46

xii

Figure 30 SEM Micrograph with Diferente Si Weight Percent a) 0 wt%,

b) 0.5 wt%Si, c) 1 wt%Si, d) 5 wt%Si, e) 10 wt%Si...... 48

Figure 31 Silicon Concentration Effect on Fiber Average Diameter...... 49

Figure 32 TEM Image of a Single Fiber...... 50

Figure 33 Particle Position from the Inlet at Different Times a) 6.0e-8s, b) 3.0e-7s, c)

6.6e-7s, d) 7.2e-7s, e) TEM Image for the Right Side of the Fiber, f) 9.6e-6s...... 51

Figure 34 Particle Position with Time for a Range of Particles...... 52

Figure 35 Particle Size Effect on Particle Lateral Velocity...... 53

Figure 36 Particle Size Effect on Stokes Number...... 54

Figure 37 Stokes Number Influence on Particles Lateral Velocity...... 54

Figure 38 SEM Micrograph for As-spun Fibers Mat with Different Si Concentrations

a) 0 wt% Si, b) 0.5 wt% Si, c) 1 wt % Si, d) 5 wt% Si e) 10 wt%Si...... 62

Figure 39 SEM Micrograph for Carbonized Fibers Mat with Different Si concentrations

a) 0 wt% Si, b) 0.5 wt% Si, c) 1 wt % Si, d) 5 wt% Si e) 10 wt%Si...... 63

Figure 40 The Effect of Silicon Content on Fiber Diameter...... 63

Figure 41 The Effect of Si Content on Fiber Mats a) Strength b) Young’s Modulus...... 65

Figure 42 The Relationship between Si Content And Weight Loss for Carbonized

Fibers At 1250 oC...... 66

Figure 43 TGA Results a) the Effect of Si Content on Char Yield, b) The Relationship

between the Silicon Content and Oxidation on Onset Temperature...... 67

Figure 44 The Effect of Silicon Content on the Raman Shift...... 69

Figure 45 The Effect of Silicon Content on ID/IG Ratio...... 69

xiii

Figure 46 Carbonized fibers at 1250 oC Pristine PAN a) Bright Field Image b) 002

Lattice Fringes of Carbon Phase Developed in the Pristine PAN Sample...... 70

Figure 47 Carbonized Fibers at 1250 oC Silicon Added PAN Nanofiber a) Bright Field,

b) 002 Lattice Fringes of the Carbon Phase...... 71

Figure 48 XRD Peaks for Different Precursors...... 72

Figure 49 a) The Relationship between the Crystal Size and Carbonizing Temperature

and Holding Times, b) The XRD Peaks for 10.0%Si At 1250 oC ...... 74

Figure 50 The Relationship between ln(K) and 1/T...... 75

Figure 51 TGA Results for 10.0 wt%Si at Different Temperatures...... 76

Figure 52 SEM Images of Treated Fibers (Holding Time four Hours) at a) 25 °C, b)

850 °C, c) 1000 °C, d) 1250 °C...... 82

Figure 53 Relationship between Fibers Diameter and Heat Treatment Temperature ...... 83

Figure 54 XRD Results of Nanofiber Mat Heat Treated at a) Three Distinct Tempera-

tures and One-Hour Holding Time, b) 1000 °C at Three Distinct Holding Times. .. 84

Figure 55 Raman Spectroscopy of Carbon Fibers at Three Distinct Temperatures...... 85

Figure 56 TGA Curves of Hybrid Fibers at Different Soaking Times at 1250 °C...... 87

Figure 57 SiC Nucleation and Growth from Si and C Phases a) TEM Image of As-spun

PAN-Si Nanofiber, b) As-Spun PAN-Si Nanofiber Illustration, c) SiC Nucleation,

d) SiC Growth, e) Conversion of All Si to SiC, f) SiC Grain Coarsening...... 88

Figure 58 The Relationship Between Mechanical Properties and Heat Treatment

Temperature...... 89

xiv

LIST OF TABLES

Table 1 Carbon Fiber Production per Ton for Distinct Industries ...... 8

Table 2 Weight Contents of PDMS and PAN...... 24

Table 3 PAN to Si Ratios...... 44

Table 4 PAN to Si Ratio...... 59

Table 5: The Relationship between the SiC Crystal Size and Initial Si Concentration .... 73

Table 6: Kinetic of Grain Growth ...... 75

Table 7 Influence of Pyrolysis Temperature and Soaking Time on Hybrid Carbon

Fibers Char Yield...... 87

LIST OF ABBREVIATIONS AND NOTATIONS

PDMS Polydimethylsiloxane

CNFs Carbon Nano Fibers

PMMA Poly Methyl Methacrylate

CVD Chemical Vapor Deposition

AN/IA Acrylonitrile / Itaconic Acid

SiC Silicon Carbide

C Carbon

Si Silicon

AN/AM Acrylonitrile /Acrylamide

DMF Dimethylformamide

MWCNT Multi-Walled Carbon Nanotubes

CNT Carbon Nanotube

AC Activated Carbon

PLLA Polylactic Acid

CF Carbon Fiber

TGA Thermogravimetric Analysis

DTA Differential Thermal Analysis

FTIR Fourier Transform Infrared Spectroscopy

TMDSC Temperature Modulated Differential Scanning Calorimetry

PDCs Polymer Derived Ceramics

PCmS PolyCarboMethylSilane

xvi

PS PolyStyrene

SEM Scanning Electron Microscope

EDS Energy Dispersive Spectrometer

TMA Thermal Mechanical Analysis

XRD X-ray Diffraction

SAD Selected Area Diffraction

BSUs Basic Structural Units

PEO Polyethylene Oxide

xvii

CHAPTER I

INTRODUCTION

1.1 Introduction

Traditional materials such as metals, ceramics, and polymers are unable to meet all technological requirements [1]. The efforts made in the past to combine different material properties in one individual material were started at the beginning of the industrial era [2].

Extensive work was performed on alloying different metals to create metal alloys that display improved properties. Later, as weight became a critical design criteria, metal alloys were replaced by fiber-reinforced composites. Such composites showed improved physical properties with better oxidation and corrosion resistance than metals. Combining the advantages of both materials within the same system resulted in hybrid materials [3]

Hybrid materials defined as “materials obtained through interaction of chemically different constituents (components), usually organic and inorganic, which form a specific

(crystal, spatial) structure that is different from the structures of initial reagents, but often inherit certain motifs and functions of the original structures”[4].

The term “hybrid organic-inorganic” material is brand new (Figure 1) which enables the researchers to fabricate more innovative structures for high-tech applications. The interest in hybrid materials came from their ability in engaging two or more chemical and structural characteristics in order to create new materials with tailored properties. For example, in organic-inorganic hybrids, the inorganic constituents are the key component in improving the mechanical and thermal properties. On the contrary, the organic

component can be added to improve optical, electrical, magnetic, and thermal properties

[5][6].

One of the cutting-edge in hybrid materials technology is the smart use of polymers and additives. The feasibility of fabrication, using low processing temperature, low cost are very attractive features [7]. In the last two decades, the number of patents and research evolving on hybrid organic/inorganic materials has increased significantly (Figure 1).

Figure 1 Statistics of the number of publications with inorganic-organic hybrid materials in the last 11 years (reprinted with permission [5]). Figure 2 shows the wide variety of organic-inorganic hybrids at different scales [8].

Multi-functional hybrid materials exhibit unique and novel properties as a result of a synergetic combination of materials [9][5]. Their properties are not only a combination of the individual components; but also the interface and grain. The nature of the interface can classify hybrid materials into two groups. The first one, the organic-inorganic bond is very weak like Van der Waals or ionic bonds; while in the second group, the constituent connected via strong chemical bond like covalent bond [10]. In addition, hybrid materials fabrication could be achieved at various scales as shown in Figure 2. 2

Figure 2 Organic-inorganic hybrid materials at different structural levels (adapted from [8]).

The overall objective of this research is to find an innovative approach to fabricate hybrid C-SiC fiber using carbon and silicon-based precursors. This Ph.D. dissertation is structured as follows: In Chapter One, hybrid material categories, applications, and advantages over one component and composite materials are introduced. In Chapter Two, a literature review of distinct types of carbon and silicon-based materials is given this chapter highlights the great advantages and disadvantages of carbon materials. In addition, many approaches to overcome carbon-based material limitations are discussed. Chapter

Three focuses on the fabrication of doped carbon nanofibers with a silicon-based polymer, while Chapter Four describes the fabrication of hybrid fibers using solid silicon nanoparticle. The silicon nanoparticles distribution within nanofiber is predicted numerically using the finite element method (FEM) by utilizing COMSOL multiphysics

software. The numerical results are validated using transmission electron microscopy images. Chapter Five describes how the hybrid fibers from chapter four were tested for oxidation resistance. The thermal durability of the various fiber formulations is studied as a function of heat treatment temperatures and silicon concentrations. The SiC nucleation and growth kinetics are predicted using the Scherrer model. Finally, in Chapter Six, the role of furnace temperature and holding times is investigated to optimize the silicon concentration (observed from chapter five).

CHAPTER II

MATERIALS AND BACKGROUND

2.1 Carbon-Based Material

Carbon has the ability to form covalent chemical bonds with variable elements possessing a wide range of electronegativities [11]. In addition, the wide range of hybridization states of carbon (sp, sp2, sp3) gives this material distinct and unique structures as shown in Figure 3. Carbon-carbon nano-structures which based on the covalent bonds linking the carbon can be categorized into two groups according to the hybridization state.

The first group includes graphitic structure which consists of graphene, carbon nanotubes, carbon onion, graphene nano-sheets, and C-dots. Such structures are mainly made from sp2 hybridization in the form of the hexagonal honeycomb lattice. However, the sp3 hybridization occupies the defects sites or edges. Secondly, another group has a combination of sp2 and sp3 hybridization. The nano-diamond and nano C-dot belong to this category [12].

Figure 3 Carbon hybridization states.

The exceptional chemical and physical properties of carbon-based material make it a target for many research areas and industrial applications like electronics, strong 5

materials and biomedical materials [13][14] [15]. The basic hexagonal honeycomb building unit in most carbon-based materials is graphene. The graphene sheets could be folded to form a 0D sphere or 1D nanotube or could be connected with weak Van der Waals force to form graphite as shown in Figure 4.

Figure 4 Graphene sheets morphologies (reprinted from permission [16]).

Most of the carbon-based materials can be used as nano-fillers to form polymeric materials which result in new nano-composites with enhanced thermal, mechanical, and electrical properties [17]. The resulted high-density defects from the substitution of graphene unit cell with five/seven rings or any atom create a porous carbon structure. The resulted porous material has sp2 structure with a high surface area which increases its reactivity [11].

Furthermore, porous carbon material has massive applications, in particular, as a separation media due to its tunable physicochemical properties [18]. Carbon-based materials provide 3D porous structure in conjunction with the required electrical properties

which makes it a functional material for sensing, photochemical applications, water processing, electrochemical storage/ conversion [19].

Kim et al [20] introduced porous carbon nanofiber (CNF) by adding polymethyl methacrylate (PMMA) polymer to polyacrylonitrile (PAN). The porosity or hollow cores density increased with increasing PMMA concentration as follow:

Figure 5 Macromorphology of two polymeric phases (PAN:PMMA) a) 5:5, b)7:3, and c) 9:1 . d, e,f) micrograph images of pyrolyzed fiber at 1000 oC. PAN:PMMA d) 5:5, e) 7:3, and f) 9:1. g) TEM images of the sample (d), showing linearly developed hollow cores along the fiber length. The inset is a magnified TEM image (reprinted from permission [20]).

2.1.1 Carbon Fibers

Carbon fiber (CF) displays high thermochemical stability in the nonoxidative environment. It has high thermal and electrical conductivities, and excellent creep resistance [21][22]. It has been selected in many applications like aerospace, automobile, military, turbine blades, construction, energy conversion and storage, and self-sensing devices. CF can be fabricated easily using polymeric precursors like acrylic, cellulosic, pitch, and PAN precursors [23][24] by subsequent heat treatment, as shown in Figure 6.

Table 1 Carbon fiber production per ton for distinct industries Industry PAN (tons) Pitch (tons) Toray Industries (small tow) 9100 Toho Tenax (Teijin) (small/large tow) 8200 Mitsubishi Rayon/Grafil (small tow) 4700 Zoltek (large tow) 3500 Hexcel (small tow) 2300 Formosa Plastics (small tow) 1750 Cytec Engineered Materials (small tow) 1500 360 SGL Carbon Group/SGL Technologies (large tow) 1500 Mitsubishi Chemical 750 Nippon Graphite Fiber 120

Figure 6 Carbon fiber precursors.

Based on the market data, PAN-based CF is the most commonly used as shown in

Table 1 [25]. Such fiber has excellent tensile and compressive strength which accompanied with a high char yield [26].

Chae et al [27] produced a high tensile strength and modulus CF using gel spinning of

PAN copolymer. The tensile strength was around 5.6 GPa and the modulus ranged

between 354 and 375 GPa. The structural analysis showed that the random nano defects around 2 nm size were the reason behind lowering the strength below 20 GPa.

Naito et al. [28] compared PAN and pitch-based carbon nanofibers tensile and compressive stresses. It was confirmed that the PAN-based single CF showed a higher compressive modulus than a pitch-based one. In general, PAN-based carbon fibers production involves three main steps which are: stabilization in air, carbonization, and graphitization [29] as shown in Figure 7.

Figure 7 Carbon fibers thermal treatment procedure (adapted from [29]).

As shown in Figure 7 many chemical and physical events involving chain scission, by- products evolution, crosslinking, and chain breakdown are connected to PAN crosslinking and pyrolysis [22].

2.1.1.1 Carbon Nanofibers

Carbon nanofibers (CNFs) are defined as fibers having more than 90% carbon element with noticeable strength (3-7 GPa) and modulus (200-500 GPa) with a very small diameter [30][31][32][33]. Carbon nanomaterials are developed via carbon diffusion which initiated and controlled using catalytic decomposition. Eventually, The resulted carbon will precipitate as a graphitic fiber [34] with various configurations such as stacked, herringbone or fishbone (cup-stacked) and nanotubular [35] (Figure 8).

Figure 8 CNF with three distinct aspect ratios a) platelet, b) tubular, c) fishbone (reprinted from permission [36]).

The platelet type CNF has a graphene layer perpendicular to the filament axes as shown in Figure 8 (a). In tubular type CNF, the graphene sheets align with the fiber direction forming many nanotubes. Finally, discrete graphitic cones stacked periodically with a growth axis is found in herringbone carbon fibers. Mostly, the fishbone CNF grew on nickel and nickel alloy with methane gas as a carbon source [37].

CNF was fabricated using catalytic thermal chemical vapour deposition (CVD) or electrospinning. In the first method, a metal such as cobalt, copper, nickel, iron is the main catalytic which conjucted with carbon source to fabricate carbon nanofibers using CVD

[38]. Although the resulted fibers yield very high modulus, the catalyst residue reduces the product mechanical properties. In addition, the CVD process relatively expensive process

[39].

Carbon nanofibers can be fabricated with electrospinning using a “top-down” approach. The main precursor for CNF fabrication should be a polymer or polymeric blends, followed by special heat treatment as mentioned earlier. The structure and properties of the carbon nanofibers governed effectively by the polymer type and heat treatment procedure [22].

Electrospinning has been extensively utilized in recent years, as a simple, inexpensive and versatile method, for the production of nanofibers from various materials

[40] [41][42][43]. The electrospinning technique is used to produce fibers with a diameter ranging from 1 to sub-micrometer [44]. In this method, a series of factors influence the morphology and the properties of electrospun fibers; for example, solution concentration, molecular weight, viscosity, surface tension, and electrical conductivity. Many researchers studied the effects of electrospinning parameters on the morphology and performance of fabricated fibers

The mechanical properties of the resulting electrospun nanofiber could be ultra-high modulus (>500 GPa), high modulus (>300 GPa), intermediate modulus (>200 GPa), low modulus (100 GPa) depending on the initial precursor [32]. For example, PAN-based carbon material has a lightweight and high mechanical properties (Figure 9).

In terms of the structure, the hexagonal planar ordering is the basic building unit in carbon fiber [45]. CF build up from the basic structural unit (Figure 10) with interatomic distance around 0.34 nm which is close to typical graphite planes (0.3345 nm) [32][46][29].

Figure 9 Tensile strength of different materials (reprinted from permission [47]).

The CF structure (Figure 10) can be considered as defected graphite with many vacancies, dislocation, bends, twin boundaries, and impurities [45]. However, the deviation from the ideal graphite structure depends on the precursors and fabrication process, so the structure may include some graphitic, turbostratic or hybrid structure [32].

Finally, the well-crystallized structure, a high degree of orientation with fiber axes, minimum defect density, small inter-atomic spacing are the main features of CF with high tensile strength, high modulus, and high electrical and thermal properties.

Figure 10 Graphite and turbostratic carbon structure.

Zhang et al [40] studied the structural, physical and mechanical changes during the heat treatment of various PAN precursors (from acrylonitrile / itaconic acid (AN/IA), acrylonitrile /acrylamide (AN/AM) copolymers and AM-containing commercial PAN fibers precursors). The fiber was pre-oxidized in the air between 125-283 oC for 50 min and then carbonized in an oxygen-free environment like nitrogen from 1003 to 1350 oC.

They concluded that the AM-containing commercial precursors have higher thermal stability and density than other PAN precursors.

Imaizumi et al [48] prepared carbon nanofiber (CNF) with a diameter of around

100 nm using phenolic resin. To control the electrospinning process poly(vinyl butyral) and electrolytes (pyridine and sodium carbonate) were added in order to control the conductivity, respectively. Such modification results in the fabrication of smaller and smother fiber which; eventually, after heat treatment, had produced, a conductive, and porous CF.

Wang et al [49] used PAN/Dimethylformamide (DMF) solution to fabricate ultrafine nanofiber. The heat treatment procedure included pyrolyzing in the vacuum around 600 oC, 800 oC, 1000 oC, and 1200 oC, respectively. By utilizing Raman 13

spectroscopy and ID/IG ratio, it was confirmed that the graphite quality and quantity were directly proportional to heat treatment temperature.

Prilutsky et al [50] fabricated hybrid nanofiber using PAN polymer and multi- walled carbon nanotubes (MWCNT) with the electrospinning method. It was shown that with increasing MWCNT, larger crystalline domains were nucleated as shown in

Figure 11.

