Structural and Physical Effects of Carbon Nanofillers in Thermoplastic and Thermosetting Polymer Systems

This thesis is based on the following papers:

I Improvement of toughness and electrical properties of

epoxy composites with carbon nanotubes prepared by industrially relevant processes

R Hollertz, S Chatterjee, H Gutmann, T Geiger, F A Nüesch and B T T Chu

Nanotechnology, 22,125702 (2011) [1]

II Comparing carbon nanotubes and graphene nanoplatelets as reinforcements in polyamide-12 composites

S Chatterjee, F A Nüesch and B T T Chu Nanotechnology, 22,275714 (2011) [2]

III Investigation of crystalline and tensile properties of carbon nanotube- filled polyamide-12 fibers melt-spun by industrially-related processes S Chatterjee, F A Reifler, B T T Chu and R Hufenus

Journal for Engineered Fibers and Fabrics (accepted for publication) [3]

IV Mecahnical reinforcement and thermal conductivity in expanded graphene nanoplates reinforced epoxy composites

S Chatterjee, J W Wang, W S Kuo, N H Tai, C Salzmann, W L Li, R Hollertz, F A Nüesch and B T T Chu

Chemical Physics Letters 531, 6 (2012) [4]

V The size and synergy effects of graphene nanoplatelets in mechanical properties of epoxy composites

S Chatterjee, F Nafezarefi, N H Tai, L Schlagenhauf, F A Nüesch and B T T Chu

under preparation [5]

Reprints were made with permission from the publishers.

Preface

This doctoral thesis presents a study carried out to understand the rein- forcing effects of carbon nanofillers on the structure and physical prop- erties of thermoplastic and thermoset polymer matrices. All the exper- iments have been performed during my tenure as a Ph.D. student be- tween 2008 - 2012 at the Swiss Federal Laboratories for Materials Science and Technology (Empa), Switzerland and the Division of Molecular and Condensed Matter Physics at the Department of Physics and Astron- omy, Uppsala University, Sweden. Most of the activities were carried out by me with the help and support of my colleagues at three departments of Empa namely, Functional Polymers, Advanced Fibers and Protection and Physiology. In addition some experiments were done at the Materi- als Science Department of Swiss Federal Institute of Technology (ETH), Zürich. Some of the important measurements were carried out at the synchrotrons MAX-lab, Sweden and Swiss Light Source, Switzerland.

The last four years have been a overwhelming journey for me as I have

worked with many talented scientists and have enriched my knowledge

with their support. In the following pages I put together the interesting

findings which have made my doctoral studies so rewarding. I take this

opportunity to thank you for your interest in my research work.

Comments on my own participation

The project has been carried out based on the framework of the pro-

posal accepted by the Swiss National Science Foundation. I have per-

formed most of the experimental techniques described in the thesis with

adequate training and support from technical personnel. Among the five

publications mentioned in this thesis I have had the main responsibility

for planning and executing the experimental part and subsequent data

analysis as well as writing the manuscript for four of them. For the pub-

lication in which I am the second author I have carried out part of the

study and have assisted the first author in manuscript preparation. I

would be happy to provide further information regarding any aspect of

the research mentioned in the thesis.

List of Abbreviations

ASTM American Society for Testing and Materials

CNT Carbon nanotube

CVD Chemical vapour diposition

DR Draw ratio

DSC Differential scanning calorimetry ECI Equatorial percentage crystallinity EGNP Expanded graphene nanoplatelet