(A) (B)

Figure 11 TEM images of a) PAN-5%MWCNT nano-fibers carbonized at 750 oC, in a region without carbon nanotubes (CNTs), b) PAN-25%MWCNT nanofiber carbonized at 1100 oC (reprinted from permission [50]). Ra et al [51] utilized PAN-based porous carbon nanofiber as a supercapacitor electrode. The fiber fabricated using a one-step carbonization/activation process. Such fiber was durable for high current loads with normalized capacitance around 67μF/mm2.

As compared to bulk activated carbon, the fiber showed 3- 4 times higher energy and power densities (Figure 12).

Figure 12 a) Specific capacitance vs discharge current density and b) Ragone plot in organic electrolyte (H2SO4 and KOH). The results obtained with activated carbon (AC) (Norit Super 50) [51]. Ji et al [52] utilized porous CF to fabricate anodes for rechargeable lithium-ion batteries. Such fiber was fabricated by the thermal treatment of electrospun PAN/ polylactic acid (PLLA) bi-component nanofiber. The special structure of such nanofibers exhibited excellent electrochemical properties (Figure 13).

Figure 13 Carbon nanofiber anode made from PAN/PLLA a) charge-discharge curve, b) cycling performance [52]. Chen et al [53] used the electrospinning technique to fabricate the nanofiber membrane for in vitro and in vivo use. The fabricated PLLA nanofiber was crafted with cationized gelatin to boost the fiber compatibility.

On the other hand, the low oxidation resistance of PAN-based CF was a focus of much research. Yuanjian et al.[30] investigated the effect of oxidative environment on mechanical properties of two types of PAN-based CF which were CF-A (carbonized at

1350 oC) and CF-B (carbonized at 1450 oC). The resulted fibers were exposed to temperature from 400 oC to 700 oC in air. The results showed that CF began to react with oxygen at a temperature of 400 oC. At 600 oC, a noticeable weight loss was found which ended with burning about 13% of total mass at 700 oC. As a result, the tensile strength of the fiber decreased significantly, but the modulus was unchanged.

The way to improve the CF is by incorporating silicon source which eventually results in nano dispersed silicon carbide and silicon oxy-carbides. The SiC and SiOC will improve the oxidation resistance of carbon fiber without affecting its excellent physical properties. The resulted hybrid fiber has the potential to work at higher temperatures and aerospace applications.

2.2 Hybrid Carbon-Preceramic Polymer Materials

Using polymer blends to form polymer composites involves using a mixture of two or more miscible or immiscible polymers [54]. Preceramic polymers can be mixed with organic polymers to form new hybrid carbon-ceramic material. Converting silicon-based polymers to ceramics is of great interest. It enables the manufacturing of materials with a wide variety of geometries at low temperatures [55][56]. SiC and Si3N4 based materials are very attractive because of their extraordinary environmental stability, high mechanical properties, extremely high resistance to oxidation (until 1400 oC) [57]. In addition, the

polymer derived ceramics can lead to a broad range of products such as fiber, 3-D printed products [58], bulk ceramics, porous ceramics, coating [59], thin-film and filters.

After doping of carbon materials with silicon-based material, many phases like free carbon (C), silicon (Si), silicon carbide (SiC), and silicon oxy carbide (SiOC) or combination of them could be created, depending on the chemical structure of precursors.

Several studies used both carbon and silicon precursors as a polymer material.

Ribeiro et al [60] fabricated highly oxidation resistant SiCN ceramics by mixing acrylonitrile (carbon-based polymer) and oligosilazane (silicon-based polymer). The resulted material showed high thermal stability until 800 oC which is about twice of carbon- based materials oxidation resistance. Kleebe and Blum [61] used organic polymer

(divinylbenzene) and inorganic polymer (polydimethylsiloxane) to fabricate carbon-rich

SiOC ceramics. The resulted hybrid material consists of carbon domain in form of turbostratic carbon, SiC nanocrystals, and SiO2 rich domain. Blum et al [62] successfully produced carbon and SiOC nano-composite using polydimethylsiloxane and divinylbenzene polymers. The resulted material showed excellent thermal durability and chemical stability at elevated temperatures. In addition, a semiconductor and high resistant hybrid C-SiOC composite made from carbon/silicon polymers studied by Lu et al [63]. The ceramics material showed high thermal stability up to 1000 oC and electrical conductivity around 4.64 S.cm-1. Saja et al. [64] used organic/inorganic polymer blend

(polydimethylsiloxane and polyacrylonitrile) to fabricate hybrid ceramic fibers. The resulted structure consists of SiC, carbon, and SiOC. The fibers showed improved mechanical properties as compared with carbon fibers. Niu et al [21] improved the

mechanical, electrical, and thermal properties of carbon materials by adding SiOC preceramic polymers.

2.3 Hybrid Carbon Materials with Solid Particles

Modifying carbon material properties is performed by doping the structure with nanoparticles or with silicon-based polymers [17]. The incorporation of inorganic particles into organic polymeric blends can lead to a wider range of hybrid materials. The key steps in the fabrication include good mixing with sonication to avoid solid particle aggregation and ensure homogeneity. Examples for metal oxides colloidal particles are Fe2O3 [65][66],

MnO2 [67], CuO [68], TiO2 [69]. Incorporating solid additives materials into organic precursors to produce a new composite leads to different material properties [70][71]. For example, PVA/Silica [72], PAN/Cu [73], MgO/Al2O3/PAN [74] systems were used to fabricate carbon materials composites. Silicon and silicon oxide are one of the important nanoparticles, which can be embedded in the polymer solution to achieve a new hybrid material [75]. For instance, C/Si composite nanofiber can be prepared by a judicious combination of colloids electrospinning and subsequent thermal treatments. Lee et al [76] added SiO2 nanoparticle in a PAN to fabricate SiC nanofibre using the electrospinning technique. It seems that a SiO2/C ratio played a key role in controlling the fibers' morphology, structure and properties. Thermogravimetric analysis (TGA) results showed an improved oxidation resistance for the hybrid fiber as compared with carbon fiber [76].

The incorporation of nanoparticles (Figure 14) within the organic hosts has no intercalation effect on the organic matrix. However, very small nanoparticles (1-100 nm) connect the small organic molecules [5].

Figure 14 Inorganic particles incorporation in polymers structure (adapted from [10]).

Streckova et al [77] utilized a needleless electrospinning process to fabricate hybrid carbon, carbon/nickel, carbon/nickel/ nickel phosphides, and carbon/nickel phosphides fibers with high aspect ratio. Such fibers could be functional as an electrode material for hydrogen evolution reaction.

Finally, in both cases, the resulted ceramic from either single-component preceramic polymers or from the whole conversion of the nanoparticles has limitations such as low toughness and low electrical properties. In addition, carbon material alone oxidizes in the air at high temperatures. Carbon materials are prone to oxidation starting at 500 oC which lower their thermo-mechanical properties [78][79][80]. As a result, the best way to overcome carbon oxidation is by doping, coating and using hybrid carbon-silicon carbide materials.

CHAPTER III

FABRICATION OF HYBRID NANOFIBER USING POLYDIMETHYLSILOXANE

AND POLYACRYLONITRILE POLYMER BLENDS

3.1 Introduction

Using polymers with Si backbone such as poly (silanes), (carbosilanes), (silazanes),

(borosilazanes) as precursors for the manufacturing of ceramic products offers many attractive advantages such as unique abilities of achieving a wide range of geometries at lower processing temperature, microstructures, and interesting properties such as high thermal stability, high mechanical properties and chemical resistance [81][82]. Recently, polysiloxane precursor has been used extensively in research in order to lower the cost

[83][84][85]. It exhibits particular characteristics, such as flexibility of the main chain, low glass transition temperature, hydrophobicity based on the alkyl side chain, thermo-stability, stability against atomic oxygen, and physiological inertness [86]. In addition, the polysiloxane is a good candidate to replace polycarbosilanes precursors to increase the efﬁciency, decrease the shrinkage and defects [87]. On the other hand, oxygen-containing

Si–O–C matrix materials derived from low-cost poly (siloxanes) are limited to relatively low-temperature applications below 1200 oC [84].

One of the most important ceramics manufactured from a polymer derived ceramics

(PDCs) is SiC which offers series of advantageous characteristics such as high melting temperature, low density, high mechanical strength, high elastic modulus, excellent thermal stability, chemical inertness, and low thermal conductivity. Due to its high performance, silicon carbide has been extensively used in high-temperature applications

(ceramic composites, membrane supports, and hot-gas filters) [82][88][27]. SiC was produced as thin films[89], nanowires [90][91], nano-powders[92], foams [93], rods

[94][95], and fibers [96][97].

Properties of the final ceramic product depend immensely on the polymer type, chemistry, curing temperature, pyrolysis conditions, and processing atmosphere [87].

Černy et al. [98] showed that precursors with O:Si =1.54 and C:Si =1.14 had the highest yield (70 wt%); however, only (27 wt%) yield exhibits for resins having O:Si =2 and C:Si

=4.87 ratios. In addition, nanocrystalline SiC was possible to achieve after heat treatment of the polymer with a C:Si ratio higher than 1. It was found that among eight polysiloxane resins from different producers polydimethylsiloxane resin was the most promising precursor for high-temperature applications, as it showed a maximum weight loss of 14%.

However, after exposing the material to elevated temperature the solid residue of polydimethylsiloxane resin was very small, as it showed a high weight loss about 20%, even 30%. Thai et al. [90] studied the thermal stability of four organofunctional polysiloxanes. In addition to the model fitting method, they used several experimental methods to find the material thermal behavior like TGA and Fourier transform infrared spectroscopy (FTIR). They found that by changing the organofunctional group in a polysiloxane, different thermal degradation mechanism was achieved. Consequently, the poly (aminopropyl) siloxane showed the lowest thermal resistance.

All the mentioned research utilizes silicon polymers to manufacture ceramics, some of them use polymer combinations (copolymers) to fabricate ceramics. For instance,

Monnier et al. [99] used graft copolymers of PAN and PDMS in their study to investigate

the relaxation process due to thermal molecules' rearrangement. They used temperature modulated differential scanning calorimetry (TMDSC) to track such changes.

PAN polymer is an excellent choice for manufacturing carbon fiber. Zhang et al.

[40] studied the structural, physical and mechanical changes during the heat treatment of various PAN precursors from acrylonitrile / itaconic acid (AN/IA), acrylonitrile

/acrylamide (AN/AM) copolymers and AM-containing commercial PAN fiber precursors).

The fibers were preoxidized in the air between 125-283 oC for 50 min, then carbonized in an oxygen-free environment like nitrogen from 1003 to 1350 oC. They concluded that the

AM-containing commercial precursors have higher thermal stability and density than other

PAN precursors. In addition, they found that the high density, sufficient strength, good crystallinity, and low defect concentration are the typical properties for PAN precursor

To fabricate nanofiber, electrospinning has been extensively utilized in recent years, as a simple, inexpensive and versatile method, for the production of nanofiber from various materials [83][100][101][41][42][43]. The electrospinning technique is used to produce fibers with a diameter ranging from 1 to 3 μm [102][103][104][105]. In this method, a series of factors influence the morphology and the properties of electrospun fibers; for example, solution concentration, molecular weight, viscosity, surface tension, and electrical conductivity.

Many researchers studied the effects of electrospinning parameters on the morphology and performance of fabricated fibers. For instance, Eick and Youngblood

[106] prepared electrospun nanofiber of nanocrystalline α-SiC with different fibers diameters and mat morphologies by changing a series of parameters such as precursors

(poly(carbomethylsilane) (PCmS) and polystyrene (PS)) ratios, the molecular weight of

PS, and concentration of the solution. Dong et al.[107] explored the effects of the concentration on the polymer solution, ratios of mixing solvent, and electric field on the electrospinning of polycarbosilane (PCS). They found that a higher DMF/toluene ratio will decrease the fiber diameter while it can keep a uniform distribution of fiber-diameter and

30% DMF/toluene with a concentration of 1.3 g ml-1 is responsible for establishing a uniform SiC web. Meanwhile, an electrical voltage higher than 20 kV is appropriate to make a uniform SiC web. Yördem [108] successfully manufactured PAN fibers; In addition, the effect of electrospinning parameters like solution concentration, voltage, and the collector distance was investigated. The results showed that all the mentioned factors are important in the fabrication of nanofibers. On the other hand, the contribution of the applied voltage to fibers size was very low when increasing the solution concentration and collector distance. In order to achieve good ceramic nanofiber by using polymer with silicon backbone like polydimethylsiloxane (PDMS) some challenges arise related to its rheological properties and molecular weight. To overcome such an issue, a carrier polymer was used to carry out the Si-based polymer. For example, Yang et al. [109] utilized

Polymethyl methacrylate polymer as a carrier for PDMS silicon polymer to overcome

PDMS low molecular weight problems. Uyar et al. used the poly (ethylene oxide) polymer as a vehicle to cyclodextrin resin[110].

In this chapter, a hybrid SiC, SiOC and carbon nanofibers were fabricated by optimizing PDMS and PAN weight ratios and using the electrospinning method. In such mixtures, the PAN is considered as a carrier for PDMS and a carbon-based precursor for

SiC nanofiber. The effect of weight ratios and heat treatment of PDMS and PAN on hybrid fiber mechanical properties of electrospun nanofibers were explored. Several instruments

were used to characterize hybrid fiber properties such as microstructure, mechanical properties, and phases formation.

3.2 Materials and Method

PAN copolymer (Scientific Polymer Products Inc.) with a molecular weight about

100,000 g/mol was used as a carbon source. Then, it was mixed with different weight ratios of PDMS. Both polymers were dissolved in a solvent mixture (DMF and acetone). The weight contents of PDMS and PAN in each sample are listed in Table 2. In all samples,

5ml DMF and 5ml acetone were used to dissolve PAN and PDMS, respectively. However,

100 % PAN was dissolved in just 10 ml DMF, because it has no PDMS which required to dissolve in acetone. The total weight of the polymer is 10 wt% and solvent 90 wt%.

Table 2 Weight contents of PDMS and PAN.

Material PDMS (wt%) PAN (wt%) PAN10 0 100 PAN8:PDMS2 20 80 PAN7:PDMS3 30 70 PAN5:PDMS5 50 50

3.2.1 Electrospinning of PDMS/PAN Nanofiber

Prior to electrospinning, the sample solution was stirred using a magnetic agitator at 55 oC for twenty hours until PDMS and PAN were fully dissolved. Then, the nanofiber was produced through a mixed solution by an electrospinning technique. The electrospinning setup apparatus is composed of a voltage controller (Stanford research systems, Inc. Model PS375), a syringe pump (new era pump systems, Inc. NE-300),

rotating collector (Dayton® DC Motor, 4Z145), collector controller, syringe, and needle.

The distance between the tip and the collector was 25 cm. A voltage of 15 kV was applied to the solution to start the spinning process. The electrospun fibers were collected on a rotating drum in order to orient the fibers. In addition, due to different rheological properties of the different solutions as shown in Figure 15, a range of the feeding rates around 0.001ml min-1 to 0.005ml min-1 was used to produce good fibers with stable Taylor cone and without dripping. Thus, higher feeding rates belonged to the fluids with higher viscosity and they are decreased toward fluids with lower viscosity.

Figure 15 Viscosity- shear rate for different polymers solutions.

3.2.2 Stabilization and Pyrolysis of PDMS/PAN Nanofiber

After electrospinning, stabilization and heat treatment of the samples were performed in a tube furnace. First, the samples were stabilized in the air from 25 oC to 225

oC using heating rate 1 oC min-1 and holding at the final temperature for 15 min. Then, the samples were carbonized from 225 oC to 600 oC with a slow heating rate (0.5 oC min-1) in argon for 15 min. Finally, from 600 oC to 1000 oC the heating rate was very slow (to avoid the thermal stresses) with the heating rate (0.25 oC min-1) in an argon atmosphere for 15 min at 1000 oC (all temperatures were optimized experimentally).

3.2.3 Characterizations

The as-spun, stabilized, and pyrolyzed fibers were characterized using microscopic and spectroscopic tools. Fiber morphology was examined using a scanning electron microscope (SEM, ProX, Phenom, Eindhoen, Netherlands) and transmission electron microscope (TEM). The average fiber diameter was estimated by measuring 50 fibers randomly using Image J software [111][112]. The chemical composition of the fibers after heat treatment was analyzed by an energy dispersive spectrometer (EDS). Fourier transform infrared spectroscopy (NICOLET iS50 FTIR,

Thermo Scientific) was used to collect FTIR spectra. The tensile test was used to investigate the effect of heat treatment on tensile strength and Young’s modulus.

Samples were cut into strips parallel to fiber longitudinal directions. They were in the form of fibers mat with dimension (6000× 700 × 5 휇푚3) and gage length about 1700

휇푚 which measured by connecting the tension stage with optan ical microscope. The thickness of fiber mats was measured using thermal mechanical analysis (TMA, TA-

Q400) in expansion mode. X-ray diffraction (XRD) (RIGAKU, JAPAN) with a voltage of 40 kV and a current of 44 mA was used to track phases formation.

Rheological properties were measured using Anton Paar MCR-302 Modular Compact rheometer at room temperature with cone fixture.

3.3 Results and Discussion

3.3.1 Nanofiber Morphologies

After spinning and regardless of the mixture ratios, all as spun nanofibers were white and ductile with sub-micron diameter. The morphology images of as-spun fibers with different PDMS concentrations are shown in Figure 16.

a b

c d

Figure 16 SEM micrographs of as-spun nanofibers with different PAN:PDMS ratios, a) PAN10, b) PAN8:PDMS2, c) PAN7:PDMS3, d) PAN5:PDMS5.

It seems that the nanofiber size was affected by two controlling factors which are: viscosity and solvent type. In this study, the influence of solvent is very obvious. From

Figure 16, PAN10 nanofiber which dissolved in DMF exhibit much smaller diameter than

PAN8:PDMS2 and PAN7:PDMS3 samples (dissolved in DMF and acetone) although it spun from a solution with much higher viscosity. This means that for PAN8:PDMS2 and

PAN7:PDMS3, the solvent effect is more dominant in defining nanofibers' diameter than the viscosity of the material. The acetone has a higher evaporation rate than DMF which gives the solution with acetone less time to elongate before deposition on the collector.

However, with increasing PDMS, the overall solution viscosity decreased. From Figure 16, the as-spun nanofibers were cylindrical, continuous, and with a smooth surface. On the other hand, some nanofibers (PAN5:PDMS5) exhibit two or more fibers sticking together and forming beads. It might be caused by low viscosity and high surface tension of the solution [113][114][85] [115][116].

However, after stabilization and carbonization, the fibers became black and brittle.