Ext Extrusion factor

FTIR Fourier transform infra red spectroscopy

GnP Graphene nanoplatelet

HPH High pressure homoginizer

FDTD Finite difference time domain

ISO International Organization for Standardization

MAS Magic angle spinning

MWCNT/MWNT Multi-walled carbon nanotube

NG Natural graphite

NMR Nuclear magnetic resonance

PA Polyamide

PA12 Polyamide-12

SENT Single edge-notched tension SEM Scanning electron microscope

Str Strain factor

SWCNT/SWNT Single-walled carbon nanotube TEM Transmission electron microscope

TGA Thermogravimetric analysis

WAXD Wide angle x-ray diffraction

Preface . . . . v

1 Introduction . . . . 1

1.1 Carbon nanofillers . . . . 2

1.2 Polymer nanocomposites . . . . 4

1.2.1 Properties of polymer nanocomposites . . . . 5

1.3 Motivation . . . . 7

2 Materials . . . . 11

2.1 Synthesis . . . . 11

2.1.1 Carbon nanotube . . . . 11

2.1.2 Graphene . . . . 13

2.2 Properties . . . . 13

2.3 Polymer composites of CNT and graphene . . . . 14

2.3.1 Dispersion . . . . 14

2.3.2 Methods for manufacturing polymer nanofiller com- posites . . . . 16

2.4 Thermoplastic polymer: Polyamide . . . . 17

2.4.1 Polyamide nanocomposites . . . . 18

2.5 Thermoset polymer: Epoxy . . . . 20

2.5.1 Epoxy nanocomposites . . . . 21

2.6 Specifications of materials used . . . . 22

3 Experimental details . . . . 25

3.1 Processing of polyamide-12 . . . . 25

3.2 Processing of epoxy . . . . 28

3.3 Characterization . . . . 28

4 Polyamide Composites . . . . 35

4.1 Dispersion of nanofillers . . . . 35

4.2 Nuclear magnetic resonance (NMR) . . . . 37

4.3 Wide angle x-ray diffraction (WAXD) . . . . 37

4.4 Thermal analysis . . . . 43

4.4.1 Differential scanning calorimetry (DSC) . . . . 43

4.4.2 Thermogravimetric analysis (TGA) . . . . 45

4.5 Mechanical properties . . . . 45

4.6 Electrical conductivity . . . . 50

5 Epoxy Composites . . . . 57

5.1 Dispersion . . . . 57

5.2 Mechanical properties . . . . 61

5.3 Electrical conductivity . . . . 65

5.4 Thermal conductivity . . . . 66

6 Conclusions and Outlook . . . . 69

Summary . . . . 75

Populärvetenskaplig sammanfattning . . . . 79

7 Bibliography . . . . 83

Acknowledgement . . . . 95

List of Figures

1.1 Graphene as building block . . . . 4

1.2 Composite application in automobile . . . . 6

2.1 TEM image of SWNT . . . . 12

2.2 TEM image of graphene. . . . 14

2.3 Surface modifications of carbon nanotubes. . . . 16

2.4 Structure of PA12. . . . 19

2.5 SEM images of PA12 composites. . . . 20

2.6 Structure of Epoxy resin. . . . 21

2.7 SEM images of nanofillers used . . . . 23

3.1 Micro-extruder . . . . 26

3.2 In-house spinning set-up . . . . 27

3.3 Processing steps for epoxy composites . . . . 29

4.1 SEM images showing agglomeration sites within the com- posites. . . . 36

4.2 SEM images showing fractured surfaces of the composites. 36 4.3 WAXD images of films . . . . 38

4.4 Deconvolution of WAXD peak for film sample . . . . 39

4.5 WAXD images of fibers . . . . 41

4.6 Deconvolution of WAXD peak for fiber . . . . 41

4.7 Herman’s orinetation factor of fibers . . . . 43

4.8 DSC curves . . . . 44

4.9 TGA curves . . . . 45

4.10 Modulus of toughness . . . . 47

4.11 specific tensile strength in relation to the strain factor . . . 50

4.12 Electrical conductivity of films . . . . 51

4.13 Electrical percolation in films . . . . 54

5.1 TEM images of CNT dispersion . . . . 58

5.2 Matlab simulation of CNT dispersion . . . . 59

5.3 STEM images of CNT network in matrix . . . . 59

5.4 TEM images of amine-EGNP dispersion . . . . 60

5.5 TEM images of CNT and GnP mixture dispersion . . . . 60

5.6 Mechanical properties of CNT composites . . . . 61

5.7 Mechanical properties of EGNP composites . . . . 63

5.8 Meachnical properties of GnP composites . . . . 64

5.9 Mechanical properties of CNT/GnP mixture composites . 65 5.10 Electrical conductivity of CNT composites . . . . 66

5.11 Thermal conductivity of EGNP and GnP composites . . . 67

List of Tables

2.1 A summary of the most common methods of graphene synthesis. . . . 14 2.2 A summary of properties of CNT and graphene. . . . 15 4.1 Chemical shifts (in ppm) of

C in pure PA12 samples. . . 37 4.2 Equatorial crystallinity index (% crystallinity) for a set of

PA12 films with 0.5 wt% CNT and GnP loading. . . . 40 4.3 Crystallinity of pure and composite films from DSC analysis 44 4.4 Summary of extrusion factors, strain factors and tensile

strengths. . . . 49 4.5 Percolation threshold in PA12 composite samples for the

different nanofillers. . . . 53

1. Introduction

There is no higher or lower knowledge, but one only, flowing out of experimentation

- Leonardo da Vinci.

Eagerness for experimentation has always gifted mankind with novel materials and the ever curious mind has never rested to improvise on all that already exist. Reinforcement is "to make a structure or material stronger, especially by adding another material to it". It is the effects of "adding another material" that fascinate a large section of the sci- entific community today. The idea of using fillers as reinforcing agents is not new, it possibly started with the use of straw to reinforce mud bricks in about 4000 BCE. Reinforcement grids, plates or fibers have been incorporated to strengthen concrete. This method was invented in 1849 and even patented in 1867. In more recent times, fibres made from materials such as alumina, glass, boron, silicon carbide and most signifi- cantly carbon have been used as fillers in composites. These conventional fibres have dimensions on the meso-scale with diameters of tens of mi- crons and lengths of the order of millimetres limiting their capabilities of mechanical reinforcement.

Polymer composites with additives mixed in thermoplastic, thermoset

and elastomer matrices are considered as an important group of materials

for their wide variety of applications. In contrast to the incorporation

of micro-scale fillers, nanofillers, due to their size ensure a very small

inter-particle distance which influences the polymer matrix properties

even at very low filler concentrations. Polymer nanocomposites came into

limelight with the discovery, at Toyota research center in the 1980s, that

on addition of a small fraction of nanoclay to PA6 dramatic improvement

could be recorded in strength, modulus, heat distortion temperature and

gas barrier properties [6].

Carbon, a remarkable element is known for its extraordinary ability of catenation. A property to combine with itself and other chemical el- ements in various ways forms the basis of organic chemistry and thus life. Since the discovery of carbon based nanostructures starting with fullerene, their unique structural and transport properties have captured interest of researchers. Graphene, another nanostructure of carbon has been tested to be the strongest material on Earth and Carbon nanotubes (CNT) do not trail behind much in their strength. From a simple rule of mixtures to more complicated models it is clear that carbon nanofillers promise ultra-strong-conducting polymer composites. In my doctoral re- search I have explored various techniques to optimize the manufacturing processes of polymer nanocomposites followed by a thorough analysis of their structure and properties of the novel matrix. In this chapter, I shall emphasize on the significance of carbon nanofillers and polymer nanocomposites that led to my strong interest to pursue my doctoral research in this field.

1.1 Carbon nanofillers

Carbon in its manifold forms has been used in technical and artistic quarters from the prehistoric ages. In nature, it is abundantly avail- able as coal, natural graphite and in smaller quantities as diamonds.

Man has also engineered several other forms of carbon such as synthetic graphite, synthetic diamonds, adsorbent carbon, cokes, carbon black, carbon and graphitic fibers, glassy carbons, diamond-like carbon etc.

for applications in various fields such as electrodes and electrical con-

tacts, lubricants, shoe polish, gemstones, cutting wheels, gas adsorption,

catalytic support, helium gas barrier, tire and elastomer reinforcement,

toner for photocopying machines and printing inks, high performance

tennis rackets, aircraft and spacecraft composites, heat sinks for ultra-

fast semiconductor based devices etc. All these various forms of carbon

can be ascribed to carbon’s unique hybridization properties [7]. In mid

1980s carbon science gained acceleration when the discovery of a first

all-carbon molecule, fullerene C60 or buckyball by Harry Kroto of the

University of Sussex and Richard Smalley of Rice University generated a lot of attention [8].

In 1991, Sumio Iijima, an electron microscopist at NEC laboratories in Japan was credited with the discovery of carbon nanotubes [9]. Ijima used arc-discharge evaporation method similar to that used for the ear- lier fullerene synthesis, to grow needle-like tubes. These needles grew at the negative end of the electrode used for the arc discharge. Carbon nanotubes (CNTs) are tubular derivatives of fullerenes which consist of graphene cylinders generally closed at both ends with caps containing pentagonal rings. CNTs can be single, double or multi walled. CNTs typically have diameter of few nanometers and length can vary up to several centimeters [10]. Before Ijima’s path breaking discovery, there were research contributing towards the discovery of CNTs, for exam- ple, Radushkevich and Lukyanovich published a report on hollow carbon fibers in 1950s [11]. In 1970s, Oberlin et al. showed an image of empty tubular-shape materials formed by benzene decomposition [12], although it was difficult to determine the number of carbon walls with the resolu- tion available at that time. The electronic properties of tubular carbon nanostructures have been theoretically studied even before Ijima’s paper was published.

Graphite is one of the allotropes of carbon having a layered, planar structure. In each layer, the carbon atoms are arranged in a hexagonal lattice with separation of 0.142 nm, and the distance between planes is 0.335 nm. Graphene is one-atom-thick planar sheet of sp

-bonded carbon atoms that are densely packed in a honeycomb crystal lattice.

The name graphene was coined in 1987. Graphene is the basic build-

ing block of carbon allotropes like graphite, charcoal, carbon nanotubes

and fullerenes. There have been several efforts to make very thin films

of graphite by methods like mechanical exfoliation between 1990 and

2004 [13]. Graphene was formally discovered by Andre Geim and Kostya

Novoselov from Manchester University in 2004 [14] for which they were

awarded the Nobel prize in 2010. They managed to extract a single-

atom-thick crystallite layer from bulk graphite by pulling out graphene

layers and transferring them onto thin SiO

on a silicon wafer in a pro-

cess called micromechanical cleavage or the Scotch tape technique. The

theory of graphene has been explored for a long time since 1947 by Wal- lace [15] and other theoreticians. Different forms of carbon based nano particles are derived from the basic graphene structure. A schematic in Fig. 1.1 shows how fullerene and nanotube are perceived to be derived from graphene.

Figure 1.1: Graphene is 2-D building block for carbon materials of all other dimensionalities. It can be wrapped into 0-D buckyballs, rolled into 1-D nan- otubes or stacked into 3-D graphite [16]

1.2 Polymer nanocomposites

Nanocomposite is a class of multi-component material in which one

of the phases is dispersed in second one in nanometer range [17]. Con-

ventional fillers in polymers have often been used to reduce cost or im-

prove material properties required for particular applications. Ceramic

fillers such as silica or alumina are typically used to reduce cost and

increase the stiffness in epoxy resins [18, 19]. The drawback is however

that addition of such rigid particles further reduces the ductility in the

already brittle epoxy. Soft particles, such as rubber, on the other hand

can be used to toughen the epoxy matrix but they consequently reduce

the stiffness [18, 20]. The outstanding properties of carbon nanotubes,

graphene nanoplatelets (few layers of graphene) and graphene can be exploited by inclusion of such nanofillers into a matrix to form nanocom- posites. Several polymer matrices my be used along with these carbon nanofillers to synthesize interesting composites which is the primary fo- cus of this present work. There are several methods for manufacturing polymer nanocomposites but the ability to disperse nanofillers is a crit- ical factor for tuning their properties.

Such composites find a wide palette of potential applications ranging from bicycle frames, badminton rackets to micro/nano devices [21], nanopackaging [22], smart materials [23], sensors and actuators [24]

among many others. Bionanocomposites used for applications such as tissue engineering and load-bearing composites for bone reconstruction are of significant interest. The most important factor driving the use of nanocomposites for mechanical reinforcement in industrial scales is the reduction in weight to its mechanical performance ratio as compared to their metallic counterparts. Major applications of nanocomposites are found in aerospace industry as critical components in aircraft.

Recently launched Boeing 787 Dreamliner is manufactured with 80 vol%

composites, featuring an unique light weight construction with 50 % carbon and glass fiber reinforced materials. The average fuel savings of all CNT-reinforced aircrafts is about 10 % along with improved physical and structural properties. Fig. 1.2 shows the possible applications of carbon nanofiller based polymer composites in the automobile industry.

Another important application is in the form of conducting films and coatings for devices like displays, touch screens etc. Enhanced electrical conductivity coupled with mechanical strength, optical and thermal properties is the key for such new-age applications.

1.2.1 Properties of polymer nanocomposites

• Mechanical

The most significant improvement in the polymer matrix has been noted in the field of mechanical reinforcement. Addition of even 1% MWCNT to polystyrene results in an improvement of 25%

in breaking stress and 36-42% in tensile modulus. Inclusion of

Figure 1.2: Possible applications of carbon nanofiller/polymer composites in automobile industry

MWCNT in ultra-high molecular weight polyethylene resulted in an improvement of 150% and 105% in toughness and ductility re- spectively [25]. Several matrices have been used for composites and reinforcement appears to be critically dependent on the polymer- nanotube interfacial interaction. It has been shown in literature that the reinforcement scales linearly with the total nanotube sur- face area in the composites, indicating that small diameter mul- tiwall nanotubes are the best tube type for reinforcement [26].

Graphite nanoplatelets are also used as fillers and improvement in flexural modulus and other mechanical properties have been ob- served [7].

• Electrical

Percolation dominated electrical conductivity is observed in several polymer/CNT composites. From more than 30 matrices studied in literature, the average percolation threshold is around 0.1 wt%

CNT loading [27]. A conducting pathway is created in the ma-

trix formed by the network of the nanofillers. Percolation thresh-

olds significantly lower than 0.1 wt% concentration are attributed

to kinetic percolation which allows for particle movement and re-

aggregation. This interpretation particularly applies to results pub- lished for CNT/epoxy composites [27]. These improvements are highly subjected to various parameters like CNT type, synthesis method, treatment, dimensionality as well as the polymer type and compounding method used. With graphene, percolation has been achieved as low as 0.15 vol% concentration in polystyrene matrix followed by an increase in conductivity with graphene content [28].

• Thermal

CNTs possess the potential to improve thermal conductivity of composites by even up to 300% with addition of 3 wt% of SWC- NTs [25]. Increase in the glass transition temperature, melting and thermal decomposition temperatures of polymer matrix due to con- straint effects on the polymer chains have also been reported.

1.3 Motivation

A primary concern in understanding the potential of nanofillers to modify polymer matrices is achieving uniform nanofiller dispersion and identi- fying the set of failure mechanisms, which may ultimately lead to rele- vant modification (and reinforcement) mechanism(s). The most common ansatz is that a network of nanofillers is required to achieve conductivity and/or mechanical reinforcement. Publications report a large variation for the concentration at which percolation is achieved. Early values are in the range of several percent or more, but some of the most recent results are of the order of 0.001 wt% . Current research has yet to satis- factorily assess how parameters such as nanofiller variety, concentration, chemical modifications, composite processing affect the achieved mod- ification in structure and properties. Given the high potential of the carbon nanofillers our goal is to investigate their effects on polymer ma- trices both in thermoplastic and thermoset polymers namely PA12 and epoxy. The reinforcing influence on the mechanical, electrical, thermal properties is of prime interest.

For thermoplastic polymer, polyamide-12 we want to understand

the effects of CNT as nucleating agents for crystalline domains since

crystallinity significantly influences the properties. Since nanofiller dispersion is a key issue governing the properties of the composite, various methods are needed to be employed to manufacture samples ensuring good nanofiller dispersion in the polymer matrix and dispersion has been investigated over different length scales ensuring better understanding. For thermoset polymer, epoxy which is widely used for industrial applications we are optimising dispersion methods suitable to be scaled up. A comparison is made between the reinforcing capabilities of CNTs and GnPs as nanofillers. A hybrid mixture of CNT and GnP is being investigated for the possible synergistic effects observed in the composites. We study three different polymeric systems namely fibers, films and mould cast composites. An in-house pilot spinning plant, which is close to the industrial standards is able to produce very high quality composite fibers. We expect to achieve reinforced materials with improved physical properties. A wide variety of analytical techniques will be used in materials characterization. Our final objective is to find out the relationship between the structure and the properties of the pristine and composite polymer systems and hence design materials effective for cutting edge applications.

Outline of Thesis

This dissertation is arranged in six chapters. The present chapter, Introduction (Chapter 1) provides a general outline regarding carbon nanofillers and the development of the polymer nanocomposite field. The motivation guiding the current project is also discussed here.