The morphology images of as-spun, cured, carbonized, and paralyzed fibers were characterized by scanning electron microscopy. The morphologies of nanofibers with a composition of PAN8:PDMS2 at each heat treatment are shown in Figure 17.

a b c

Figure 17 SEM micrographs of PAN8:PDMS2 nanofibers at different treatments temperatures a) 225 oC, b) 600 oC, c) 1000 oC. From Figure 17, the fibers became wavy after the curing heat treatment. In addition, a noticeable shrinkage was observed after carbonization at 600 oC as a result of weight loss,

structural and chemical changes which combined with various gas releases such as CO2,

CO, NH3, and H2O [113]. Further gas emissions, which combined with a reduction in diameter was achieved after subjecting the fibers to an elevated temperature (1000 oC) as shown in Figure 18.

Figure 18 The average diameter of electrospun nanofibers for different heat treatment.

3.3.2 Energy Dispersive Spectroscopy

To analyze the element distribution across the nanofiber diameter, it decided to use

SEM instrument coupled to EDS detector. Only carbonized samples at 1000 oC were analyzed. Elements like Si (from PDMS precursor), C, O, and N (from PAN precursor) were looked for and detected. After multiple scans and element mapping, we have found that the silicon element is concentrated at the center of nanofibers and decreased gradually as we become close to the skin as shown in Figure 19. In contrast to Nayak et.al. [114] who found uniform Si distribution across fiber diameter after adding PDMS to polypropylene.

Figure 19 Si distribution across fiber diameter for PAN8:PDMS2 sample.

3.3.3 Fourier Transform Infrared Spectroscopy

Figure 20 shows the FTIR spectrum of as-spun nanofibers consisting of PAN to

PDMS weight ratios.

Figure 20 FTIR spectrum of the as-spun nanofibers.

For the as-spun nanofiber with the composition of PAN7:PDMS3, the characteristic absorption bands at 1019 cm-1 and 1081cm-1 were attributed to the Si-O-Si stretching vibration at PDMS [115][85]. The absorption bands at 799cm-1, 1260cm-1 were related to

-1 deformation in Si-CH3 bond at PDMS. The peak around 2962cm corresponded to C-H

-1 stretching in Si-CH3 [102][85][116]. In addition, the peak around 1732cm was related to

C=O or C-O in co-monomers such as methacrylate in PAN. Furthermore, in the sample of

PAN7:PDMS3, the intensity of the peaks about silicon was much higher than that in the sample of PAN8PDMS2 and PAN10 due to the higher composition of silicon. In addition, peaks 2241 – 2243 cm-1 are due to the CN nitrile group [117].

The FTIR spectrum of nanofiber with the composition of; for example,

PAN7:PDMS3 is shown in Figure 21 which explains the results for as-spun, cured, and carbonized samples, respectively.

Figure 21 FTIR spectrum of nanofibers with composition of PAN7:PDMS3.

From Figure 21, the intensity of most peaks decreased after curing at 225 oC due to cross-linking. After curing, the intensity of many peaks decreased, especially for the peaks

-1 -1 -1 -1 at 799 cm , 1019 cm , 1081cm , and 1260 cm which represent deformation in Si-CH3 bond and Si-O-Si stretching vibration. This indicated that a large amount of oxygen reacted with the sample and replaced the weak bonds such as methyl bonds or hydrogen during oxidation curing

After pyrolyzing at 1000 oC a peak caused by Si-O bond vibrations appeared at

1030 cm-1[118]. This proved the formation of Si-O-C ceramic nanofibers from the conversion of preceramic polymers. Another phase that evaluated at high temperature is indicated by the peak around 800 cm-1 which is related to amorphous SiC phase vibration[91][119]. In general, the FTIR vibrations between 700 to 1000 cm-1 can be related to SiC phase stretch [120][121][122]. The location of band vibration is associated with many structural parameters like phase size, crystallinity, chemical composition, preparation method, and the morphology of SiC.

3.3.4 Raman Spectroscopy

The reason behind performing Raman is that it is an excellent tool to characterize carbon-based materials. Furthermore, some peaks are not IR active, so Raman is a good tool to find such peaks. The Raman spectroscopy for three distinct concentrations after heating to 1000 oC is shown in Figure 22. The results showed many sharp peaks, the first two (around 1370 cm-1 and 1596 cm-1) are called a D band and G band, respectively. The appearance of D band is related to lack of bond ordering in sp2 graphitic domain or slight coherence between neighbors graphitic layers. In another word, in carbon-based materials, the D band is one characteristic of sp3 bonding; however, the G band around 1596 cm-1 is one feature of sp2 hybridization in graphene nano-domain [123][124]. In addition, the

intensity of G band is higher than D band which indicates an increase in sp2 domain [125]

[126] [127][128] [129].

Figure 22 Raman spectroscopy for pyrolyzed samples at 1000 oC.

3.3.5 X-Ray Diffraction

The crystallization behavior of PDMS and PAN after heat treatment to 1000 oC was investigated by XRD. All peaks were broad and revealed that the material was still mostly amorphous at 1000 oC. From Figure 23, a broad hump around 20-30 (2θ) was observed for all samples. In addition, the wide peak, which centered at 2θ = 24.5o is related to (002) graphitic layer. Another wide peak at 2θ = 44.0o was very obvious which corresponds (100) signal of turbostratic carbon plane [130][131].

Figure 23 XRD for PAN8:PDMS2 at 1000 oC.

3.3.6 Transmission Electron Microscopy

TEM was used to confirm or deny the previous analytical techniques with additional details and more precision. To do TEM, only carbonized samples were considered. The sample preparation was performed using the ultra-microtomy technique with a very thin slice of less than 70 nm. From the previous results, the sample showed a skin-core effect similar to that observed in Nicalon fibers. In addition, TEM images shown in Figure 24 a-c confirmed the formation of nano-SiC crystals. However, when selected diffraction was performed, both carbon and SiC phases were present.

Figure 24 TEM images a) Bright field image, b) Dark field image, c) Selected Area Diffraction (SAD) of the PAN8:PDMS2. On the other hand, depending on the location, the skin seems to be is carbon-rich and the core is silicon-rich. Furthermore, the core is mainly dominated by silicon-based phases in the form of SiC or Si-O-C phases. Using lattice fringes imaging technique of the surface skin of the nanofiber sample (Figure 25) where L1 represents the length of basic structural units (BSUs) and L2 the length of distorted and aggregated BSUs and N of carbon stacking layers, it seems that sample surface is rich of turbostratic carbons with the 2D order.

Figure 25 Lattice fringes imaging of skin surface.

In addition, regardless of PAN:PDMS ratios, once these two phases are mixed, after stabilization and carbonization at 1000 oC, three phases are present. These phases are silicon carbide, carbon, and oxy silicon carbide. Spectroscopic and microscopic techniques had confirmed the presence of nanocrystalline SiC and turbostratic carbon. These phases formed an intertwined network at the nanometric scale

Figure 26 Lattice fringes of nanofiber core a) the small arrows indicate the carbon basic units, b) large arrows indicate the existence of 111 SiC. 3.3.7 Tensile Test

The effect of adding PDMS on tensile strength and Young’s modulus was calculated by taking the average of five samples for each weight ratio. The results in Figure

27-a and b showed that adding of PDMS in the percentage of 20% and 30% can improve tensile strength and Young’s modulus of uncured fibers. However, increasing PDMS to

50% showed a negative effect on fiber mechanical properties. For instance, the

PAN5:PDMS5 sample showed about 80% reduction in tensile strength as compared with

PAN8:PDMS2 samples.

Figure 27 The effect of PDMS percentage on fiber a) Tensile strength, b) Young’s modulus.

The effect of heat treatments on fibers Young modulus is shown in the stress-strain curve for as-spun, cured at 225 oC, carbonized, and pyrolyzed samples (Figure 28). The tensile test for the carbonized samples at 600 oC was very challenging as the material was very fragile and hard to handle with low strength. On the other hand, after pyrolyzing at

1000 oC with holding 15min, tensile strength improved slightly, but the elastic modulus showed a very noticeable increase due to the disappearance of the organic character of

material and formation strong ceramics like SiC which detected in FTIR, Raman and TEM tests.

Figure 28 Stress-strain curve at different heat treatments for PAN8:PDMS2.

3.4 Conclusions

Nanofibers with a diameter of several hundred nanometers were successfully fabricated using the electrospinning process and a mixture of two types of polymers

(PMDS and PAN) as a precursor. After stabilization and carbonization at 1000 oC, three phases such as silicon carbide, carbon, and oxy-silicon carbide were presented.

Spectroscopic and microscopic techniques had confirmed the presence of nano-crystalline

SiC and turbostratic carbons. These phases formed an intertwined network at the nanometric scale. In addition, the resulted fibers showed a core skin structure with the skin richer in carbon and a core mainly dominated by silicon-based phases in the form SiC or

Si-O-C phases. A significant improvement was observed in both tensile strength and elastic modulus in these hybrid fibers as compared to SiC free carbon fiber. In terms of

crystallography, such nanofiber seems to exhibit similar microstructure that was observed in Nicalon fiber. However, it was difficult to determine the ratio of these phases and their influence on the physical properties of these hybrid fibers.

CHAPTER IV

EXPERIMENTAL AND NUMERICAL INVESTIGATION OF THE SILICON

PARTICLE DISTRIBUTION IN ELECTROSPUN NANOFIBER

4.1 Introduction

One of the vital research areas in fiber fabrication is the spinning of a colloidal system which consists of polymer and nanoparticles to fabricate a multi-component nanofiber. The incorporation of inorganic materials into organic precursors to produce hybrid composites lead to new material properties

[132]. Silicon and silicon oxide are considered to be a significant source of nanoparticles, which can be embedded in the polymer solution to achieve new hybrid material [75]. For instance, C/Si composite nanofiber can be prepared by a judicious combination of colloidal electrospinning and subsequent thermal treatments. As a result, it is possible to tailor properties with various particles that ultimately will be transformed into a gradient of crystal size. For example, the oxidation resistance of SiC is dependent on its particle size [133][134].

Quanli et al [135] confirmed that silicon carbides particles have a massive effect on oxidation kinetics, as the weight gain and rate constant were inversely related to SiC particle size. In addition, they observed that the oxidation temperature had increased by changing the particle size from a nanometric to a micrometric scale.

The dispersion, miscibility, and particle location inside the fibers are limiting factors in determining the final fiber properties [136]. The process

needs to be very well understood and controlled. The general protocol to control particle distribution falls into two main categories: active and passive particle distribution [137][138]. The first group is based on directing external sources of energy to regulate the particle trajectories. These sources involve a dielectrophoresis [139], electrophoresis [140], acoustic, optical fractionation

[141] [142] magnetic [143][144], electric, and thermal ones [145]. For example, in electrospinning, the electric forces can effect on charged elements distribution. Furthermore, the surface and core properties can be tailored to fit different applications by monitoring the species' charge. Sun et al [146] investigated two polymer components distribution within the electrospun fibers.

One of the components was neutral (polyethylene oxide (PEO)) while the other one was charged (peptide-polymer conjugate). The results showed that the outer surface collected the charged peptide, and the PEO was localized in the fiber center. In addition, Biono [147] used alginate and PEO blends to investigate components segregation during the electrospinning process. The observations showed that negatively charged alginate clustered on the wall while the PEO was concentrated at the core.

On the other hand, a passive separation, which based on morphological variations between particle groups like (shape, size, deformability, and mass), was used to isolate the particles by utilizing inertial forces or hydrodynamic effect [148][137][149]. It is important to study and control this separation category as it is responsible for clustering the particles according to their size, which is crucial in controlling the final fiber properties. Szczech and Jin [150]

manipulated reversible capacity values in lithium-ion batteries by changing the silicon particle size. As a result, the accumulation of large or small particles at the wall seems to lead to different electrochemical properties. To perform the separation, Park et al [151] utilized inertial forces to isolate multi-sized spherical particles that flowing in a series of microchannels. As a result of initial forces, the particles bent from the main fluid streamlines and accumulated on the channel walls. Furthermore, Bec [152] found a critical value for Stokes number in which the particles began to cluster. The threshold for this number was around 10-4 and below this number, no particle separation occurred.

Controlling the double-phase dispersion in the nano-scale is very challenging. One method is using the electrospinning technique, which is an easy and reliable method for the fabrication of nanofiber [110]. In this process, the electrical field has the functionality of inducing electrical charge along the liquid surface. Increasing the intensity of the electrical field, induce a Taylor cone as a result of elongation of the semicircular surface of the solution [109].

The electrospinning apparatus involves high voltage power supply (1-30 kV), a syringe, needle, connected to a terminal, and collector, connected to the ground

[153]. To fabricate ceramics or carbon nanofiber, electrospinning has been extensively utilized in recent years [83] [100] [101] [41] [42] [43]. In this method, fiber with sizes from sub-microns to few nanometers can be fabricated

[154] [155] [71] [156]. To fabricate successful fibers, several factors can be related to properties and morphology of the spun fibers; for instance, solution concentration, molecular mass, rheology, molecular alignment, and electrical

conductivity play a vital role in this process [64] [23]. Kim and Um [157]

investigated the influence of the electrospinning parameters on silk fibroin

nanofiber properties. The effect of the concentration and viscosity showed a

dominant role in defining the fiber morphology and defect formation. They

found that the viscosity threshold to fabricate a bead free fibers was around 0.13

and 0.2 Pa.s. In addition, the authors showed that increasing the silk

concentration resulted in increasing the diameter of the fiber [144]. Mit-

uppatham et al [158] studied the effect of polymer molecular weight,

concentration, and viscosity on ultra-fine polyamide-6 (PA-6) fibers. Their

results showed a monotonic increase in fibers' diameter that occurred with

increasing the polymer viscosity and concentration.

In this chapter, the electrospinning technique was utilized to fabricate

hybrid nanofibers. The main polymer was polyacrylonitrile which mixed with

silicon nanoparticle with very low percentages to form the colloidal suspension.

The effect of silicon percentage on fiber morphologies was examined using a

scanning electron microscope (SEM). As the silicon particles are neutral, the

active separation is not possible and the passive one is dominant. The particle

size distribution inside the fiber was characterized using Transmission Electron

Microscope and numerically using software package COMSOL Multiphysics

Version 5.

4.2 Experimental Procedure

The electrospinning solution was prepared by dissolving PAN ( Mn =100,000

g/mol) which purchased from Scientific Polymer Inc. Ontario, NY and Si nanoparticles

with an average size of 200 nm (Sky Spring Inc. Houston, TX) in 10 ml DMF. The weight percentages of the solution components were 10 wt% of PAN and Si, and 90 wt% DMF.

The weight contents of Si and PAN in each sample are listed in Table 3.

Table 3 PAN to Si ratios.

PAN wt% Si wt% 100 0.00 99.5 0.50 99.0 1.00 95.0 5.00 90.0 10.0

The mixture was vigorously stirred using a magnetic agitator at 60 oC for 8 hours until Si and PAN were fully mixed. Then, the mixture was loaded to electrospinning apparatus to produce nanofibers. The electrospinning setup is composed of a voltage controller (Stanford research systems, Inc. Model P.S375, Sunnyvale, CA), a syringe pump

(New Era Pump Systems, Inc., NE-300, Farmingdale, NY), rotating collector (Dayton®

DC Motor, 4Z145). A voltage of 15 kV was applied to the solution to start the spinning process. The electrospun fibers were collected on a rotating drum. Finally, the feeding rate for all samples was around 0.003 ml min-1. The distance between the syringe tip and the collector was 25 cm.

SEM was utilized to characterize the morphology of the fibers. To estimate the average fiber diameter, about 50 fibers were measured randomly using Image J software

[111]. TEM was utilized to track particle distribution within the fiber. Finally, to find fluid rheological properties, Anton Paar MCR-302 Modular Compact Rheometer was used to find the fluid viscosity.

4.3 Numerical Modeling

COMSOL Multiphysics Version 5 was selected to simulate silicon particle size distributions within electro-spun fibers. Due to the difficulty of including the whole electrospinning machine, a small part of the fiber was included by representing it in 2-D symmetric geometry (to reduce the computational cost). The modeled part was taken from the stable straight jet to avoid electrohydrodynamic instabilities in whipping and to ensure that most of the solvent still in place which means that the solution is almost fluid. One of the challenges in modeling such a process is that everything in electrospinning is in macro- scale while the fibers and particles in micro to nano-scale. In this model, the main fluid

(PAN/DMF solution) flow was modeled using a laminar flow module and the particle position with time was tracked using the particle tracing module. The electric field effect on particle position was neglected, due to the neutral nature of silicon nanoparticles.

Finally, solvent evaporation was considered negligible in this model.

4.3.1 Meshing

Triangular and quadrilateral mesh types (Figure 29) with a total of 90460 elements were used to divide the computational domain. The element size and type were selected after performing the grid refinement study.

Figure 29 Mesh generation.

4.3.2 Governing Equations

The fluid flow inside the computational domain is governed by the Navier-Stokes equation which solves the conservation of momentum and the continuity equation for conservation of mass. This equation for incompressible and stationary flow [159] can be simplified as:

휌(푢. ∇)푢 = ∇.[-PI+휇(∇푢 + (∇푢)푇)]+F ….…(1)

휌∇. (푢) =0 ….…(2)

Where 휌 is the fluid density (978 kg.m-3) which calculated using mixture law and

휇 is the dynamic viscosity (0.175 Pa.s). In addition, to find the fluid velocity (u), the volumetric flow rate (Q) which is known from the pump that connected to the syringe is used as the following [160] [161]:

푄 = 휋푟2푢 ….…(3)

Where r is the fiber radius, u is the jet velocity, and after the calculations, the fluid velocity was found around 15 ms-1. For the boundary conditions, the velocity is considered as a normal inlet velocity. Finally, due to the absence of the solid-liquid boundary and all the computational domain was fluid, the wall boundary condition was chosen as a slip wall.

To find the particle position with time, the particle tracing for the fluid flow module was used. The governing equation [162] for such modules is:

푑 ….…(4) (푚 휐) = 퐹 푑푡 푝 푇

-1 Where푚푝: particle mass (kg), 휐: particle velocity (ms ), and 퐹푇: total force (N).

The total force in our modeling is a combination of drag and inertial forces. The drag force (equation 5) is built-in force in the module [162]:

1 ….…(5) 퐹퐷= 푚푝(푢 − 푣) 휏푝

휏푝: Particle velocity response time for spherical particles in a laminar flow

2 푑푝휌푝 ….…(6) 휏 = 푝 18휇

3 푑푝 : is the particle diameter (m), and 휌푝 is the particle density (2328 kg/m ), and u is fluid velocity

The inertial force [151] [163] that controls the lateral-particles migration is defined using user-defined force in COMSOL.