Chapter 2: Materials introduces the nanofillers and the polymers that are used in this project. Methods of synthesis and properties of nanofillers, structure of polymers and properties of nanocomposites are described and the technical details of the materials utilized are mentioned.

Chapter 3: Experimental Details describes the processing techniques

used in synthesizing the polymer composites and the different analytical

methods employed in characterization.

The results of the study are analysed and discussed in Chapter 4:

Polyamide Composites and Chapter 5: Epoxy Composites. The influence of the nanofillers on the morphology and properties of the polymers and possible explanations to the behaviours of the composites are the focus of these chapters.

Finally, chapter 6, Conclusion and Outlook summarizes the results

and presents a comparative discussion on the reinforcing capabilities of

different nanofillers in the two matrices used. The chapter also provides

a general outlook on possible future work.

2. Materials

Exotic properties of materials that we use make them the basis of our investigations. Our main focus is on the properties of polymer nanocom- posites, a class of materials composed of nanoscale particles dispersed in a polymer matrix. Most of the raw materials used in this project are commercially available which makes the processes suitable for industrial scaling. In this chapter we discuss the synthesis and properties of the carbon nanofillers, the polymer matrices and nanocomposites that are relevant in our work.

2.1 Synthesis

2.1.1 Carbon nanotube

As the name suggests, CNTs are nano-scaled tubular structures with very high aspect ratio and can be used as 1-dimensional fillers. They can be single, double or multi-walled depending on their synthesis methods.

Three most important methods for CNT synthesis used both at labora- tory and industrial scale are discussed below

• Arc-evaporation

Arc-evaporation method, which produces the best quality nan-

otubes, has undergone a wide development over a decade. This

is a physical vapour deposition technique in which an electric arc

is used to vaporize material from a cathode target. This vaporized

material condenses on a substrate to form desired structures. The

rate of nanotube synthesis by this method is quite high with depo-

sition rates of typically 20-100 mg/min [29]. This method was first

reported by Ebbesen, Ajayan, Iijima and Ando in 1992 [30, 31] and

has a typical yield of 30 %. Further modifications to this conven-

tional method is under way e.g. the use of hexane or other organic

vapours to increase the yield, arc-discharge under liquid N

or wa-

ter and by the use of magnetic field.

• Chemical vapour deposition (CVD)

This method consists of producing carbon in gaseous phase and using an energy source such as plasma or a resistively heated fur- nace to transfer energy to gaseous carbon containing molecules and passing the gases through metal catalysts. This method was first used by Jose-Yacaman in 1993 [32] and has yield of upto 100 %.

This method has higher productivity than the arc-evaporation and laser vaporization processes. However, the main disadvantage of catalytic CVD-based production is the lower quality of the nan- otube structures as compared to those produced by other tech- niques.

• Laser vaporization

In this method a furnace is heated and an inert gas is made to flow through a tube at a constant pressure. A doped cylindrical graphite target is mounted at the center of the furnace. Vaporiza- tion of the target is performed by pulsed laser. This technique was first reported by Smalley in 1995 [33]. This method produces high quality single walled CNT but the production cost is high and yield is low.

In Fig. 2.1 we observe a transmission electron microscopic (TEM) im- age of single walled carbon nanotubes.

Figure 2.1: TEM image showing SWNT bundles, [inset] A high resolution TEM

image of a 2 nm diameter SWNT [34].

2.1.2 Graphene

To isolate an individual graphene layer Geim et al. used the most ba- sic peeling method utilizing a common Scotchtape to successively re-

move layers from a graphite flake. The tape was ultimately pressed down against a substrate to deposit a sample [16]. SiO

on Si is being used suc- cessfully as the substrate for graphene deposition [35]. Graphene can also be obtained by heating silicon carbide to high temperatures (≥ 1100

C) to reduce it to graphene [36]. This process produces epitaxial graphene with dimensions dependent upon the size of the SiC substrate.

Recently it has been shown that graphite spontaneously exfoliates into single-layer graphene in chlorosulphonic acid, and dissolves at high isotropic concentrations [37]. This occurs without the need for any cova- lent functionalization, surfactant stabilization or sonication, which may cause decrease in the flake size and the properties of graphene. Hernan- dez et al. demonstrated graphene dispersions with concentrations up to 0.01 mg/ml, produced by dispersion and exfoliation of graphite in organic solvents such as 1-Methyl-2-pyrrolidone [38]. Novel methods for synthesis of graphene in bulk are an ongoing topic of research interest and in the near future we shall witness many more such techniques. Fig. 2.2 shows a TEM image of a suspended single layer graphene membrane and in Table 2.1 different methods of synthesizing graphene is summarized.

For practical applications in some fields e.g. polymer nanocomposites, it is extremely challenging to be able to use individual exfoliated sheets of graphene, therefore few layers of graphene termed as graphene/graphite nanoplatelet (GnP) are used.

2.2 Properties

Carbon nanotubes and graphene are known for their excellent mechan-

ical, electrical, thermal and electronic transport properties which are

summarized in Table 2.2.

Figure 2.2: Bright-field TEM image of a suspended graphene membrane. Its central part (homogeneous and featureless region indicated by arrows) is mono- layer graphene (scale bar 500 nm) [39].

Table 2.1: A summary of the most common methods of graphene synthesis [40].

Single layer Few layers

Micromechanical cleavage of highly ordered pyrolytic graphite

Chemical reduction of exfoliated graphene oxide (2-6 layers)

Chemical vapour deposition on

metal surfaces Aerosol pyrolysis (2-40 layers) Epitaxial growth on an insulator

(Silicon Carbide)

Arc exfoliation in presence of Hydro- gen

Reduction of single layer graphene oxide

Thermal exfoliation of graphite ox- ide (2-7 layers)

Dispersion of graphite in water, 1- Methyl-2-pyrrolidone (NMP) Intercalation of graphite

2.3 Polymer composites of CNT and graphene

2.3.1 Dispersion

High aspect ratio, strong van der Waals force between tubes and flexi-

bility of CNTs leads to physical entanglement. For graphenes additional

interplaner π − π interactions makes it even more challenging to disperse

these nano-fillers into matrices. They always aggregate together to form

Table 2.2: A summary of properties of CNT and graphene [41–43].

Property Carbon nanotube Graphene

Young’s modulus ∼ 1 TPa for SWNT ∼ 0.5 - 1 TPa

∼ 0.3 - 1 TPa for MWNT

Tensile strength ∼ 50 GPa for SWNT ∼ 0.5 TPa

∼ 10 - 150 GPa for MWNT

Resistivity ∼ 5 - 50 µΩcm ∼ 50 µΩcm (in-plane)

Electrical conductiv-

ity ∼ 10

Ωcm ∼ 10

Ωcm

Thermal conductiv- ity

∼ 3500 Wm

K

(along the axis)

∼ 4000 - 6000

Wm

K

∼ 1.5 Wm

K

(in the radial direction)

Oxidation in air 700

C 450 − 650

C

bundles or ropes and in composites this can lead to reduction of the load carrying capability between the reinforcing phase and the matrix [44].