2 3 휋푈 휌푝푑푝 ….…(7) 퐹𝑖= 6퐷ℎ

Where 퐹𝑖 : Inertial force, U: average fluid velocity, and 퐷ℎ is the characteristic channel dimension. Although the behavior of heavy particles in micro fluids is not totally understood [163], fluid and particle actions inside the microfibers can be analyzed using dimensionless numbers. For instance, the dimensionless number that controls the inertial

and viscous forces [151] is the Reynolds number (Re). Another critical number is Stokes

number (St) which defined as the ratio of particle relaxation time 휏푝 to flow characteristic

time 휏푓 as follow:

2 휏푝 휌푝 푑 /18 휇 ….…(8) 푆푡= = 휏푓 퐷ℎ/푈푚

Were 푈푚 is the maximum velocity.

4.4 Results and Discussions

4.4.1 SEM Characterization

The SEM images of the as-spun nanofiber with various Si weight percentages are

shown in Figure 30.

Figure 30 SEM micrograph with diferente Si weight percent a) 0 wt%, b) 0.5 wt %Si, c) 1 wt%Si, d) 5 wt%Si, e) 10 wt%Si.

For all concentrations, the resulted fibers were very uniform with limited defects like beads or surface irregularities. However, for fibers made with higher silicon contents

(Figure 2 (d) and (e)), the resulted fibers show a noticeable amount of silicon particles on fiber surface which led to a rough fiber surface. Another effect of silicon content on fibers morphology was the fibers' dimensional change (Figure 31).

0% 0.5% 1% 5% 10% Si c

Figure 31 Silicon concentration effect on fiber average diameter.

Although the total solid content was kept constant (10 wt%), the PAN/Si ratio showed a vital role in controlling the fiber diameter. The diameter for the fiber with 0.5 wt% Si showed around 14 % increase as compared with the fiber without Si nanoparticles.

In addition, bigger fiber was fabricated by adding 1 wt% Si. Further increase in Si content showed an adverse effect on fiber diameter [164] [165] as illustrated in Figure 31. These morphological changes were controlled by the polymer concentration which regulates the solution viscosity.

4.4.2 TEM Characterization

The ultra-microtomy technique was performed for TEM sample preparation. TEM was used to characterize the particle size distribution in the fiber. The analysis was carried out for the fibers with 1 wt% Si and 99 wt% PAN as shown in Figure 32.

Figure 32 TEM image of a single fiber.

From the above Figure, silicon large particles were located near the wall and small particles were evenly distributed. This result was analyzed and confirmed using COMSOL multi-physics modeling by tracking the particle position with time.

4.4.3 Modeling Results

4.4.3.1 Visualization of Particle Distribution

Particles with a varying size between 2.5E-7 m to 1.0E-8 m were used to understand the particle lateral migration under the influence of drag and inertial forces. A fluid with

9.9 wt% PAN and 90 wt% DMF was the carrier fluid. The velocity profile of this fluid was

a uniform with 15 m.s-1 which results from setting the slip wall boundary condition. After solving the governing equations that mentioned earlier, the visualization of two particles set (2.5E-7 m and 1.0E-8 m) with different times is shown in Figure 33 (a) to (e). The big and small particles at the beginning of the release (Figure 33 (a)) have almost the same position without lateral movement. After spending the particles short time in vertical movement, the particle size effect began to be obvious. At 6.6 E-7s the big particles start to hit the wall (Figure 33 (c)) while the small ones kept flowing with the fluid main streamlines. After bumping the wall, the big particles freeze and solidify on it (Figure 33

(d)) which solves the clue of populating the big particles on the wall. Finally, most of the big particles concentrated on the wall after spending 9.6E-6s (Figure 33(f)).

Figure 33 Particle position from the inlet at different times a) 6.0E-8s, b) 3.0E-7s, c) 6.6E- 7s, d) 7.2E-7sec, e) TEM image for the right side of the fiber f) 9.6E-6s.

To sum up, the particle clustering on the wall was possible only for the big particle as the small particle confirmed a negligible transverse movement. Such behavior can be related to a distinctive particle inertial force for different sized particles. For big particles, the inertial effect was sufficient to enable the large particles to cross the fluid main streamlines.

Figure 34 illustrates the effect of particle size on the transverse shift for five particle sizes. From this Figure, at the beginning of the electrospinning (time = 0), different particles with different dimensions were distributed randomly throughout the fiber. At that time, the smallest particle 1.0E-8 has an initial position around 0.44 휇푚 from the fiber centerline, while the largest particle has an initial position around (0.38 휇푚). After a short time, particles with diameters ranging between 2.5E-7 m to 1.0E-8 m had different migration rates due to changes in inertial forces. For example, the largest particles in this model (2.5E-7 m) took about 0.66 휇sec to reach the wall and settle down on it, while the smallest particles (1.0E-8 m) followed the fluid inward velocity with negligible transverse movement. Numerical results were in agreement with various TEM characterizations.

Figure 34 Particle position with time for a range of particles.

From Figure 34, the particle lateral velocity can be predicted by calculating the slope of the straight line for each particle size. The resulted points were fitted (Figure 35) using Origin Lab 8.5 software. The high transverse velocities of large particles explain the short time they took to settle on the wall.

Figure 35 Particle size effect on particle lateral velocity.

From the mentioned Figures, the threshold of particle diameters that affected by inertial forces was defined by Stokes number (equation 8). From this equation, the Stokes number is directly proportional to the square of particle diameter and decreasing the particle diameter (other parameters constant) will lower Stokes number.

Figure 36 confirmed that the highest Stokes number around 1.4E-3 observed from particles with 2.5E-7 m diameter, caused the most significant particle-fluid streamlines mismatch which; eventually, induced large particles to concentrate at the fiber wall.

Figure 36 Particle size effect on Stokes number.

Finally, the relationship between the Stokes number and the lateral shift is shown in Figure 37. It is very obvious that the particle lateral velocity increases linearly with increasing Stokes number. In addition, the threshold for Stokes number after which a negligible lateral particle velocity occurred, was around 2.2E-4. This means that particles having a diameter of less than 1.0E-7 m were unable to cross the streamlines. Bec [152] simulation results showed that there was a critical value of Stokes number around 10-4 in which particle clustering happened.

Figure 37 Stokes number influence on particles lateral velocity.

4.5 Conclusions

A hybrid C-SiC nanofiber with controlled phases dispersion was performed using electrospinning of a mixed binary system. The system consists of PAN polymer and silicon nanoparticles. To track the particle dispersion process, TEM characterization was carried out to visualize particle size and their distribution which followed by computational study using COMSOL Multiphysics. Both experimental and numerical results showed that the larger silicon particles were located on the fiber surface and; eventually, the smaller particles concentrated at the core. The larger particles have higher inertial forces as compared with the smaller ones which enable them to migrate to a different location within the fiber. In addition, it found the threshold for particle size to make the inertial effect effective was around 1.0E-7 m. The threshold for Stokes number was found to be around

2.2E-4 with a critical particle size of 1.0E-7 m in diameter. The results demonstrated a novel way for the fabrication of PAN/Si ceramic nanofiber with a gradient of particle size and properties from the skin to the core.

CHAPTER V

THE ROLE OF CARBON AND SIC CRYSTALLINITIES IN THE OXIDATION AND

MECHANICAL PROPERTY IMPROVEMENT OF HYBRID NANOFIBER

5.1 Introduction

The high mechanical, electrical, and physical properties of carbon fiber, introduce

it as a candidate for many high-tech applications. However, CFs are prone to oxidation

starting at 400 oC which lower significantly their thermo-mechanical properties

[78][166][80]. The interaction between carbon materials and oxygen is associated with the

features of carbon-oxygen reactions and adsorption/ desorption of oxygen on the surface

[167]. CF can be fabricated using polymeric precursors like acrylic, cellulosic, pitch-based,

and PAN precursors [31][23]. PAN-based carbon fibers offers numerous advantages, such

as lightweight and high tensile strength properties. Thus, it is worth finding methods to

enhance the oxidation resistance of such important material. Many approaches were carried

out in the past 50 years to find an effective way to settle down this issue [168]. The efforts

fall into three categories which are: using inhibitors [169] coating [170][171][170][171],

and replacing carbon fibers with silicon carbide fibers. Inhibitors retard the reaction rate,

preventing oxygen from diffusion into carbon fiber [172]. In addition, coating works as an

oxygen diffusion barrier and protect carbon fibers from the oxidative environment. Kern

and Gadow [173] used silazanes, siloxanes, borasilazanes, phenolic resins and coal tar

pitch as a polymer precursor for continuous liquid coating of carbon fibers. Kim et al.[174]

enhanced carbon nanofiber oxidation resistance by deposition SiC/SiO2 through a sol-gel

process. The sample characterization shows the formation of a homogenous and continuous

coating that offers good protection from 1000-1400 oC. Fukuoka et al. [21] proposed a new approach to deposit silicon carbide on graphite material. The coating is carried out by thermal spraying of silicon powder on a graphite substrate, then heating the two components in an inert atmosphere at a temperature between 1100 oC and 1700 oC.

Eventually, a uniform SiC layer with low gas permeability is formed which offer good protection at high temperature. However, coating with the SiC layer results in cracks development which creates an easy path for oxygen diffusion into carbon fibers [168][173].

The coating thickness should be manipulated to be thick enough in order to isolate carbon from the oxygen and thin to keep the mechanical properties [170][175][176]. In addition, coating- related issues such as thermal expansion, crack formation, uniformity, and cost hinder the applicability of coating.

The third approach is summarized by the whole replacement of carbon fibers by

SiC fibers to overcome the low oxidation resistance of carbon fibers. One of the most common routes in the fabrication of ceramic nanofibers is electrospinning, which involves a mixture of organic/inorganic materials and solvent [177] [110]. To fabricate such fibers, a polymer derived ceramics (PDCs) offers a series of advantageous characteristics such as a high-temperature structural material, including high melting temperature, low density, high mechanical strength, high elastic modulus, excellent thermal stability, chemical inertness, and low thermal conductivity. Brewer et al.[178] investigated the oxidation behavior of SiOC which derived from four distinct silsesquioxane. After heat treatment, the resulted ceramics were tested qualitatively and quantitatively. To do so, the material was exposed to elevated temperatures (600, 800, 1000, and 1200 oC) with a hold for a period of 500 hours. The results showed that the oxidation resistance is connected to the

initial carbon content in the precursor. During oxidation, the silsesquioxane with an excess of carbon showed a higher weight loss. However, a precursor with low carbon content showed excellent oxidation resistance. From the data, one of the silsesquioxanes has only

0.6% weight loss after exposure to 600 oC and holding for 500 hours. For such precursors, the carbon content was over its half value even after exposing it to 1200 oC for 500 hours.

However, replacing carbon fiber with SiC nanofiber or ceramic fibers is not a valid solution because SiC-based fibers are brittle.

Another approach was considered by the coupling inorganic materials with organic moieties to form a hybrid composite. Silicon and silicon oxide are ones of the important nanoparticle which can be embedded in the polymer solution to achieve a new hybrid material[75]. Lee et al.[76] incorporated SiO2 nanoparticle in PAN polymer to fabricate

SiC nanofiber using electrospinning technique. After using many characterization tools, the SiO2/C ratio played a key role in controlling the fibers' morphology and structure. In addition, TGA results proved that the fibers had excellent oxidation resistance. However, the fabricated nanofibers were porous, especially with increasing SiO2 content which makes it unsuitable for application with high mechanical loading. Another advantage of Si and carbon hybrids is that SiC is formed after heat treatment and the carbon phase is well- ordered. Wu et al [179] successfully fabricated graphene on the cubic SiC/Si substrate.

This procedure is considered as an effective method in ordering graphitic carbons.

In addition, the SiC crystals, their size, and distribution played a key role in controlling the oxidation resistance and mechanical properties. The kinetics and thermodynamics of the SiC grain growth in ceramics are the same as in metals except for some exclusions that relate to ceramic porosity and non-stoichiometry [180]. Tobias et

al.[181] studied the impact of carbon content on the crystallization kinetics of SiC. They

successfully determined crystallization mechanism and activation energy for SiC growth.

The objective of this chapter is to fabricate hybrid C-SiC nanofibers from

precursors with varying PAN:Si ratios. The resulted fibers take advantage of high

mechanical properties of carbon fibers and high oxidation resistance of SiC fibers.

Furthermore, an experimental procedure like silicon to carbon precursor ratios and heat

treatment was optimized to fabricate hybrid nanofiber with high oxidation resistance and

mechanical properties. Finally, to optimize the heat treatment temperature, which controls

the ceramic phase size and then the properties, the kinetic study was carried out.

5.2 Experimental Setup

5.2.1 Materials

Polyacrylonitrile (6% methyl acrylate copolymer) with a molecular weight of about

100,000 g/mol was supplied by Scientific Polymer Products Inc. The polymers were

dissolved in 10 ml DMF. Different weight percentages of Si nanoparticles (from sky spring

Inc.) were added to the mixture as shown in Table 4. The total solid (PAN and Si

nanoparticles) to solvent weight percentages were kept constant ( 90 wt% DMF:10 wt%

solid).

Table 4 PAN to Si ratio.

PAN wt% Si wt% 100 0.00 99.5 0.50 99.0 1.00 95.0 5.00 90.0 10.0

A hot plate with a magnetic agitator was used to heat up the mixture and stir it for

12 hours until all the components were fully mixed. Nanofibers were produced using the electrospinning apparatus. This machine includes a pump (New Era pump systems, Inc.

NE-300), a voltage controller (Stanford Research Systems, Inc. Model P.S375), a rotating collector (Dayton® DC Motor, 4Z145), a syringe, and needle. The applied voltage was around 15 kV, the feeding rate was about 2휇m min-1, the distance between the syringe tip and the collector was 25 cm.

The stabilization of as-spun fibers was performed in air using a programmed tube furnace to carry out the multi-step procedure. The thermal stabilization was set at 230 oC for 1 hour; then, it increased to 250 oC and hold for 2 hours. Finally, the temperature was increased to 280 oC and held for another 2 hours. The heating rate for all steps was kept constant at around 1 oC min-1. The cured samples were cooled in the furnace until room temperature.

The stabilized fibers were exposed to carbonization heat treatment in argon gas, which includes heating the fibers to 1250 oC with the slow heating rate (1.5 oC min-1) and holds for 1 hour at the final temperature.

SEM was utilized to examine the morphology of the fibers and to measure the nanofiber dimensions (using Image J software) [111][112].

The tensile test was used to investigate the effect of silicon content on tensile strength and Young’s modulus. Samples were cut into strips parallel to fiber longitudinal directions. These specimens were in a form of fiber mat which, measured by connecting the tension stage with an optical microscope. XRD (RIGAKU, JAPAN) with a voltage of

40 kV and a current of 44 mA was used to track phases formation. The CuKα radiation

(휆=1.541 Ao) and a scan rate of 1 deg/min were used to perform XRD. The carbon yield of the precursors was measured using a TGA (Q500, TA Instruments, New Castle, DE) in the

o o N2 atmosphere, the heating program was set to be ramping to 800 C at 10 C/min. Raman spectrum of carbonized fiber mats was studied using a Renishaw in-Via Raman

Microscope with a 633 nm laser.

5.3 Results and Discussion

5.3.1 Scanning Electron Microscopy

To understand the effect of silicon content on fibers morphology, SEM micrographs of varying ranges of silicon contents were shown in Figure 38. The fiber morphology consists of long and straight fibers. Some Si nanoparticles are very noticeable on the surface, especially for samples with high silicon contents (Figure 38 c and d). With increasing the silicon content, the surface of the fiber appeared very rough.

After exposing the fibers to elevated temperatures, the fibrous mat showed some waviness and relaxed morphology as shown in Figure 39. Furthermore, some of the silicon migrates outside the fibers due to the fibers shrinkage after heat treatment which limits their ability to hold a large amount of silicon.

Figure 38 SEM micrograph for as-spun fibers mat with different Si concentrations a) 0 wt% Si, b) 0.5 wt% Si, c) 1 wt % Si, d) 5 wt% Si e) 10 wt%Si. Finally, the average nanofiber diameters before and after heat treatment have changed significantly (Figure 40).

Figure 39 SEM micrograph for carbonized fibers mat with different Si concentrations a) 0 wt% Si, b) 0.5 wt% Si, c) 1 wt % Si, d) 5 wt% Si e) 10 wt%Si.

Figure 40 The effect of silicon content on fiber diameter.

5.3.2 Mechanical Test

Although the single fiber segment mechanical testing is very helpful in examining

the nature of carbon fibers, such studies are not an indicator of defects or properties at a

larger scale [23]. In addition, due to the nano-metric size of the fibers, it was very

challenging to find the mechanical properties of the single fiber. As a result, a fibers mat

with many fibers and several microns thickness in the form of a strip was used to estimate

the mechanical properties of the fibers. The effect of adding Si on fibers strength and

stiffness was calculated by taking the average of five samples for each weight ratio. The

resulted mechanical properties, like tensile strength and Young’s modulus of carbonized

PAN/Si nanofiber (at 1250 oC for one hour) with different silicon contents, are given in

Figure 41. For lower concentrations, fiber stiffness was increased with increasing silicon

content till 1 wt% Si due to SiC formation which adds a ceramic character to the material.

After this number, the modulus decreased. The decline in Young’s modulus for fiber with

0.5 Si and may be attributed to the bad distribution of Si nanoparticle within the fiber.

The strength of the material also increased with adding silicon to the PAN, and the

highest strength was recorded for 0.5 wt% Si. After this percentage, the strength of

carbonized samples begun to decline due to ceramic phase formation which has lower

strength. Although the threshold for best strength was 0.5 wt% Si, increasing silicon

content over the threshold kept the strength higher than the pure PAN strength.

Figure 41 The effect of Si content on fiber mats a) strength b) Young’s modulus.

5.3.3 Thermogravimetric Analysis

TGA results for pyrolyzed PAN-Si nanofibers with different silicon weight percentages are shown in Figure 42. TGA test was conducted in the air with a heating rate of 10 oC/min from room temperature to 800 oC. From the TGA curves, carbon fibers without silicon undergo a dramatic thermal degradation at 500 oC with very high weight loss (9% char). On the other hand, carbon fibers with 10 wt%Si showed a noticeable improvement in oxidation resistance and char yield as compared with pure carbon fibers.

These results attributed to the SiC formation in carbon-silicon hybrid fibers at high carbonization temperatures as it was approved from XRD peaks. The resulted SiC improved the oxidation results for carbon fibers. A ceramic phase could be on the fiber surface which works as a coating layer isolating fibers from oxygen or as a second phase within the fiber working as a composite.