Recently, efforts have been made to develop effective methods to dis- perse SWNTs [45] and at the same time retain their length [46]. Fig. 2.3 illustrates some of the methods that are commonly employed to disperse CNTs. The strategies used for dispersion include the following techniques 1. Using π − π induced interactions, nanofillers can be dispersed in solvents using the physical adsorption of p-aromatic compounds [47].

2. Making other non-covalent induced dispersions (van der Waals, physical adsorption), by wrapping with surfactants and polymers [47, 48] or by the use of charged nanoparticles (ZrO

) [49]. Surfac- tants have assisted the dispersion process to some extent as they de-bundle CNTs by steric or electrostatic repulsions [50] and im- provement in properties compared to pristine CNT composites of polymers have been recorded [51–53].

3. Making covalent bonded side wall functionalization/oxidation in-

duced dispersions. The chemical modification of nanofiller wall by

forming covalent bond with substituents on the walls can greatly increase the dispersibility of the modified CNTs. For instance, nan- otubes could be functionalized with various functional groups, such as aliphatic amines, carboxylates [46, 54, 55] resulting in their im- proved dispersibility [56, 57].

Figure 2.3: Surface modifications of carbon nanotubes. a) aromatic molecules can be attached to nanotubes using non-covalent interactions (π − π inter- actions), b) non-covalent interaction can be used to wrap polymers around nanotubes, c) chemical groups can be covalently attached to nanotubes [58].

2.3.2 Methods for manufacturing polymer nanofiller composites

There are several techniques used in industry and research to produce well dispersed nanofiller-polymer composites. Here we discuss four effec- tive and most commonly used methods

1. Melt compounding: This method involves high shear mixing of

molten polymer with the nanofillers and is mostly applied for ther-

moplastic polymers. Physical mixing using extruders, kneaders etc

is performed where the temperature of the melt governs the viscos-

ity [59, 60]. We use this technique for our study of CNT/graphene filled polyamide 12 fibers and films.

2. Solution mixing: For polymers which form stable solutions, this is a good method to mix nanofillers. The solution together with the nanofillers is sonicated to uniformly disperse the fillers in the matrix and then subsequently the solvent is removed [61]. For graphene it is also possible to chemically reduce the material during solution phase mixing [28].

3. High shear mixing - calendaring: This method is mostly used for thermoset polymers like epoxy resins. Three roll milling is a com- mon type of calendaring where the material is placed in between rotating rollers and they get mixed under a high shear force [62].

This technique can be scaled up for industrial use [1].

4. In-situ polymerization: In-situ polymerization is another approach to ensure better distribution and good interaction of CNTs in the matrix [18, 63]. Here the nanofillers are added to the polymeriz- able master solution and physically mixed or sonicated followed by polymerization of the matrix. It is also possible to directly spin composite fibers from such a master solution of Caprolactam con- taining CNT [64].

Other methods include noncovalent attachment techniques like poly- mer wrapping and polymer adsorption. grafting to and grafting from methods [25] are also employed effectively.

2.4 Thermoplastic polymer: Polyamide

Thermoplastic polymers can be melted into a liquid above the melting

point and it freezes to a very glassy state when cooled sufficiently without

involvement of any chemical processes. They can be moulded reversibly

and re-used multiple times. Polyamides are an important class of ther-

moplastic polymers which have a wide range of industrial and household

applications. These thermoplastic polymers are known for their exotic

properties e.g. extreme toughness, abrasion resistance, good chemical re- sistance, light weight, low water absorption, good electrical insulation etc. and are therefore extensively used. As seen in Fig. 2.4 there are two polymorphic phases of polyamide-12(PA12), namely the α and the γ phase, the more stable one being the γ form. Depending on the crys- tallization conditions, four PA12 polymorphs designated as α, α

, γ, and γ

, were reported [65]. The polymorph obtained by quenching from melt and subsequent crystallization is denoted as γ

modification. The γ and γ

phases of PA12 have similar structure and x-ray diffraction patterns.

A hexagonal lattice has been assigned to both, characterized by only one strong x-ray reflection; these two phases can be differentiated by NMR spectroscopy [65, 66].

The mechanical properties of the PA12 are largely dependent on the crystalline structure of the matrix and inclusion of CNTs affects this crystalline structure which can be quantified from x-ray (both wide and small angle) diffraction, nuclear magnetic resonance (NMR) and differential scanning calorimetry(DSC) [59, 65, 67] studies.

2.4.1 Polyamide nanocomposites

Polyamides are universally used as matrix material for composites. Com-

mon methods for preparing such composites are melt-compounding, in-

situ polymerization and grafting. With MWCNT loading of less than

2 wt% there has been significant improvement in the modulus, strength

and hardness in PA6. Electrically conducting non-woven PA6 membranes

have been made with MWCNT adsorbed on the surface [25]. CNTs also

influence the crystallinity of PA6 as detected by x-ray scattering and dif-

ferential scanning calorimetry (DSC) studies [59]. Reports suggest that

nanofillers act as nucleation sites for the formation of new crystalline

domains [69]. Recently graphene-PA6 composites prepared by in-situ

ring opening polymerization showed excellent improvement in mechan-

ical properties of the polymer, even at a low graphene concentration

of 0.1 wt% [70]. With PA6 electrical percolation could be achieved at

2.5 wt% or higher CNT loading [27]. Melt spun PA12 fibers reinforced

Figure 2.4: Schematic showing structure of PA12 a) α phase, where the polymer chains run anti-parallel to each other and the H-bonds are in plane b) γ phase, where the polymer chains run parallel to each other and the H-bonds are twisted out of the plane [68].

with CNTs showed variation in the crystallite sizes and also an improve- ment in the stiffness. For fiber samples, with or without nanofillers, draw- ing makes a vital difference in its mechanical properties [60]. PA12 CNT composite with improved mechanical properties and reduced thrombo- genicity (antithrombotic property) is very promising for use in medical applications, such as hemodialysis, cardiopulmonary bypass, intravascu- lar catheters, left ventricular support, and vascular dilating devices [71].

From literature we see that laser sintering of PA12/carbon nanofiber composites resulted in improvement in tensile properties [72].

Recently it has been reported that with carbon nanofibers in PA12

the thermal and thermo-oxidative stabilities of the matrix could be

improved along with the stiffness of the material [57]. Composite of

PA12 with a small amount of the functionalized expandable graphene

caused a significant improvement in tensile strength, elongation to

break, impact energy and toughness of the polymer, although no

significant improvement in Young’s modulus was reported [73]. Fig. 2.5

Figure 2.5: SEM images of a) PA12/CNT composite showing well-dispersed CNTs protruding from the surface [60] b) PA12/functionalized graphene com- posite with flakes visible in the polymer matrix [73].

shows scanning electron microscopic (SEM) images of fractured surfaces of PA12 composites with CNT and graphene. Several other polyamides have also been used as the matrix material for composites and have proven to be effective in enhancing the mechanical, electrical and thermal properties indicating a bright future for such materials.