Figure 42 The relationship between Si content and weight loss for carbonized fibers at 1250 oC. The char yield for different Si contents is shown in Figure 43 (a). In this Figure, the oxidation resistance showed a significant increase with increasing Si content. The relationship between silicon content and char yield was fitted to be as y=푎푥푏, where “a” is around 37.2 and “b” is about 0.3.

(a)

(b)

Figure 43 TGA results a) the effect of Si content on char yield, b) the relationship between the silicon content and oxidation on onset temperature. In addition, the onset temperatures of oxidation were increased with increasing silicon content as shown in Figure 43(b). This increase; again, was related to an increase in the SiC domain with the increasing silicon content.

5.3.4 Raman Spectroscopy

One of the most accurate tools in examining structural features such as crystalline defects in carbon-based materials is Raman spectroscopy [182]. Two distinguishable

regions are always present in Raman spectra which are: first-order region (1100 to 1800

-1 -1 cm ) and second-order region (2200 to 3400 cm ). In the first region, graphite E2g vibration mode resulted from an aromatic layer of sp2 carbon around 1580 cm-1 frequency

(well-known as G band). In well-ordered graphite, the only peak in the first-order region is G band. However, for less perfect carbon materials, and a band called D band is possible to appear around 1350 cm-1 which attributed to defected or less ordered carbon [183][184].

The position, width, and intensity of these two bands are indicators for carbon materials properties and structure. The results shown in Figure 44, confirmed that the effect of silicon content on the position of D band (1330 cm-1) and G band (1570 cm-1) was almost negligible.

To estimate the contents of the defect in graphite, the ratio between D and G band

(ID/IG) was used [185]. The effect of silicon content on this ratio was given in Figure 45.

It is very obvious that silicon content has an inverse effect on the ID/IG ratio. The decrease in this ratio with increasing Si means that silicon or later SiC phases facilitate carbon ordering. The calculated ID/IG was relatively low (around 1.58) after adding silicon to PAN which means it helps to order carbon structure. In contrast, Ji et al [70] found a very high

ID/IG ratio (more than 4) for PAN/Si nanofiber. In addition to nanofiber with 10 wt% Si, a small 2D band around 2750 cm-1 was very obvious. The 2D peak is a secondary D band, which results from a second-order or two-resonance process [186]. This peak was generated due to the high percentage of SiC content for 10 wt% Si samples which help graphene or nanotubes growth [187]. It was reported that Si and SiC are excellent catalysts in the ordering of carbons and growth of graphene layers on their surfaces [188] [189].

However, none of them succeed in ordering and forming graphene within nanofiber using the freshly formed SiC phase.

Figure 44 The effect of silicon content on the Raman shift.

Figure 45 The effect of silicon content on ID/IG ratio.

5.3.5 Transmission Electron Microscopy

Two samples were sliced using an ultramicrotomy tool and then analyzed using the bright field and lattice fringes imaging techniques. The magnification of bright-field image

(Figure 46 (a)) showed that the pure PAN samples consist of carbonaceous material with a short-range two-dimensional carbon ordering (single arrow indicating turbostratic carbons in Figure 46(b). However, the sample with added silicon displays a very different carbon texture. As shown in Figure 47 (a) and (b), small silicon particles are well dispersed and large particles are located at the fiber skin, which coincides with the previous researcher work [190][64]. However, the carbon phase is very distinct. The carbon molecular orientation is very apparent and the order is well established (single arrow in Figure 47(b)).

This advanced change in carbon ordering might be a result of a catalytic reaction or a result of stress graphitization that occurred during heat treatment. This result was confirmed by the Raman data listed above.

Figure 46 Carbonized fibers at 1250 oC pristine PAN a) Bright field image, b) 002 lattice fringes of carbon phase developed in the pristine PAN sample.

Figure 47 Carbonized fibers at 1250 oC silicon added PAN nanofiber a) bright field, b) 002 lattice fringes of the carbon phase.

5.3.6 X-Ray Diffraction

To confirm or deny SiC formation, XRD was used. The two main SiC structures are 훽 −SiC (cubic structure) and 훼 − SiC (hexagonal structure). The 훽 −SiC forms at lower temperatures (below 1700 oC), while 훼 − SiC began to form at higher temperatures.

In XRD test, to calculate the interatomic spacing, Bragg equation [191][192] was used as:

휆=2dsin휃 ….…(9)

Where 휆 is the wavelength of XRD (1.5418 Ao), d is interatomic distance, and 휃 is diffraction angle.

The results of XRD for the nanofiber with different silicon contents are shown in

Figure 48. For 0.5 wt% Si and 1.0 wt% Si two broad peaks around 2θ = 25.19 o which is corresponded to (002) graphitic layer and another hump around 2θ = 44.0o which is related to (100) signal of turbostratic carbon plane [193][113][130]. The interatomic spacing (d002

and d100) was calculated using Bragg’s law. After calculations, the d002 was around 0.35312 nm and d100 about 0.20555 nm. To find the interatomic distance with increasing Si content above 5.0 wt% Si a more characteristics SiC peaks begun to develop. For such high silicon contents, XRD pattern confirmed a formation of SiC major peaks which recorded for 2θ

o o o =35.5 , 60 , and 72 [194]. The interatomic spacing for such peaks was d111= 2.51,

o o, o d220=1.53, and d311=1.309 for 2θ =35.5 , 60 and 72 , respectively. The intensities of these peaks were related to the initial silicon percentages. For instance, the sample with 10 wt%

Si showed the most developed SiC peaks, and below 5.0 wt% Si is hard to see SiC peaks.

Figure 48 XRD peaks for different precursors.

To estimate crystal size using XRD diffraction peaks, a Scherrer equation was used

[193] as follows:

푘.휆 L= ….…(10) 퐵. 퐶푂푆휃

This equation is used mainly to find the relationship between the crystalline dimension (L) and peak broadening B. The constant K depends immensely on the crystal

shape. For spherical crystals with cubic symmetry, K is around 0.95 [195]. The crystal size estimation for SiC resulted from different precursors is shown in Table 5.

Table 5 The relationship between the SiC crystal size and initial Si concentration Sample 2휽o Crystal size (nm) Phase name (hkl) 10.0%Si 35.6 10.25 SiC (111) 10.0%Si 60.16 10.59 SiC (220) 10.0%Si 71.8 10.24 SiC (311) 5.00%Si 35.74 11.62 SiC (111) 5.00%Si 60.08 9.58 SiC (220) 5.00%Si 72.1 11.40 SiC (311) From the oxidation resistance results, the highest char (the best oxidation properties) was given to 10.0% Si nanofiber. The same concentration showed improved mechanical properties compared to pure PAN. As a result, an extensive study on including the kinetics of SiC growth was performed for 10.0% Si. By utilizing the Scherrer equation to find the crystal size, mathematical calculations were performed to study the SiC growth mechanism as shown in the following equations [181][194][196]:

r= 푛√퐾푡 ….…(11)

Where r is crystal size, t is time, n is an indicator of grain growth mechanism, and

K is related to diffusion coefficient DB of the atom which calculated using the following formula[181]:

퐸 (− 푎) ….…(12) K∝ DB ∝ 푒푥푝 푅푇

Where R is the universal gas constant and T is the absolute temperature. The crystal size of SiC at three temperatures (850 oC, 1000 oC, and 1250 oC) with three holding times

(1, 4, 8 hours) is shown in Figure 49(a). In addition, the X-ray results for nanofiber treated to 1250 oC at three holding times is shown in Figure 49(b).

a b

Figure 49 a) The relationship between the crystal size and carbonizing temperature and holding times, b) the XRD peaks for 10.0%Si at 1250 oC. By correlating equation 11 and Figure 49, the ‘n’ for each temperature was found to be constant around 4, which indicates that the main growth mechanism is the coalescence of the second phase by grain boundary diffusion as shown in

Table 6. To find the activation energy for such a growth mechanism, the parameter

K for 850 oC, 1000 oC, and 1250 oC should be calculated using equation 12. The linear relationship between the natural logarithm of K and 1/T can be utilized to calculate the activation energy Ea (slope =Ea/R) as shown in Figure 50. From the Figure, the activation energy was calculated to be Ea= 35 KJ/mol.K. The activation energy and “n“ estimation in the kinetic calculations are very crucial in predicting the SiC phase growth. The control of SiC crystal size directly affected the mechanical and oxidative properties of fibers as mentioned earlier.

Figure 50 The relationship between ln(K) and 1/T.

Table 6: Kinetic of grain growth [197]. Pore control Surface diffusion 4 Lattice diffusion 3 Vapor transport P=const. 3 Vapor transport (P=2S/r) 2 Boundary control Pure system 2 Impure system Coalescence of the second phase by lattice diffusion 3 Coalescence of the second phase by grain boundary diffusion 4 A solution of the second phase 1 Diffusion through the continuous second phase 3 Impurity drag (low solubility) 3 Impurity drag (high solubility) 2

All the mentioned calculations were beneficial in the hybrid systems, as the most important criteria in such a system are the second phase growth. For example, the coarsening of the SiC phase will give the fibers ceramic properties and they are no longer hybrid fibers. On the other hand, very small SiC grains will not be very effective in protecting the fibers from oxygen, as shown in Figure 51. As a result, knowing the kinetic behind grain growth did help in controlling the second phase size.

Figure 51 TGA results for 10.0 wt%Si at different temperatures.

5.4 Conclusions

Hybrid C-SiC nanofiber was successfully fabricated by mixing carbon-based polymer (PAN) and silicon nanoparticle as precursors using the electrospinning technique.

After stabilization and carbonization to 1250 oC, the structure was mainly composed of turbostratic carbons and SiC nano-phases. These structural changes showed a positive effect on the mechanical and thermal properties of the nanofiber. TGA curves showed that carbon nanofiber without silicon underwent a dramatic thermal degradation at 500 oC with

very high weight loss (9% char). On the other hand, carbon nanofiber with 10 wt% Si showed a noticeable improvement in the oxidation resistance and char yield as compared to a control sample. In terms of mechanical properties, a very noticeable improvement, especially for low Si concentrations, in both tensile strength and modulus was confirmed.

Finally, the SiC grain growth analysis demonstrated that the determining factor in the growth mechanism was the interfacial diffusion of the second phase.

CHAPTER VI

HYBRID CARBON-SiC NANOFIBER WITH IMPROVED OXIDATION

RESISTANCE

6.1 Introduction

CFs are fibers consisting of about 90% carbon elements with high strength (3–7 GPa) and modulus (200–500 GPa) [30]. They display good thermo-chemical stability in nonoxidative environments. CFs can have good thermal and electrical conductivities, and excellent creep resistance [21]. They have been utilized in many applications, such as aerospace [198], highway transportation [199], military [200], turbine blades [201], construction [202], energy conversion and storage [203], and self-sensing devices [204].

CFs can be fabricated using polymeric precursors, such as acrylic, cellulosic, pitch, and

PAN precursors [31][23][24] by subsequent heat treatment.

The mechanical properties of the resulting fibers could be ultra-high modulus (>500

GPa), high modulus (>300 GPa), intermediate modulus (>200 GPa), or low modulus (100

GPa), depending on precursor and fabrication method [32]. For example, Polyacrylonitrile

(PAN)-based carbon materials have lightweight and high mechanical properties. At the same time, the low oxidation resistance of carbon fibers has been the focus of significant research. Yuanjian et al. [30] investigated the effect of oxidative environment on the mechanical properties of two types of PAN-based CFs, namely CF-A (carbonized at 1350

°C) and CF-B (carbonized at 1450 °C). The fibers were subjected to a temperature between

400 °C to 700 °C in air. The results showed that CNF began to react with oxygen at 400

°C. At 600 °C, a noticeable weight loss was found, which ended with burning about 13%

of total mass at 700 °C. As a result, the tensile strength of fiber decreased significantly, but the modulus was found to remain constant.

One approach to improve the properties of CFs is by utilizing a multi-functional hybrid materials approach, which yields unique and novel properties as a result of the synergistic combination of materials properties [9]. In the past, extensive work was performed on alloying different metals to create metal alloys that display improved properties. As lightweight becomes critical design criteria, metal alloys are being replaced with fiber- reinforced composites. Theses composite showed improved physical properties with better oxidation and corrosion resistance than metals. The same strategy can be applied to CFs in order to improve its low oxidation resistance.

To fabricate carbon and silicon-based ceramics, smart use of polymer (organic and inorganic) precursors can be utilized. It enables forming complex shapes at low temperatures [55]. Several studies used both carbon precursors with solid additives to produce a new nanocomposite, which leads to different material properties [70],[71]. For example, PVA/Silica [72], PAN/Cu [73], and MgO/Al2O3/PAN [74] systems were used to fabricate carbon materials composites. Silicon and silicon oxide are two important nanoparticles, which can be embedded in the polymer solution to achieve a new hybrid material. For instance, C-Si hybrid nanofiber can be prepared electrospinning which followed by curing and pyrolysis heat treatments. Lee et al. [76] added SiO2 nanoparticles in a PAN to fabricate SiC nanofibre using the electrospinning technique. After using many characterization tools, the SiO2/C ratio played a key role in controlling the fibers' morphology, structure, and properties. TGA results showed that the fibers had an excellent oxidation resistance. Saja et al. [64] used an organic/inorganic polymer blend

(polydimethylsiloxane and polyacrylonitrile) to fabricate hybrid ceramic fibers. The resulting structure consists of SiC, carbon, and SiOC. The fibers showed improved mechanical properties as compared to pristine carbon fibers.

In this chapter, we report results from a study with the objective of improving carbon fibers oxidation resistance by incorporating silicon source, which after heat treatment, converts into nano-dispersed silicon carbide, thus providing protection to oxidation, without affecting the intrinsic fiber properties. The resulted hybrid fibers have the potential for being used at higher temperatures in a wide range of applications. The carbonization temperature and time were optimized to achieve the best combination of thermal and mechanical properties.

6.2 Experimental Setup

A polyacrylonitrile copolymer (molecular weight about 100,000 g/mol) with 6% methyl acrylate copolymer was acquired from scientific polymer products Inc. A 10 mL of

DMF was used to dissolve PAN and form a viscous fluid. Si nanoparticles from Sky Spring

Inc. (Houston, TX, USA), which served as a silicon source, were added to the solution. The

PAN/Si ratio was 90/10 wt %. However, the total weight percentage of PAN and silicon in the DMF solvent was kept constant (10 wt % Si) in 90 wt % DMF.

A magnetic agitator with a heated stage was utilized for dissolving and mixing. After stirring for 12 h, the mixture was loaded into an electrospinning apparatus to fabricate nanofibers. The machine has a voltage controller (Stanford Research Systems, Inc.

(Sunnyvale, CA, USA) Model P.S375), a pump (New Era pump systems, Inc.

(Farmingdale, NY, USA) NE-300), a rotating collector (Dayton® DC Motor, 4Z145), a

syringe, and a needle. The electrospinning parameters were: 15 kV voltage, 2 휇m min−1 feeding rate, 25 cm distance from the syringe tip and collector. The as-spun fibers were stabilized (in air) in a tube furnace, and by using a multi-step procedure, which involves:

1. Heating from 25 °C to 230 °C and holding for 1 h.

2. Heating from 230 °C to 250 °C and holding for 2 h.

3. Heating from 250 °C to 280 °C and holding for 2 h.

The heating rate was kept constant (1°C/min), while the cooling segment followed the cooling rate of the furnace until room temperature.

The cured fibers were carbonized in a tube furnace (in argon gas) to thermally process the fibers at 850 °C, 1000 °C, and 1250 °C, and holding times 1, 4, and 8 h, respectively.

The heating rate was kept constant (1.5 °C min−1).

Fiber morphology and dimensions were characterized using SEM coupled with Image

J software [111][112]. Mechanical properties of different heat-treated fibers were carried out by cutting the fibers mat into small strips (parallel to the fibers longitudinal direction).

The fibers' mat dimensions were measured by connecting the tension stage to the optical microscope.

X-ray diffraction (RIGAKU, Tokyo, Japan) was utilized to examine the phase formation. The X-ray parameters included 40 kV, a current of 44 mA, CuKα radiation (휆

= 1.541 A°), and a scan rate of 1 deg/min. To complement the XRD data, Renishaw in-Via

Raman Microscope with 633 nm laser was used. Oxidation resistance test was performed using thermal gravity analysis (TGA, Q500, TA Instruments, New Castle, DE, USA).

6.3 Results and Discussion

6.3.1 Scanning Electron Microscopy

To understand the morphological changes during heat treatments, SEM characterization was carried out for all fibers, subjected to different temperatures (Figure

52). As-spun fibers were straight and regular with limited segregation of Si nanoparticle to the surface. After heat treatment, the fibers begun to shrink due to the evaporation of volatile compounds. The fibers size reduction lowers their ability to hold a large amount of nanoparticle and most of the silicon migrates to the surface of the fiber, which forms a protective SiC layer. In addition, the fibers become wavy with increasing the carbonization temperature. Average values from the characterization of 20 fibers were estimated using

Image J software and the results are shown in Figure 53.

Figure 52 SEM images of treated fibers (holding time four hours) at a) 25 °C, b) 850 °C, c) 1000 °C, d) 1250 °C.

Figure 53 Relationship between fibers diameter and heat treatment temperature. Data points correspond to the average from 20 fibers. Error bars correspond to one standard deviation with respect to the average value.

6.3.2 Nanofiber Structural Examinations

X-ray diffraction was used to confirm and track SiC formation. In general, SiC has around 200 polytypes, which are called polymorphs [205]. The two common crystalline forms of SiC are cubic (훽 −SiC) and hexagonal (훼 − SiC) forms. To identify various phases in XRD test, the interatomic spacing was calculated using the Bragg equation [192] as:

휆 = 2dsin휃 (1)

Where 휆 is the wavelength of X-ray (1.5418 A°), d is interlayer distance, and 휃 is diffraction angle.

X-Ray results of treated nanofiber at different temperatures are shown in

Figure 54a. For fibers treated at temperatures 1000 °C and below, the fibers have three main domains, which were carbon, SiC, and silicon. This means that SiC has begun to form at temperatures as low as 850 °C, but it was not enough to convert the whole silicon into

SiC. Furthermore, increasing the holding times at lower temperatures did not help in reacting the remaining Si (Figure 54 b). Longer holding times improved the carbon structures in the fibers as the hump around 2θ = 26° became more characteristics toward longer times, but it does not affect the SiC to Si amounts. Carbonization at elevated temperatures (more than 1250 °C) was enough to convert all Si to SiC, which was confirmed by the X-ray results. At this temperature, the SiC major peaks were detected at

2θ = 35.5° and 72° [194], which belonged to an interatomic spacing d111 = 2.51, d220 =

1.53, and d311 = 1.309, respectively.

Figure 54 XRD results of nanofiber mat heat treated at a) three distinct temperatures and one- hour holding time, b) 1000 °C at three distinct holding times.

Raman spectroscopy is another powerful tool for characterizing the structure of carbon-based materials. Nanofiber measurements were done at three different temperatures, as shown in Figure 55. The two main peaks around at 1352 and 1588 cm−1 are related to D and G band [206], respectively. Such peaks describe the molecular picture of carbon materials. The D band represents the defect in carbon structure or low ordered carbon structure in the graphite domain, which means that the D band is characteristic of sp3 hybridization. On the other hand, G band represents sp2 hybridization in the graphitic structure [124]. In addition, one of the important ratios in Raman spectroscopy is D to G ratio (ID/IG), which indicates the degree of graphitizability. From Figure 55, the ID/IG is inversely related to the heat treatment temperature. The ratio for higher temperatures showed a lower ratio, which means a better graphite structure. In addition, the highest heat treatment temperature (1250 °C) showed a 2D peak around 2750 cm−1.

Figure 55 Raman spectroscopy of carbon fibers at three distinct temperatures.

6.3.3 Thermogravimetric Analysis

TGA results for carbonized PAN-Si nanofiber at different temperatures and holding times are shown in Figure 5. After carbonization, the fibers exposed to the oxidative environment to measure their thermal resistance. To do so, the TGA test was conducted in the air with a heating rate of 10 °C/min from room temperature up to 800 °C. The TGA results are summarized in Table 7. It was found that the carbonized fibers at 850 °C showed the lowest oxidation resistance and char yield. Carbonized fibers at higher temperatures showed a noticeable improvement in oxidation resistance and char yield. This confirms the effect of SiC phase growth at high temperatures in improving the carbon fibers' oxidation resistance. TGA results (Figure 56) showed that the oxidation resistance of the fibers doubled due to increasing the holding time from 4 h to 8 h. The reason behind the significant improvement explained by a previously published study regarding the silicon particle distribution [190]. It was confirmed experimentally using TEM (Figure 57a) and numerically that most of the large silicon nanoparticles were located on fibers skin and the small particles were distributed randomly. This means that the intensity of silicon, initially, on the wall was higher than the center Figure 57b. Silicon atoms began to combine with C to form SiC around 850, which results in the consumption of Si, as shown in Figure 57c, d. After 1250 °C, the whole silicon was consumed and converted to SiC Figure 57e–g).

The long holding times (Figure 57f) showed the highest oxidation resistance as the SiC formed a continuous silicon carbide coating layer in addition to the reinforcing effect within the fibers. TGA results confirm that the current method is an excellent and cheap way of protecting carbon fibers. Prakash et al. [207] improved the oxidation resistance of carbon fibers by achieving 16% char after coating CFs with SiC nanowires as compared

with 0% char for uncoated carbon fibers. Zhou et al. [208] improved the carbon fibers oxidation resistance by growing silicon carbide on carbon fibers using catalysis chemical vapor deposition (CCVD), which increased the char yield to around 40%. In our observation, we have maximum 60% char for hybrid carbon-SiC nanofiber.

Table 7 Influence of pyrolysis temperature and soaking time on hybrid carbon fibers char yield.

Heat Treatment Holding Time Char Yield Temperature (°C) (Hours) (%) 850 1 29 1000 1 33 1 37 1250 4 41 8 62

Figure 56 TGA curves of hybrid fibers at different soaking times at 1250 °C.

Figure 57 SiC nucleation and growth from Si and C phases a) TEM image of as-spun PAN-Si nanofiber, b) as-spun PAN-Si nanofiber illustration, c) SiC nucleation, d) SiC growth, e) conversion of all Si to SiC, f) SiC grain coarsening.

6.3.4 Mechanical Test

The mechanical properties of fiber mats instead of single fibers were used to avoid the challenges that are related to single fibers measurements. The fibers mat consisting of several microns thick was cut into strips to measure the fibers' tensile strength. The effect of carbonization temperature on fibers' mechanical properties was measured and is shown in Figure 58. The tensile strength increased from 10 MPa to 21 MPa after carbonizing the material to 850 °C. After this temperature, the strength of carbonized samples begun to decline due to ceramic phase formation, which has lower strength. The slope of the stress- strain curve showed a continuous increase with increasing the temperature due to the SiC growth which gives the material more ceramic character.

Figure 58 The relationship between mechanical properties and heat treatment temperature.

6.4 Conclusions

PAN and Si nanoparticle were used to fabricate carbon-silicon carbide (C-SiC) hybrid nanofiber. The pyrolysis temperature and holding time played a key role in changing the microstructure and the resulting thermal and mechanical properties. The complete conversion of silicon to SiC happened at 1250 °C, which means that, below this temperature, three phases co-exist (SiC, Si, and C), as confirmed from XRD results.

Therefore, the char yield was boosted with increasing the carbonization temperatures and soaking time due to SiC phase growth, which works as a coating and reinforcing phase. An excellent char yield around 60% was recorded for carbon hybrid fibers with 8 h holding time at 1250 °C. Such observation provides high oxidation protection for carbon fibers at low cost. Furthermore, the slope of the stress-strain curve increased its values with raising the heat treatment temperatures; however, the highest strength was recorded for lower carbonization temperature.

CHAPTER VII

CONCLUSIONS AND RECOMMENDATIONS

7.1 Conclusions

An innovative processing method to fabricate C-SiC fibers using the electrospinning technique was successfully demonstrated in this work. The fiber was processed using two precursor routes. The first route used blends of PAN and a silicon- based polymer (PDMS), while the second one involved dispersions of silicon nanoparticles in PAN.

The polymer blend approach resulted in fibers having a core-skin structure with the core rich in ceramic phases (SiC and SiOC) and the skin rich in carbon. In addition, the resulting fibers showed improved mechanical properties as compared to SiC free carbon fiber produced with the same process. In comparison, the silicon nanoparticle approach resulted in the fabrication of hybrid C-SiC fiber with a gradient of particle sizes from the skin to the core. After experimental and numerical investigations, it was confirmed that silicon particles with large size have the ability to migrate to the fiber skin during the electrospinning process, while the smaller ones did not have enough inertia to move laterally and therefore retained their initial location. Numerical simulation of Si particles- size distribution validated the experimental results.

Furthermore, studying the effect of silicon to carbon ratio on the fiber oxidation resistance was very crucial in the fabrication of hybrid C-SiC fibers. Increasing the silicon content to 10 wt% resulted in a four-time improvement in char yield as compared to SiC free carbon fibers. SiC phase helped to form a two-dimensional ordered carbon phase

which enhanced the hybrid fibers' mechanical properties. To help quantify and study SiC nucleation and growth, a kinetic model was used to calculate the activation energy for SiC phase growth (35 KJ/mol.K) and it is exponent (n=4). This result led to the conclusion that the growth mechanism was grain boundary diffusion. In addition, it was demonstrated that the pyrolysis temperature and holding time had a major role in the SiC phase nucleation and growth process. It was confirmed that a complete transformation of Si to SiC occurred at 1250 °C. However, for heat treatments below 1000 °C, three distinct coexisting phases, including Si, C, and SiC were produced. From TGA data, the char yield showed linear increasing growth with increasing the pyrolysis temperature. In addition, holding for long times at higher temperatures (1250°C) showed a significant increase in oxidation resistance of the hybrid fiber because of SiC grain growth. At longer holding times, the SiC phase was observed to function as both an antioxidation coating and a mechanical reinforcing phase

7.2 Recommendations and Future Work

For recommendations and future work, it will be interesting to fabricate hybrid carbon-ceramic fibers using different preceramic polymers. Such polymers could be polycarbosilanes, polyborosiloxane, and polysilazanes and depending on the initial preceramic polymer chemistry the resulted hybrid material could be SiC, SiBOC, and

SiCN, respectively. Furthermore, the resulted hybrid fibers could be implemented in ceramics matrix composites to increase the mechanical properties.

REFERENCES

[1] G. Kickelbick, Introduction to Hybrid Materials, Citeseer, 2007.

[2] C. Sanchez, G.J.D.A.A. Soler-Illia, F. Ribot, D. Grosso, Design of functional

nano-structured materials through the use of controlled hybrid organic–inorganic

interfaces, Comptes Rendus Chim. 6 (2003) 1131–1151.

[3] Q. Lu, Synthesis of PDMS-Metal Oxide Hybrid Nanocomposites Using an in Situ

Sol-Gel Route by Qiaoyu Lu, Ph.D dissertation, Michigan Technological

University, Michigan Technological University, 2012.

[4] Glossary of nanotechnology and related terms, (n.d.).

http://eng.thesaurus.rusnano.com/ (accessed February 24, 2019).

[5] S.H. Mir, L.A. Nagahara, T. Thundat, P. Mokarian-Tabari, H. Furukawa, A.

Khosla, Review—Organic-Inorganic Hybrid Functional Materials: An Integrated

Platform for Applied Technologies, J. Electrochem. Soc. 165 (2018) B3137–

B3156.

[6] C. Sanchez, B. Julián, P. Belleville, M. Popall, Applications of hybrid organic–

inorganic nanocomposites, J. Mater. Chem. 15 (2005) 3559–3592.

[7] G. Kickelbick, Hybrid Materials – Past, Present and Future., Hybrid Mater. 1

(2014).

[8] P. Gómez-Romero, C. Sanchez, Functional hybrid materials, John Wiley & Sons,

2006.

[9] B.A.B. Morgan, M. Matters, V. Issue, Polymer-Clay Nanocomposites : Design and

Application of Multi-Functional Materials, Mater. Matters. (2011).

[10] P. Judeinstein, C. Sanchez, Hybrid organic–inorganic materials: a land of 93

multidisciplinarity, J. Mater. Chem. 6 (1996) 511–525. [11] M. Oschatz, R.

Walczak, Crucial Factors for the Application of Functional Nanoporous Carbon-

Based Materials in Energy and Environmental Applications, C . 4 (2018).

[12] V. Georgakilas, J.A. Perman, J. Tucek, R. Zboril, Broad Family of Carbon

Nanoallotropes: Classification, Chemistry, and Applications of Fullerenes, Carbon

Dots, Nanotubes, Graphene, Nanodiamonds, and Combined Superstructures,

Chem. Rev. 115 (2015) 4744–4822.

[13] C. Cha, S.R. Shin, N. Annabi, M.R. Dokmeci, A. Khademhosseini, Carbon-Based

Nanomaterials: Multifunctional Materials for Biomedical Engineering, ACS Nano.

7 (2013) 2891–2897.

[14] D. Xia, M. Otyepka, X. Li, W. Liu, Q. Zheng, Carbon-based materials at

nanoscale, J. Nanomater. 2015 (2015) 2–4. doi:10.1155/2015/750242.

[15] A.K. Geim, K.S. Novoselov, The rise of graphene, Nat. Mater. 6 (2007) 183.

[16] S. Das, P. Sudhagar, Y.S. Kang, W. Choi, Graphene synthesis and application for

solar cells, J. Mater. Res. 29 (2014) 299–319.

[17] T. Gatti, N. Vicentini, M. Mba, E. Menna, Organic Functionalized Carbon

Nanostructures for Functional Polymer-Based Nanocomposites, European J. Org.

Chem. 2016 (2016) 1071–1090.

[18] M.E. Casco, M. Martínez-Escandell, E. Gadea-Ramos, K. Kaneko, J. Silvestre-

Albero, F. Rodríguez-Reinoso, High-Pressure Methane Storage in Porous

Materials: Are Carbon Materials in the Pole Position?, Chem. Mater. 27 (2015)

959–964.

[19] L. Borchardt, Q.-L. Zhu, M.E. Casco, R. Berger, X. Zhuang, S. Kaskel, X. Feng,

Q. Xu, Toward a molecular design of porous carbon materials, Mater. Today. 20

(2017) 592–610.

[20] C. Kim, Y. Il Jeong, B.T.N. Ngoc, K.S. Yang, M. Kojima, Y.A. Kim, M. Endo, J.-

W. Lee, Synthesis and Characterization of Porous Carbon Nanofibers with Hollow

Cores Through the Thermal Treatment of Electrospun Copolymeric Nanofiber

Webs, Small. 3 (2006) 91–95.

[21] Hirofumi Fukuoka, GRAPHITE-SILICON CARBIDE COMPOSITE AND

MAKING METHOD, (2008).

[22] L. Feng, N. Xie, J. Zhong, Carbon Nanofibers and Their Composites: A Review of

Synthesizing, Properties and Applications, Mater. (Basel, Switzerland). 7 (2014)

3919–3945.

[23] C. Liu, K. Lafdi, Fabrication and characterization of carbon nanofibers from

polyacrylonitrile/pitch blends, J. Appl. Polym. Sci. 134 (2017) 45388.

[24] Z. Xu, C. Gao, Graphene fiber: a new trend in carbon fibers, Mater. Today. 18

(2015) 480–492.

[25] E. Frank, F. Hermanutz, M.R. Buchmeiser, Carbon Fibers: Precursors,

Manufacturing, and Properties, Macromol. Mater. Eng. 297 (2012) 493–501.

[26] O.P. Bahl, Z. Shen, J.G. Lavin, R.A. Ross, Manufacture of carbon fibers, Carbon

Fibers. 3 (1998) 1.

[27] H.G. Chae, B.A. Newcomb, P. V Gulgunje, Y. Liu, K.K. Gupta, M.G. Kamath,

K.M. Lyons, S. Ghoshal, C. Pramanik, L. Giannuzzi, High strength and high

modulus carbon fibers, Carbon N. Y. 93 (2015) 81–87.

[28] K. Naito, Y. Tanaka, J.-M. Yang, Transverse compressive properties of

polyacrylonitrile (PAN)-based and pitch-based single carbon fibers, Carbon N. Y.

118 (2017) 168–183.

[29] M. MInus, S. Kumar, The processing, properties, and structure of carbon fibers,

Jom. 57 (2005) 52–58.

[30] Y. Tong, X. Wang, H. Su, L. Xu, Oxidation kinetics of polyacrylonitrile-based

carbon fibers in air and the effect on their tensile properties, Corros. Sci. 53 (2011)

2484–2488.

[31] S.-J. Park, G.-Y. Heo, Precursors and Manufacturing of Carbon Fibers BT -

Carbon Fibers, in: S.-J. Park (Ed.), Springer Netherlands, Dordrecht, 2015: pp. 31–

66.

[32] X. Huang, Fabrication and Properties of Carbon Fibers, Materials (Basel). 2 (2009)

2369–2403.

[33] P. Lu, Q. Huang, A. Mukherjee, Y.-L. Hsieh, SiCO-doped Carbon Fibers with

Unique Dual Superhydrophilicity/Superoleophilicity and Ductile and Capacitance

Properties, ACS Appl. Mater. Interfaces. 2 (2010) 3738–3744.

[34] J. Gavillet, A. Loiseau, C. Journet, F. Willaime, F. Ducastelle, J.-C. Charlier,

Root-Growth Mechanism for Single-Wall Carbon Nanotubes, Phys. Rev. Lett. 87

(2001) 275504.

[35] K. Teo, C. Singh, Catalytic synthesis of carbon nanotubes and nanofibers, Encycl.

Od Nanosci. Nanotechnol. X (2003) 1–22.

[36] H.-Y. Cheng, Y.-A. Zhu, Z.-J. Sui, X.-G. Zhou, D. Chen, Modeling of fishbone-

type carbon nanofibers with cone-helix structures, Carbon N. Y. 50 (2012) 4359–

4372.

[37] K.P. DE JONG, J.W. GEUS, Carbon Nanofibers: Catalytic Synthesis and

Applications, Catal. Rev. 42 (2000) 481–510.

[38] I. Martin-Gullon, J. Vera, J.A. Conesa, J.L. González, C. Merino, Differences

between carbon nanofibers produced using Fe and Ni catalysts in a floating

catalyst reactor, Carbon N. Y. 44 (2006) 1572–1580.

[39] R.L. Poveda, N. Gupta, Carbon Nanofibers: Structure and Fabrication BT -

Carbon Nanofiber Reinforced Polymer Composites, in: R.L. Poveda, N. Gupta

(Eds.), Springer International Publishing, Cham, 2016: pp. 11–26.

[40] Z. Wangxi, L. Jie, W. Gang, Evolution of structure and properties of PAN

precursors during their conversion to carbon fibers, Carbon N. Y. 41 (2003) 2805–

2812.

[41] A.G. Odeshi, H. Mucha, B. Wielage, Manufacture and characterisation of a low

cost carbon fibre reinforced C/SiC dual matrix composite, Carbon N. Y. 44 (2006)

1994–2001.

[42] F. Lamouroux, S. Bertrand, R. Pailler, R. Naslain, M. Cataldi, Oxidation-resistant

carbon-fiber-reinforced ceramic-matrix composites, Compos. Sci. Technol. 59

(1999) 1073–1085.

[43] R.M. Rocha, C.A.A. Cairo, M.L.A. Graça, Formation of carbon fiber-reinforced

ceramic matrix composites with polysiloxane/silicon derived matrix, Mater. Sci.

Eng. A. 437 (2006) 268–273.

[44] M. Inagaki, Y. Yang, F. Kang, Carbon Nanofibers Prepared via Electrospinning,

Adv. Mater. 24 (2012) 2547–2566.

[45] S. Kumar, Advances in high performance fibres, (1991).

[46] W.P. Hoffman, W.C. Hurley, P.M. Liu, T.W. Owens, The surface topography of

non-shear treated pitch and PAN carbon fibers as viewed by the STM, J. Mater.

Res. 6 (1991) 1685–1694.

[47] N.R. Council, High-Performance Structural Fibers for Advanced Polymer Matrix

Composites, National Academies Press, 2005.

[48] S. Imaizumi, H. Matsumoto, K. Suzuki, M. Minagawa, M. Kimura, A. Tanioka,

Phenolic Resin-Based Carbon Thin Fibers Prepared by Electrospinning: Additive

Effects of Poly(vinyl butyral) and Electrolytes, Polym. J. 41 (2009) 1124.

[49] Y. Wang, S. Serrano, J.J. Santiago-Avilés, Raman characterization of carbon

nanofibers prepared using electrospinning, Synth. Met. 138 (2003) 423–427.

[50] S.P. and E.Z. and Y. Cohen, The effect of embedded carbon nanotubes on the

morphological evolution during the carbonization of poly(acrylonitrile) nanofibers,

Nanotechnology. 19 (2008) 165603.