2.5 Thermoset polymer: Epoxy

Thermoset polymers cure or harden (set) into a given shape through irreversible chemical processes by cross-linking. Unlike thermoplastics, post curing the polymer cannot be melted or recast in any other form without degrading the material.

Epoxy is perhaps the most widely used thermoset polymer today with

its manifold applications such as paints, coatings, electrical insulators,

molds, castings, plastic tooling and recently it has gained considerable

attention as the primary matrix material used in the aerospace

industry. Epoxy composites are capable of being processed in bulk

industrial quantities and find various applications. Epoxy is essentially

a co-polymer made from two chemical components namely, a resin and

a hardener. The resin consists of monomers or short chain polymers

with an epoxide group at either end. Epoxy resin as seen in Fig. 2.6

is commonly produced from a reaction between epichlorohydrin and

Figure 2.6: Structure of Epoxy resin. "n" denotes the number of polymerized subunits and can be upto 25.

bisphenol-A or a similar chemical. The hardener consists of polyamine monomers e.g. triethylenetetramine (TETA). When these compounds are combined together, the amine groups react with the epoxide groups to form a covalent bond. Each NH group can react with an epoxide group, so that the resulting polymer is heavily cross-linked contributing to the strength and rigidity of the material. The process of polymer- ization is called curing and can be controlled through temperature, choice of resin and hardener compounds and the ratio of the compounds.

2.5.1 Epoxy nanocomposites

When cured, the highly-cross linked microstructure provides the epoxy

matrix with high modulus of strength, good resistance to creep, good

performance at elevated temperatures but poor ductility. The earliest

record of epoxy-nanotube composites is by Ajayan in 1994 [74] where

nanotubes were aligned within the epoxy matrix by the shear forces

induced by cutting with a diamond knife. Schadlerwt et al. [75] first

measured the stress - strain properties of a MWNT - epoxy composite in

both tension and compression mode and reported an increase in modulus

from 3.1 GPa to 3.71 GPa on addition of 5 wt% nanotubes. Since then

significant improvement in mechanical properties have been reported for

epoxy nanocomposites by several groups. Xu et al. [76] observed a signif-

icant increase in modulus from 4.2 to 5 GPa at only 0.1 wt% MWCNT

loading, Allaoui et al. obtained good all-round properties for MWCNT

epoxy composites with an increase in modulus from 0.12 GPa to 0.47

GPa on addition of 4 wt% CNT [77].

Epoxy composites boast of particularly low values of percolation threshold for electrical conductivity compared to other polymer composites. Percolation thresholds significantly lower than 0.1 wt % were reported which can be attributed to kinetic percolation that allows for particle movement and re-aggregation [78, 79]. At 5 wt% graphene filled epoxy resin showed a thermal conductivity of 1 Wm

K

, which is 4 times higher than pure epoxy and at 20 wt% loading the conductivity increased to 6.44 Wm

K

[80].

2.6 Specifications of materials used

Two different types of industrial grade multi-walled carbon nanotubes (MWCNT) were used from suppliers Nanocyl SA, Belgium (NC 3150) having pure carbon content higher than 95 wt% and Bayer Material Sci- ence AG, Germany (Baytubes C150P). The MWCNTs (Fig. 2.7a) have average length greater than 1 µm and diameter of 10-20 nm. Graphene nanoplatelets (GnP) from XG Sciences Inc. (xGnP), Michigan, USA, as shown in Fig. 2.7, were used for the reinforcements of the composites.

The GnPs are constituted of a few sheets of graphene stacked together and the sheet thickness ranged from 6-15 nm. GnPs of two different sheet sizes namely 5 µm and 25 µm were used to study the size effects of such nanofillers. Expanded GnPs prepared from natural graphite flakes (NG) were used for a study on epoxy composites. The flakes were immersed in a mixture of concentrated 98% sulfuric acid and 65% nitric acid for 15 min. at room temperature. This caused intercalation of sulfuric acid into the graphene layers of the NG flakes. The intercalated compound was then washed by distilled water, neutralizing to a pH value of about 7 and dried in an oven for three hours at & 100

C. The dried inter- calated compound was then heated at & 1000

C for about 10 seconds.

The thermal shock led to sudden volatilization of the intercalate within

the graphene layers resulting in a 50-100 times expansion of the flake

along the thickness direction (c-axis). This material is termed expanded

graphene nanoplatelets (EGNPs). The EGNPs are extremely light, one

gram of material typically occupies a volume of 200 − 300 cm

even

compared to CNTs which have a density of 1.4 g/cm

.

A copolymer MV 03 Micro from Deurex AG (Germany) was used as surfactant. For acid oxidation the pristine nanofillers were refluxed with 9 M nitric acid for 24 hrs to carboxylate the CNT walls and GnP sur- faces in order to remove catalytic metal particles also present in the raw material [81, 82]. Thereafter, the CNTs and GnPs were washed several times and neutralized with distilled water to pH 7. For composite fibers with MWCNT industrial grade masterbatch (pellets), Plasticyl PA1502 from Nanocyl SA, Belgium comprising of 15% multi walled CNTs of 90%

purity in PA12 (Grilamid L16) was used.

Two different polymer matrices were studied to analyze the reinforcing properties of the nanofillers. The thermoset matrix, Epoxy resin used is Epikote 828LVEL (Hexion, USA), which is produced from bisphe- nol A and epichlorohydrin. We used an aromatic di-amine curing agent, Epikure 3402 (Hexion, USA).The thermoplastic polymer used is PA12 and was provided by EMS-GRILTECH (GRILAMID L 16) in the form of pellets.

Figure 2.7: SEM images of a) GnP of size 5 µm b) GnP of size 25 µm c)

Expanded GnP d) CNT.

3. Experimental details

A variety of experimental methods has been used in different phases of the project. We can broadly divide the experiments into two parts namely the procedures for preparation of nanocomposites and the different tech- niques used for characterization of these composites. The experiments were performed at Empa, St Gallen and Dübendorf and at ETH, Zürich.

X-ray diffraction measurements were carried out at MAX-lab, Sweden and Swiss Light Source, Switzerland. In this chapter, we discuss in de- tails the different methods we have used during the study.

3.1 Processing of polyamide-12

Extrusion: Twin screw extrusion is an integral part of polymer process- ing technology. They are widely used for reactive processing including both polymerization and grafting reactions, for compounding, blending, de-volatilization, as well as for thermoplastic final shaping operations [83]. The screws are rotated in the same or opposite directions to ensure high sheer mixing of the components mostly at elevated temperatures.

We use extrusion to disperse the nanofillers in the polymer matrix oper- ating at a temperature higher than the polymer melting point. The high sheer created in the molten polymer by the rotating screws facilitates the compounding process.