[51] E.J. Ra, E. Raymundo-Piñero, Y.H. Lee, F. Béguin, High power supercapacitors

using polyacrylonitrile-based carbon nanofiber paper, Carbon N. Y. 47 (2009)

2984–2992.

[52] L.J. and X. Zhang, Fabrication of porous carbon nanofibers and their application as

anode materials for rechargeable lithium-ion batteries, Nanotechnology. 20 (2009)

155705.

[53] J.-P. Chen, C.-H. Su, Surface modification of electrospun PLLA nanofibers by

plasma treatment and cationized gelatin immobilization for cartilage tissue

engineering, Acta Biomater. 7 (2011) 234–243.

[54] J. Karger-Kocsis, ed., Polypropylene Structure, blends and composites, Springer

Netherlands, Dordrecht, 1995.

[55] F. Xu, D. Wu, R. Fu, B. Wei, Design and preparation of porous carbons from

conjugated polymer precursors, Mater. Today. 20 (2017) 629–656.

[56] W. Lengauer, A. Liersch, Polymer precursor ceramics and their applications,

(2004).

[57] M. Birot, J.-P. Pillot, J. Dunogues, Comprehensive Chemistry of Polycarbosilanes,

Polysilazanes, and Polycarbosilazanes as Precursors of Ceramics, Chem. Rev. 95

(1995) 1443–1477.

[58] H. Elsayed, P. Colombo, E. Bernardo, Direct ink writing of wollastonite-diopside

glass-ceramic scaffolds from a silicone resin and engineered fillers, J. Eur. Ceram.

Soc. 37 (2017) 4187–4195.

[59] T. Konegger, C.C. Tsai, R.K. Bordia, Preparation of Polymer-Derived Ceramic

Coatings by Dip-Coating, Mater. Sci. Forum. 825–826 (2015) 645–652.

[60] L.F.B. Ribeiro, O. Flores, P. Furtat, C. Gervais, R. Kempe, R.A.F. Machado, G.

Motz, A novel PAN/silazane hybrid polymer for processing of carbon-based fibres

with extraordinary oxidation resistance, J. Mater. Chem. A. 5 (2017) 720–729.

[61] H.-J. Kleebe, Y.D. Blum, SiOC ceramic with high excess free carbon, J. Eur.

Ceram. Soc. 28 (2008) 1037–1042.

[62] Y.D. Blum, D.B. MacQueen, H.-J. Kleebe, Synthesis and characterization of

carbon-enriched silicon oxycarbides, J. Eur. Ceram. Soc. 25 (2005) 143–149.

[63] K. Lu, D. Erb, M. Liu, Thermal stability and electrical conductivity of carbon-

enriched silicon oxycarbide, J. Mater. Chem. C. 4 (2016) 1829–1837.

[64] S. Al-Ajrah, K. Lafdi, Y. Liu, P. Le Coustumer, Fabrication of ceramic nanofibers

using polydimethylsiloxane and polyacrylonitrile polymer blends, J. Appl. Polym.

Sci. 135 (2017) 45967.

[65] M. Wan, W. Zhou, J. Li, Composite of polyaniline containing iron oxides with

nanometer size, Synth. Met. 78 (1996) 27–31.

[66] G. Bidan, O. Jarjayes, J.M. Fruchart, E. Hannecart, New nanocomposites based on

“tailor dressed” magnetic particles in a polypyrrole matrix, Adv. Mater. 6 (1994)

152–155.

[67] H. Yoneyama, A. Kishimoto, S. Kuwabata, Charge–discharge properties of

polypyrrole films containing manganese dioxide particles, J. Chem. Soc. Chem.

Commun. (1991) 986–987.

[68] C.-L. Huang, R.E. Partch, E. Matijević, Coating of Uniform Inorganic Particles

with Polymers: II. Polyaniline on Copper Oxide, J. Colloid Interface Sci. 170

(1995) 275–283.

[69] M. Matsumura, T. Ohno, Concerted transport of electrons and protons across

conducting polymer membranes, Adv. Mater. 9 (1997) 357–359.

[70] L. Ji, K.-H. Jung, A.J. Medford, X. Zhang, Electrospun polyacrylonitrile fibers

with dispersed Si nanoparticles and their electrochemical behaviors after

carbonization, J. Mater. Chem. 19 (2009) 4992–4997.

[71] Z.M. Huang, Y.Z. Zhang, M. Kotaki, S. Ramakrishna, A review on polymer

nanofibers by electrospinning and their applications in nanocomposites, Compos.

Sci. Technol. (2003).

[72] C. Shao, H.-Y. Kim, J. Gong, B. Ding, D.-R. Lee, S.-J. Park, Fiber mats of

100

poly(vinyl alcohol)/silica composite via electrospinning, Mater. Lett. 57 (2003)

1579–1584.

[73] P. Han, T. Yuan, L. Yao, Z. Han, J. Yang, S. Zheng, Copper Nanoparticle-

Incorporated Carbon Fibers as Free-Standing Anodes for Lithium-Ion Batteries,

Nanoscale Res. Lett. (2016).

[74] S. Dadvar, H. Tavanai, M. Morshed, Fabrication of nanocomposite PAN

nanofibers containing MgO and Al2O3 nanoparticles, Polym. Sci. Ser. A. 56

(2014) 358–365.

[75] A. Neilde, R. Da Silva, M.L. Pereira Da Silva, Silicon Compound Inclusion on

PAN Nanofibers, ICSMVA 2014

[76] S.-H. Lee, S.-M. Yun, S.J. Kim, S.-J. Park, Y.-S. Lee, Characterization of

nanoporous β-SiC fiber complex prepared by electrospinning and carbothermal

reduction, Res. Chem. Intermed. 36 (2010) 731–742.

[77] M. Streckova, E. Mudra, R. Orinakova, L. Markusova-Buckova, M. Sebek, A.

Kovalcikova, T. Sopcak, V. Girman, Z. Dankova, M. Micusik, J. Dusza, Nickel

and nickel phosphide nanoparticles embedded in electrospun carbon fibers as

favourable electrocatalysts for hydrogen evolution, Chem. Eng. J. 303 (2016) 167–

181.

[78] B.-H. Kim, S.Y. Kim, C.H. Kim, K.S. Yang, Y.-J. Lee, Preparation of anti-

oxidative carbon fiber at high temperature, Appl. Surf. Sci. 257 (2010) 662–664.

[79] K.L. Luthra, Oxidation of carbon/carbon composites—a theoretical analysis,

Carbon N. Y. 26 (1988) 217–224.

[80] M.A. Shayed, H. Hund, R.D. Hund, C. Cherif, Thermal and oxidation resistant

101

barrier on carbon fiber with Si and Si–Ti based pre-ceramic coatings for high

temperature application, Text. Res. J. 86 (2016) 1901–1913.

[81] Q. Wei, E. Pippel, J. Woltersdorf, M. Scheffler, P. Greil, Interfacial SiC formation

in polysiloxane-derived Si–O–C ceramics, Mater. Chem. Phys. 73 (2002) 281–

289.

[82] Y.-S. Jang, M. Jank, V. Maier, K. Durst, N. Travitzky, C. Zollfrank, SiC ceramic

micropatterns from polycarbosilanes, J. Eur. Ceram. Soc. 30 (2010) 2773–2779.

[83] M. NARISAWA, H. YASUDA, R. MORI, H. MABUCHI, K. OKA, Y.-W. KIM,

Silicon carbide particle formation from carbon black

—polymethylsilsesquioxane mixtures with melt pressing, J. Ceram. Soc.

Japan. 116 (2008) 121–125.

[84] P. Greil, Polymer Derived Engineering Ceramics, Adv. Eng. Mater. 2 (2000) 339–

348.

[85] Q.-S. Ma, Z.-H. Chen, W.-W. Zheng, H.-F. Hu, Processing and characterization of

three-dimensional carbon fiber-reinforced Si–O–C composites via precursor

pyrolysis, Mater. Sci. Eng. A. 352 (2003) 212–216.

[86] N. Nemoto, Y. Watanabe, J. Umemiya, Y. Otomo, Synthesis and Characterization

of Poly(tetramethyl-1,4-silphenylenesiloxane) Derivatives with Oxyethylene

Substituent on Phenylene Moiety, Polym. Bull. 57 (2006) 661–670.

[87] T. Gumula, C. Paluszkiewicz, S. Blazewicz, Study on thermal decomposition

processes of polysiloxane polymers—From polymer to nanosized silicon carbide,

J. Anal. Appl. Pyrolysis. 86 (2009) 375–380.

[88] G.D. Soraru, F. Babonneau, J.D. Mackenzie, Structural evolutions from

102

polycarbosilane to SiC ceramic, J. Mater. Sci. 25 (1990) 3886–3893.

[89] H.Q. Ly, R. Taylor, R.J. Day, F. Heatley, Conversion of polycarbosilane (PCS) to

SiC-based ceramic Part 1. Characterisation of PCS and curing products, J. Mater.

Sci. 36 (2001) 4037–4043.

[90] Q.-T. Pham, C.-S. Chern, Thermal stability of organofunctional polysiloxanes,

Thermochim. Acta. 565 (2013) 114–123.

[91] T. Gumula, C. Paluszkiewicz, M. Blazewicz, Structural characterization of

polysiloxane-derived phases produced during heat treatment, J. Mol. Struct. 704

(2004) 259–262.

[92] H. Li, L. Zhang, L. Cheng, Y. Wang, Z. Yu, M. Huang, H. Tu, H. Xia, Effect of

the polycarbosilane structure on its final ceramic yield, J. Eur. Ceram. Soc. 28

(2008) 887–891.

[93] R. Riedel, M. Seher, J. Mayer, D.V. Szabó, Polymer-derived Si-based bulk

ceramics, part I: Preparation, processing and properties, J. Eur. Ceram. Soc. 15

(1995) 703–715.

[94] M. Monthioux, O. Delverdier, Thermal behavior of (organosilicon) polymer-

derived ceramics. V: Main facts and trends, J. Eur. Ceram. Soc. 16 (1996) 721–

737.

[95] M.A. Schiavon, E. Radovanovic, I.V.P. Yoshida, Microstructural characterisation

of monolithic ceramic matrix composites from polysiloxane and SiC powder,

Powder Technol. 123 (2002) 232–241.

[96] L. Duan, Q. Ma, Effect of pyrolysis temperature on the pore structure evolution of

polysiloxane-derived ceramics, Ceram. Int. 38 (2012) 2667–2671.

103

[97] E. Müller, A. Hilbig, R. Wenzel, K. Trommer, G. Roewer, O. Sciurova, J.R.

Niklas, Amorphous SiC:H- Layers from Precursors, Adv. Eng. Mater. 4 (2002)

880–883.

[98] M. Černý, P. Glogar, Z. Sucharda, V. MacHovič, Properties and performance of

polysiloxane-derived ceramic matrix in heat resistant composites reinforced with

R-glass or fine ceramic fibres, Ceram. - Silikaty. 49 (2005) 145–152.

[99] X. Monnier, J.A.S. Puente, N. Basson, A. Saiter, E. Dargent, P.E. Mallon,

Investigation of the Amorphous Domain Cooperativity in Complex Copolymer

Nanofibers By Temperature Modulated Dsc, 34 (2014) 5539.

[100] Q.-S. Ma, Z.-H. Chen, W.-W. Zheng, H.-F. Hu, Effects of pyrolysis processes on

microstructure and mechanical properties of Cf/SiOC composites fabricated by

preceramic polymer pyrolysis, Ceram. Int. 31 (2005) 305–314.

[101] Y.-J. Jin, Y.-W. Kim, Low temperature processing of highly porous silicon carbide

ceramics with improved flexural strength, J. Mater. Sci. 45 (2010) 282–285.

[102] A.M. Wilson, G. Zank, K. Eguchi, W. Xing, B. Yates, J.R. Dahn, Polysiloxane

Pyrolysis, Chem. Mater. 9 (1997) 1601–1606.

[103] R.E. Myers, High char yield silazane derived preceramic polymers and cured

compositions thereof, (1993).

[104] G.M. Bayley, P.E. Mallon, Porous microfibers by the electrospinning of

amphiphilic graft copolymer solutions with multi-walled carbon nanotubes,

Polymer (Guildf). 53 (2012) 5523–5539.

[105] B. V Manoj Kumar, K.-Y. Lim, Y.-W. Kim, Influence of submicron SiC particle

addition on porosity and flexural strength of porous self-bonded silicon carbide,

104

Met. Mater. Int. 17 (2011) 435–440.

[106] B.M. Eick, J.P. Youngblood, SiC nanofibers by pyrolysis of electrospun

preceramic polymers, J. Mater. Sci. 44 (2009) 160–165.

[107] D.-G. Shin, D.-H. Riu, H.-E. Kim, Web-type silicon carbide fibers prepared by the

electrospinning of polycarbosilanes, J. Ceram. Process. Res. 9 (2008) 209–214.

[108] O.S. Yördem, M. Papila, Y.Z. Menceloğlu, Effects of electrospinning parameters

on polyacrylonitrile nanofiber diameter: An investigation by response surface

methodology, Mater. Des. 29 (2008) 34–44.

[109] D. Yang, X. Liu, Y. Jin, Y. Zhu, D. Zeng, X. Jiang, H. Ma, Electrospinning of

Poly(dimethylsiloxane)/Poly(methyl methacrylate) Nanofibrous Membrane:

Fabrication and Application in Protein Microarrays, Biomacromolecules. 10

(2009) 3335–3340.

[110] T. Uyar, P. Kingshott, F. Besenbacher, Electrospinning of Cyclodextrin–

Pseudopolyrotaxane Nanofibers, Angew. Chemie Int. Ed. 47 (2008) 9108–9111.

[111] M. Kancheva, A. Toncheva, N. Manolova, I. Rashkov, Enhancing the mechanical

properties of electrospun polyester mats by heat treatment, Express Polym. Lett. 9

(2015) 49–65.

[112] R. Rošic, J. Pelipenko, P. Kocbek, S. Baumgartner, M. Bešter-Rogač, J. Kristl, The

role of rheology of polymer solutions in predicting nanofiber formation by

electrospinning, Eur. Polym. J. 48 (2012) 1374–1384.

[113] Y.-T. Peng, C.-T. Lo, Effect of Microstructure and Morphology of Electrospun

Ultra-Small Carbon Nanofibers on Anode Performances for Lithium Ion Batteries,

J. Electrochem. Soc. 162 (2015) A1085–A1093.

105

[114] R. Nayak, R. Padhye, I.L. Kyratzis, Y.B. Truong, L. Arnold, Effect of viscosity

and electrical conductivity on the morphology and fiber diameter in melt

electrospinning of polypropylene, Text. Res. J. 83 (2012) 606–617.

[115] Y. Li, L. Fernandez-Recio, P. Gerstel, V. Srot, P.A. van Aken, G. Kaiser, M.

Burghard, J. Bill, Chemical Modification of Single-Walled Carbon Nanotubes for

the Reinforcement of Precursor-Derived Ceramics, Chem. Mater. 20 (2008) 5593–

5599.

[116] H.B. Park, C.H. Jung, Y.K. Kim, S.Y. Nam, S.Y. Lee, Y.M. Lee, Pyrolytic carbon

membranes containing silica derived from poly(imide siloxane): the effect of

siloxane chain length on gas transport behavior and a study on the separation of

mixed gases, J. Memb. Sci. 235 (2004) 87–98.

[117] S.N. Arshad, high strength carbon nanofibers derived from electrospun

polyacrylonitrile, University of Illinois at Urbana-Champaign, 2010.

[118] T. Liang, Y.-L. Li, D. Su, H.-B. Du, Silicon oxycarbide ceramics with reduced

carbon by pyrolysis of polysiloxanes in water vapor, J. Eur. Ceram. Soc. 30 (2010)

2677–2682.

[119] C. Vix-Guterl, I. Alix, P. Gibot, P. Ehrburger, Formation of tubular silicon carbide

from a carbon–silica material by using a reactive replica technique: infra-red

characterisation, Appl. Surf. Sci. 210 (2003) 329–337.

[120] Y. Li, C. Chen, J.-T. Li, Y. Yang, Z.-M. Lin, Surface charges and optical

characteristic of colloidal cubic SiC nanocrystals, Nanoscale Res. Lett. 6 (2011)

454.

[121] J.Y. Fan, H.X. Li, W.N. Cui, Microstructure and infrared spectral properties of

106

porous polycrystalline and nanocrystalline cubic silicon carbide, Appl. Phys. Lett.

95 (2009) 21906.

[122] E. Bouillon, F. Langlais, R. Pailler, R. Naslain, F. Cruege, P. V Huong, J.C.

Sarthou, A. Delpuech, C. Laffon, P. Lagarde, M. Monthioux, A. Oberlin,

Conversion mechanisms of a polycarbosilane precursor into an SiC-based ceramic

material, J. Mater. Sci. 26 (1991) 1333–1345.

[123] J. Mukherjee, A. Ranjan, A.K. Saxena, S. Karan, D.K. Dutta Majumder, A. Ghosh,

S. Ghosh, P.K. Das, R. Banerjee, A novel carbon rich crystalline (C) SiC–Si(n)

interface using liquid polycarbosilane as a precursor – a unique Schottky junction,

J. Mater. Chem. C. 1 (2013) 6945–6951.

[124] P. Lu, Q. Huang, A. Mukherjee, Y.-L. Hsieh, Effects of polymer matrices to the

formation of silicon carbide (SiC) nanoporous fibers and nanowires under

carbothermal reduction, J. Mater. Chem. 21 (2011) 1005–1012.

[125] S. Nakashima, H. Harima, Raman Investigation of SiC Polytypes, Phys. Status

Solidi. 162 (1997) 39–64.

[126] H. Okumura, E. Sakuma, J.H. Lee, H. Mukaida, S. Misawa, K. Endo, S. Yoshida,

Raman scattering of SiC: Application to the identification of heteroepitaxy of SiC

polytypes, J. Appl. Phys. 61 (1987) 1134–1136.

[127] T. Tomita, S. Saito, M. Baba, M. Hundhausen, T. Suemoto, S. Nakashima,

Selective resonance effect of the folded longitudinal phonon modes in the Raman

spectra of SiC, Phys. Rev. B. 62 (2000) 12896–12901.

[128] R. Prakash, S. Amirthapandian, D.M. Phase, S.K. Deshpande, R. Kesavamoorthy,

K.G.M. Nair, Study of ion beam induced mixing in nano-layered Si/C multilayer

107

structures, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with

Mater. Atoms. 244 (2006) 283–288.