PA12 Films

Multiwalled CNT and GnP were used for preparing composites. Before

compounding the PA12 pellets were dried at 60

C for 24 hours to

remove the water content. The PA12 pellets with systematically

varying loading of MWCNT/GnP were added in batches of 4 g each to

a twin-screw micro-extruder (HAAKE Minilab) operating at 190

C

and 180 rpm for 180 s. A picture of the micro-extruder used in our

laboratory is shown in Fig. 3.1. The temperature governs the viscosity of

the mixture. A wide range of temperatures was tried above the melting

point of PA12 which is 179

C [84] but we learnt that above 190

C the polymer gets oxidized/degraded very fast. Longer mixing time also causes degradation in PA12 which visibly turn yellow. An optimal set of parameters was found using systematic sets of trials. For surfactant mixed samples the CNT and GnP were physically pre mixed with the surfactant powder in 1:1 ratio by an agate mortar and pestle for 30 minutes before compounding with the PA12 in the micro-extruder. The composite material was hot pressed and then immediately quenched at a cold press and films of average thickness of 100 µm were obtained.

Three types of MWCNT and GNPs were used for the film samples.

a) Pristine b) Acid oxidized c) With a surfactant

Figure 3.1: Twin screw micro-extruder used for making the composites.

PA12 Fibers

The PA12/CNT masterbatch used for the composites needed to be di- luted with PA12 to reach desired CNT concentrations. PA12 together with the CNT masterbatch was extruded in a twin-screw extruder to obtain different CNT concentrations within the polymer. Virgin PA12 fibers were spun at the in-house melt pilot spinning plant SPIDER [85]

with draw ratios 3, 4, and 4.5. Draw ratio (DR) is given by Draw ratio = Length of fiber segment after drawing

Original length of fiber segment . (3.1)

The ratio of the speed of the rotating wheels that take up the fiber

governs the drawing and defines the draw ratio. Another set of fibers

were spun with varying concentrations of CNT mixed with PA12. CNT mixed fibers were manufactured with CNT concentrations of 0.003 wt%, 0.0075 wt%, 0.015 wt%, 0.03 wt%, 0.075 wt% and 0.15 wt% each with draw ratios: 3, 4 and 4.5. Some of the as-spun fibers were annealed for 3 hrs at 160

C and allowed to cool slowly.

Figure 3.2: a) schematic of the in-house melt-spinning pilot plant, SPIDER b) picture of the set-up.

SPIDER: Spider (Spinning - development - research), as the melt spin-

ning pilot plant has been dubbed at Empa was built by Fourné Poly-

mertechnik (Alfter-Impekoven, Germany) in 2004 based on Empa speci-

fications is shown in Fig. 3.2. It produces fibers consisting of one or two

different thermoplastic polymers. The bi-component fibers can be laid

side by side or arranged in a core/cladding structure and may be round,

square, filled or hollow in structure. The machine consists essentially of

two thermoplastic extruders and a piston extruder that melts and ho-

mogenizes the polymer granules. Pumps ensure exact proportioning of

the plastic and determine the fineness of the fibers according to the haul-

off and winding speeds. The spinneret, a plate with a specific number

of holes of defined cross section, dictates the number of filaments and

their diameter. The solidification and cooling process is controlled by an

air stream in the blowing shaft. The draw take-up machine enables the

fibers to be drawn and heat-treated, thereby allowing their mechanical properties to be specifically adjusted.

3.2 Processing of epoxy

Three different manufacturing processes were used for nanofiller/epoxy composites as described in a schematic flow chart in Fig. 3.3. In pro- cess (a) pristine nanofillers were pre-dispersed in acetone by ultrason- ication for 15 min at 35 kHz. The resulting nanofiller suspension was then processed by a high pressure homogenizer (HPH) from M110Y, Microfluidics, USA at 5.5 MPa for 10 min. The suspension was subse- quently mixed with epoxy resin using a high shear mixer at 20,000 rpm for 10 min. Acetone was removed completely by ultrasound assisted ro- tary evaporation at 100 mbar and 60

C with a vibration frequency of 35 kHz in order to retain uniform nanofiller dispersions after removing ace- tone. Process (b) is similar to process (a) but includes three-roll milling (SDY200 from Bühler AG, Switzerland) as an additional step after the pre-dispersion using HPH. In process (c) the pristine nanofillers were dispersed directly within the epoxy resin using three-roll milling at 30

C and at a gap pressure of 1MPa. Eventually, the mixture was passed three times through the three-roll mill. All nanofiller/epoxy composite resins were mixed with the hardener using an epoxy:hardener ratio of 100:24.

Resulting composite was then degassed within a vacuum oven at 40

C in order to remove air bubbles and the residual solvent. Subsequently, the composite was moulded at 80

C for 12 h, followed by post-curing at 120

C for 4 h. The same techniques were used for all the different types of nanofillers used namely CNT, EGNP and GnP.

3.3 Characterization

The composite samples were characterized by a wide variety of techniques

to study their morphological and physical properties. Different methods

used are as following

Figure 3.3: Processing steps for epoxy composites.

1. Optical microscopy: Modern compound optical microscopes utilise transmitted visible light and a series of lenses to focus and magnify a given object. Extensive optical microscopy (Olym- pus SZX16) was performed to visually examine the nanofiller dispersion at a macroscopic level and for re-agglomeration studies.

2. Scanning electron microscopy (SEM): SEM is used to im- age a sample by scanning the surface with a beam of electrons in a raster scan pattern. The electron beam interacts with the electrons in a sample producing secondary electrons, back scattered electrons etc which contain information about the sample’s surface topogra- phy, composition and other physical properties. In order to obtain micro-structural and micro-analytical information, the composite materials were characterized by high resolution scanning electron microscope (Nova NanoSEM 230).

3. Transmission electron microscopy (TEM): This is an electron

microscopy technique where a beam of electrons is transmitted

through an ultra thin specimen, interacting with the specimen as it passes through. An image is formed from the interaction of the electrons transmitted through the specimen containing information about the structure of the sample. This image is also obtained using a scanning technique and then magnified and focused onto a detector. JEOL 2200FS has been used for high resolution imaging.

4. Tensile testing: For films, tensile testing was carried out on dog bone like samples with a Zwick Roell Z010 in order to determine the Young’s modulus, tensile strength and strain for all specimens. The midsection of a dog bone shaped sample has a narrower width than the grip section (where it is attached to the instrument). This concentrates the stress in the test area, so that fracture and most of the strain occur in the central portion. Strain was measured in this section, and stress was calculated from the force load on the grips. The tests were performed according to the standard ASTM D882-10 [86]. For the fibers tensile testing was carried out on Instron apparatus using testing standard ISO 2062:2009 [87]. For epoxy plates two different testing techniques were used:

Three point bending: In this testing method, the simple is supported on two outer points, and deformed by driving the third central point downwards. The maximum stress is located at the centre. The force and deformation is utilized to calculate the elastic modulus, the flexural modulus and the elongation of the sample. The tests are performed abiding by the ISO-standard 178:2001 [88].