[129] M. Montejo, F.P. Ureña, F. Márquez, A.M. Ingraín, J.J.L. González, Raman and

IR spectra of the unstable ionic species potassium trimethylsilanolate: The role of

the counterion in its theoretical interpretation, J. Raman Spectrosc. 39 (2008) 460–

467.

[130] Y. Yang, F. Simeon, T.A. Hatton, G.C. Rutledge, Polyacrylonitrile-based

electrospun carbon paper for electrode applications, J. Appl. Polym. Sci. 124

(2011) 3861–3870.

[131] Y.-T. Peng, C.-T. Lo, Effect of microstructure and morphology of electrospun

ultra-small carbon nanofibers on anode performances for lithium ion batteries, J.

Electrochem. Soc. 162 (2015) A1085–A1093.

[132] L. Ji, K.H. Jung, A.J. Medford, X. Zhang, Electrospun polyacrylonitrile fibers with

dispersed Si nanoparticles and their electrochemical behaviors after carbonization,

J. Mater. Chem. (2009).

[133] J. Roy, S. Chandra, S. Das, S. Maitra, Oxidation behaviour of silicon carbide-a

review., Rev. Adv. Mater. Sci. 38 (2014).

[134] X. Hou, K. Chou, Model of oxidation of SiC microparticles at high temperature,

Corros. Sci. 50 (2008) 2367–2371.

[135] J. Quanli, Z. Haijun, L. Suping, J. Xiaolin, Effect of particle size on oxidation of

silicon carbide powders, Ceram. Int. 33 (2007) 309–313.

[136] C. Balzer, M. Armstrong, B. Shan, Y. Huang, J. Liu, B. Mu, Modeling

Nanoparticle Dispersion in Electrospun Nanofibers, Langmuir. 34 (2018) 1340–

108

1346.

[137] C. Wyatt Shields IV, C.D. Reyes, G.P. López, Microfluidic cell sorting: a review

of the advances in the separation of cells from debulking to rare cell isolation, Lab

Chip. 15 (2015) 1230–1249.

[138] A. Chatterjee, Size-Dependant Separation of Multiple Particles in Spiral

Microchannels, University of Cincinnati, 2011.

[139] L. Yang, P.P. Banada, M.R. Chatni, K. Seop Lim, A.K. Bhunia, M. Ladisch, R.

Bashir, A multifunctional micro-fluidic system for dielectrophoretic concentration

coupled with immuno-capture of low numbers of Listeria monocytogenes, Lab

Chip. 6 (2006) 896–905.

[140] N. Pamme, A. Manz, On-Chip Free-Flow Magnetophoresis: Continuous Flow

Separation of Magnetic Particles and Agglomerates, Anal. Chem. 76 (2004) 7250–

7256.

[141] M.P. MacDonald, G.C. Spalding, K. Dholakia, Microfluidic sorting in an optical

lattice, Nature. 426 (2003) 421–424.

[142] K. Ladavac, K. Kasza, D.G. Grier, Sorting mesoscopic objects with periodic

potential landscapes: Optical fractionation, Phys. Rev. E. 70 (2004) 10901.

[143] T. Deng, M. Prentiss, G.M. Whitesides, Fabrication of magnetic microfiltration

systems using soft lithography, Appl. Phys. Lett. 80 (2002) 461–463.

[144] C.W. Yung, J. Fiering, A.J. Mueller, D.E. Ingber, Micromagnetic–microfluidic

blood cleansing device, Lab Chip. 9 (2009) 1171–1177.

[145] N. Pamme, Continuous flow separations in microfluidic devices, Lab Chip. 7

(2007) 1644–1659.

109

[146] X.-Y. Sun, R. Shankar, H.G. Börner, T.K. Ghosh, R.J. Spontak, Field-Driven

Biofunctionalization of Polymer Fiber Surfaces during Electrospinning, Adv.

Mater. 19 (2007) 87–91.

[147] C.A. Bonino, K. Efimenko, S.I. Jeong, M.D. Krebs, E. Alsberg, S.A. Khan, Three-

Dimensional Electrospun Alginate Nanofiber Mats via Tailored Charge

Repulsions, Small. 8 (2012) 1928–1936.

[148] J. Kanbar, Hydrodynamic Focused Passive Separation under Continuous Flow in a

Microfluidic Chip, (2012).

[149] H. Tsutsui, C.-M. Ho, Cell separation by non-inertial force fields in microfluidic

systems, Mech. Res. Commun. 36 (2009) 92–103.

[150] J.R. Szczech, S. Jin, Nanostructured silicon for high capacity lithium battery

anodes, Energy Environ. Sci. 4 (2011) 56–72.

[151] J.-S. Park, H.-I. Jung, Multiorifice Flow Fractionation: Continuous Size-Based

Separation of Microspheres Using a Series of Contraction/Expansion

Microchannels, Anal. Chem. 81 (2009) 8280–8288.

[152] J. Bec, Multifractal concentrations of inertial particles in smooth random flows, J.

Fluid Mech. 528 (2005) 255–277.

[153] D. Crespy, K. Friedemann, A.-M. Popa, Colloid-Electrospinning: Fabrication of

Multicompartment Nanofibers by the Electrospinning of Organic or/and Inorganic

Dispersions and Emulsions, Macromol. Rapid Commun. 33 (2012) 1978–1995.

[154] J. Yao, W.C. Bastiaansen, T. Peijs, High Strength and High Modulus Electrospun

Nanofibers, Fibers . 2 (2014).

[155] C. Huang, S. Chen, C. Lai, D.H. Reneker, H. Qiu, Y. Ye, H. Hou, Electrospun

110

polymer nanofibres with small diameters, Nanotechnology. 17 (2006) 1558–1563.

[156] V. Beachley, X. Wen, Effect of electrospinning parameters on the nanofiber

diameter and length, Mater. Sci. Eng. C. 29 (2009) 663–668.

[157] H.J. Kim, I.C. Um, Relationship between rheology and electro-spinning

performance of regenerated silk fibroin prepared using different degumming

methods, Korea-Australia Rheol. J. 26 (2014) 119–125.

[158] C. Mit-uppatham, M. Nithitanakul, P. Supaphol, Ultrafine Electrospun Polyamide-

6 Fibers: Effect of Solution Conditions on Morphology and Average Fiber

Diameter, Macromol. Chem. Phys. 205 (2004) 2327–2338.

[159] I. Microfluids Moduls, Version 5, COMSOL

[160] L.M. Bellan, H.G. Craighead, J.P. Hinestroza, Direct measurement of fluid

velocity in an electrospinning jet using particle image velocimetry, J. Appl. Phys.

102 (2007) 94308.

[161] A. Menon, K. Somasekharan, Velocity, acceleration & jerk in electrospinning,

Internet J. Bioeng. 4 (2008).

[162] Particle Tracing Module, Version 5, COMSOL, Inc., (n.d.).

[163] P. O’Malley, Trajectories of Inertial Particles and Fluid Elements, Haverford Coll.

Haverford. (2008).

[164] R. Nain, D. Singh, M. Jassal, A.K. Agrawal, Zinc oxide nanorod assisted rapid

single-step process for the conversion of electrospun poly(acrylonitrile) nanofibers

to carbon nanofibers with a high graphitic content, Nanoscale. 8 (2016) 4360–

4372.

[165] T.E. Newsome, S. V Olesik, Electrospinning silica/polyvinylpyrrolidone

111

composite nanofibers, J. Appl. Polym. Sci. 131 (2014).

[166] M.C. dos Santos, M.C. Maynart, L.R. Aveiro, E.C. da Paz, V.B.T.-R.M. in M.S.

and M.E. dos Santos Pinheiro, Carbon-Based Materials: Recent Advances,

Challenges, and Perspectives, in: Elsevier, 2017.

[167] B. Ilkiv, S. Petrovska, R. Sergiienko, O. Foya, O. Ilkiv, E. Shibata, T. Nakamura,

Y. Zaulychnyy, Electronic Structure of Hollow Graphitic Carbon Nanoparticles

Fabricated from Acetylene Carbon Black, Fullerenes, Nanotub. Carbon

Nanostructures. 23 (2015) 449–454.

[168] R. Gadiou, S. Serverin, P. Gibot, C. Vix-Guterl, The synthesis of SiC and TiC

protective coatings for carbon fibers by the reactive replica process, J. Eur. Ceram.

Soc. 28 (2008) 2265–2274.

[169] X. Wu, L.R. Radovic, Inhibition of catalytic oxidation of carbon/carbon

composites by boron-doping, Carbon N. Y. 43 (2005) 1768–1777.

doi:https://doi.org/10.1016/j.carbon.2005.02.029.

[170] A.K. Roy, S. Schulze, M. Hietschold, W.A. Goedel, Oxidation protection of

carbon fibers by coating with alumina and/or titania using atomic layer deposition,

Carbon N. Y. 50 (2012) 761–770.

[171] G. Hackl, H. Gerhard, N. Popovska, Coating of carbon short fibers with thin

ceramic layers by chemical vapor deposition, Thin Solid Films. 513 (2006) 217–

222.

[172] Y.-J. Lee, H.-J. Joo, L.R. Radovic, Preferential distribution and oxidation

inhibiting/catalytic effects of boron in carbon fiber reinforced carbon (CFRC)

composites, Carbon N. Y. 41 (2003) 2591–2600.

112

[173] F. Kern, R. Gadow, Deposition of ceramic layers on carbon fibers by continuous

liquid phase coating, Surf. Coatings Technol. 180–181 (2004) 533–537.

[174] B.-H. Kim, C.H. Kim, K.S. Yang, K.-Y. Kim, Y.-J. Lee, SiC/SiO2 coating for

improving the oxidation resistive property of carbon nanofiber, Appl. Surf. Sci.

257 (2010) 1607–1611.

[175] Y.-Q. Wang, B.-L. Zhou, Z.-M. Wang, Oxidation protection of carbon fibers by

coatings, Carbon N. Y. 33 (1995) 427–433.

[176] N. Popovska, S. Schmidt, E. Edelmann, V. Wunder K., H. Gerhard, G. Emig,

Coating of continuous carbon fibers with double layers by chemical vapor

deposition, J. Phys. IV Fr. 11 (2001) Pr3-885-Pr3-892.

[177] P. Bhattarai, K. Thapa, R. Basnet, S. Sharma, Electrospinning: How to Produce

Nanofibers Using Most Inexpensive Technique? An Insight into the Real

Challenges of Electrospinning Such Nanofibers and Its Application Areas, Int. J.

Biomed. Adv. Res. 5 (2014).

[178] C.M. Brewer, D.R. Bujalski, V.E. Parent, K. Su, G.A. Zank, Insights into the

Oxidation Chemistry of SiOC Ceramics Derived from Silsesquioxanes, J. Sol-Gel

Sci. Technol. 14 (1999) 49–68.

[179] H.-C. Wu, A.N. Chaika, M.-C. Hsu, T.-W. Huang, M. Abid, M. Abid, V.Y.

Aristov, O. V Molodtsova, S. V Babenkov, Y. Niu, B.E. Murphy, S.A. Krasnikov,

O. Lübben, H. Liu, B.S. Chun, Y.T. Janabi, S.N. Molotkov, I. V Shvets, A.I.

Lichtenstein, M.I. Katsnelson, C.-R. Chang, Large positive in-plane

magnetoresistance induced by localized states at nanodomain boundaries in

graphene, Nat. Commun. 8 (2017) 14453.

113

[180] C.M.L.M.H. Kapadia, The mechanism of grain growth in ceramics, 1973.

[181] T. Lehmann, J. Baier, A. Leineweber, A. Kienzle, J. Bill, Influence of the Carbon

Content on the Crystallization and Oxidation Behavior of Polymer-Derived Silicon

Carbide (SiC), Adv. Eng. Mater. 17 (2015) 1631–1638.

[182] C. Kim, S.-H. Park, J.-I. Cho, D.-Y. Lee, T.-J. Park, W.-J. Lee, K.-S. Yang, Raman

spectroscopic evaluation of polyacrylonitrile-based carbon nanofibers prepared by

electrospinning, J. Raman Spectrosc. 35 (2004) 928–933.

[183] M.B. Vázquez-Santos, E. Geissler, K. László, J.-N. Rouzaud, A. Martínez-Alonso,

J.M.D. Tascón, Comparative XRD, Raman, and TEM Study on Graphitization of

PBO-Derived Carbon Fibers, J. Phys. Chem. C. 116 (2012) 257–268.

[184] R. Ramachandran, M. Saranya, P. Kollu, B.P.C. Raghupathy, S.K. Jeong, A.N.

Grace, Solvothermal synthesis of Zinc sulfide decorated Graphene (ZnS/G)

nanocomposites for novel Supercapacitor electrodes, Electrochim. Acta. 178

(2015) 647–657.

[185] J. Liu, Q. Li, Y. Zou, Q. Qian, Y. Jin, G. Li, K. Jiang, S. Fan, The Dependence of

Graphene Raman D-band on Carrier Density, Nano Lett. 13 (2013) 6170–6175.

[186] C. Cheng, H. Li, Q. Fu, L. Li, L. Guo, X. Yin, X. Tian, Effects of pyrocarbon on

morphology stability of SiC nanowires at high temperatures, J. Am. Ceram. Soc.

101 (2018) 3694–3702. doi:10.1111/jace.15524.

[187] K. Share, R.E. Carter, P. Nikolaev, D. Hooper, L. Oakes, A.P. Cohn, R. Rao, A.A.

Puretzky, D.B. Geohegan, B. Maruyama, C.L. Pint, Nanoscale Silicon as a

Catalyst for Graphene Growth: Mechanistic Insight from in Situ Raman

Spectroscopy, J. Phys. Chem. C. 120 (2016) 14180–14186.

114

[188] V.Y. Aristov, G. Urbanik, K. Kummer, D. V Vyalikh, O. V Molodtsova, A.B.

Preobrajenski, A.A. Zakharov, C. Hess, T. Hänke, B. Büchner, I. Vobornik, J.

Fujii, G. Panaccione, Y.A. Ossipyan, M. Knupfer, Graphene Synthesis on Cubic

SiC/Si Wafers. Perspectives for Mass Production of Graphene-Based Electronic

Devices, Nano Lett. 10 (2010) 992–995.

[189] A.N. Chaika, V.Y. Aristov, O. V Molodtsova, Graphene on cubic-SiC, Prog.

Mater. Sci. 89 (2017) 1–30.

[190] S.M.N. Al-Ajrash, K. Lafdi, E.S. Vasquez, F. Chinesta, P. Le Coustumer,

Experimental and Numerical Investigation of the Silicon Particle Distribution in

Electrospun Nanofibers, Langmuir. 34 (2018) 7147–7152.

[191] R. Cheung, Silicon Carbide Microelectromechanical Systems for Harsh

Environments, (2006) 192.

[192] S.Y. Kim, B.-H. Kim, K.S. Yang, K.-Y. Kim, The formation of silica nanoparticles

on the polyacrylonitrile-based carbon nanofibers by graphene via electrospinning,

Mater. Lett. 71 (2012) 74–77.

[193] B.S. Girgis, Y.M. Temerk, M.M. Gadelrab, I.D. Abdullah, X-ray Diffraction

Patterns of Activated Carbons Prepared under Various Conditions, Carbon Sci. 8

(2007) 95–100.

[194] Y. Yue, Synthesis of Silicon Carbide Fibers From Polycarbosilane By

Electrospinning Method, 2014.

[195] . N.S.B., . S.B.T., . B.N.R., Structural and Morphological Features of Graphitised

Carbon Fibres, Trends Appl. Sci. Res. 1 (2006) 204–213.

[196] H. V Atkinson, Overview no. 65: Theories of normal grain growth in pure single

115

phase systems, Acta Metall. 36 (1988) 469–491.

[197] R.J. BROOK, Controlled Grain Growth, in: F.F.Y.B.T.-T. on M.S.& T. Wang

(Ed.), Ceram. Fabr. Process., Elsevier, 1976: pp. 331–364.

[198] I.M. De Rosa, F. Sarasini, M.S. Sarto, A. Tamburrano, EMC Impact of Advanced

Carbon Fiber/Carbon Nanotube Reinforced Composites for Next-Generation

Aerospace Applications, IEEE Trans. Electromagn. Compat. 50 (2008) 556–563.

[199] H.A. Al-Qureshi, Automobile leaf springs from composite materials, J. Mater.

Process. Technol. 118 (2001) 58–61.

[200] M. Tarfaoui, A. El Moumen, K. Lafdi, Progressive damage modeling in carbon

fibers/carbon nanotubes reinforced polymer composites, Compos. Part B Eng. 112

(2017) 185–195.

[201] N. Tual, N. Carrere, P. Davies, T. Bonnemains, E. Lolive, Characterization of sea

water ageing effects on mechanical properties of carbon/epoxy composites for tidal

turbine blades, Compos. Part A Appl. Sci. Manuf. 78 (2015) 380–389.

[202] A.H. Ali, H.M. Mohamed, B. Benmokrane, A. ElSafty, Theory-based approaches

and microstructural analysis to evaluate the service life-retention of stressed

carbon fiber composite strands for concrete bridge applications, Compos. Part B

Eng. 165 (2019) 279–292.

[203] M. Pervaiz, M.M. Sain, Carbon storage potential in natural fiber composites,

Resour. Conserv. Recycl. 39 (2003) 325–340.

[204] C. Luan, X. Yao, C. Liu, L. Lan, J. Fu, Self-monitoring continuous carbon fiber

reinforced thermoplastic based on dual-material three-dimensional printing

integration process, Carbon N. Y. 140 (2018) 100–111.

116

[205] G. Chikvaidze, N. Mironova-Ulmane, A. Plaude, O. Sergeev, Investigation of

silicon carbide polytypes by Raman spectroscopy, Latv. J. Phys. Tech. Sci. 51

(2014) 51–57.

[206] A. Kaniyoor, S. Ramaprabhu, A Raman spectroscopic investigation of graphite

oxide derived graphene, AIP Adv. 2 (2012) 32183.

[207] J. Prakash, K. Dasgupta, B. Kumar, S.K. Ghosh, J.K. Chakravartty, Role of SiC

nanowire coating on oxidation behavior of carbon fibers: Kinetic and

thermodynamic study, Surf. Coatings Technol. 259 (2014) 637–646.

[208] W. Zhou, L. Long, P. Xiao, Y. Li, H. Luo, W. Hu, R. Yin, Silicon carbide nano-

fibers in-situ grown on carbon fibers for enhanced microwave absorption

properties, Ceram. Int. 43 (2017) 5628–5634.

117