Single edge notch tension (SENT): This test is car-

ried out to calculate the fracture toughness which is an indication

of the amount of stress required to propagate a pre-existing flaw

such as a crack through the sample. A crack (notch) is initiated

in the sample to be tested, the force and geometry of the sample

and the induced crack is used to calculate the value of fracture

toughness using the methods described in [89]. ASTM standard, ASTM E1820-11 is followed to perform the measurements [90].

5. Resistivity measurement: A four-point probe device was used to measure electrical conductivity of the materials. The probe used was Resistivity Test Rig, Model RM3-AR from Jandel Engineering, U. K. The rig consists of four tungsten metal tips of radius 100 µm with equal spacing of 1 mm with 100 g applied load. The cur- rent was passed across the outer two probes and the corresponding voltage across the inner two probes and the sheet resistance were measured, from which the conductivity is deduced.

6. Thermal conductivity measurement: The thermal diffusiv- ity of EGNP/epoxy composites was measured using a LFA447 Nanoflash apparatus at room temperature. Samples were prepared in disc-shaped forms, with diameter of ca. 10 mm and a thickness of 1.0 mm. The specimens were mounted in a graphite holder located inside a graphite susceptor.

7. Differential scanning calorimetry (DSC): From DSC, the glass transition temperature, the melting point and the crystallization/melting curve can be determined. The value of entropy of fusion can be deduced and subsequently the percentage crystallinity in the sample can be calculated from this technique.

DSC measurements were carried out with Perkin Elmer DSC 7 in the range of 0

C to 220

C in nitrogen atmosphere at a heating rate of 20

C/min.

8. Thermogravimetric analysis (TGA): This method measures the changes in weight with temperature for a physical system. The weight loss curve is used to determine the degradation tempera- tures, absorbed moisture content of materials and the level of in- organic and organic components in materials. Measurements were performed with Perkin Elmer TGA 7 at 5

C /min in oxygen at- mosphere in the temperature range of 25

C to 800

C.

9. Nuclear magnetic resonance (NMR): This technique relies on

a phenomenon which occurs when the nuclei of certain atoms are

immersed in a static magnetic field and exposed to a second os- cillating magnetic field. Some nuclei experience this phenomenon, and others do not, depending upon whether they possess a property called spin. The magnetic field at the nucleus is not equal to the applied magnetic field; electrons around the nucleus shield it from the applied field. In our case of

C NMR the radio frequency elec- tromagnetic radiation when sent through the nucleus interacts with the spin to split it into two energy states. The difference between the applied field and the field at the nucleus is termed the nuclear shielding or the chemical shift. Chemical shift is a function of the nucleus and its environment and thus we can derive information about the elements in our samples. The solid state

C magic angle spinning (MAS) was carried out at the Bruker AVANCE-400 MHz NMR Spectrometer at 100.61 MHz using a 7 mm CP-MAS probe at MAS rates of 3500 Hz.

10. Wide angle x-ray diffraction (WAXD): WAXD is an effective method to determine the crystalline structure of polymers. This technique specifically refers to the analysis of Bragg peaks scattered to wide angles, which implies that they are caused by sub-nanometer sized structures. From the angle of diffraction/scattering it is possible to calculate the inter-planer spacing of the polymer chains within the crystals. The diffraction is governed by Bragg’s law given by

2d sin θ = nλ (3.2)

where d is the inter-planar spacing, θ is the scattering angle and

λ is the x-ray wavelength. For fiber samples WAXD measurements

were carried out at synchrotron beamline I 711 at MAX-lab, Swe-

den. The detailed description of the beamline can be found in liter-

ature [91]. The beamline has a single crystal monochromator which

provides a high photon flux in expense of easy tunability and high

energy resolution. The experimental setup has a 4-circle diffrac-

tometer with kappa geometry, capable of doing both single crystal

and powder diffraction using a large area CCD detector. X-ray

wavelength of 1.1 Å was used for our study. For the film samples WAXD was carried out using a table top setup at ETH, Zürich consisting of a four-circle diffractometer, Xcaliber with CCD from Oxford diffraction with a Mo K

source at a wavelength of 0.71 Å.

The films before and after tensile testing were used as samples with

an exposure time of 5 minutes. For both table top setup and syn-

chrotron radiation a blank spectra is recorded and subtracted from

the image frames to eliminate background noise.

4. Polyamide Composites

Polyamide-12 is used as matrix material to understand the effects of addition of carbon nanofillers on structure and properties of the resul- tant composites. Two different forms of PA-12 nanofiller composites are studied, namely, polymer films and melt spun fibers. For poly-crystalline polymers like PA-12, crystallinity plays an important role in their me- chanical properties. Structure, crystallinity, mechanical and electrical properties of the composites are thoroughly studied by a variety of tech- niques which shall be discussed in this chapter.

4.1 Dispersion of nanofillers

As already discussed in section 2.3.1, it is a challenging job to achieve uniform dispersion of CNTs and GnPs in polymer matrices. In Fig. 4.1a, we observe certain discrete sites with CNT agglomeration in 5 wt% pris- tine CNT composite sample after extrusion. Large areas with CNT ag- glomeration can also be seen in Fig. 4.1b on an oxidized CNT composite sample. Fig. 4.1c, shows agglomerated GnP particles in the same poly- mer matrix system. We have observed some of the larger agglomerates with sizes ranging from 5-40 µm. Oxidized CNTs and GnPs are no longer loose powders. During processing (mainly while drying) the nanoparti- cles became clustered together and it is extremely difficult to break them apart even with extrusion. The oxidized fillers are supposed to make bet- ter linkage with polymer matrix due to increased adherence and chemical bonding but on account of their initial state of agglomeration it is not possible to achieve composites with well dispersed oxidized nanofillers.

However, individual CNTs have been observed in several fractured sur-

faces as shown in Fig. 4.2a for composites prepared using the surfactant

at 5 wt% CNT loading, suggesting uniform dispersion over large areas. As

seen in figure Fig. 4.2b, for GnP with surfactant, there is a distribution

Figure 4.1: SEM images showing agglomeration sites within the composites a) Pristine CNT-PA12 composite, magnified image of one such agglomeration site is provided in the inset b) Oxidized CNT-PA12 composite c) Oxidized GnP-PA12 composite (arrow indicates the location of GnP agglomerate)[2].

of the fillers. However, it is not possible to achieve uniform dispersion of individual graphene layers and there are agglomeration sites present.

Figure 4.2: SEM images of fracture surfaces of a) surfactant + CNT composite b) surfactant + GnP composite at 5 wt% nanofiller loading [2].

Practically it is more challenging to disperse GnPs compared to CNTs.

Possible reasons may be attributed to the fact that in addition to the

van der Waals forces a strong π − π interaction between individual GnP

sheets is also present. During melt processes, such as extrusion and sub-

sequently hot pressing, there is a considerable scope for re-agglomeration

or secondary agglomeration for thermoplastic polymers, like PA12, as re-

ported in literature [92]. This phenomenon may eventually help in elec-

trical percolation.

4.2 Nuclear magnetic resonance (NMR)

From solid state NMR measurements, chemical shifts and line widths of individual resonances are determined by non-linear least-square fit of a sum of Gaussian/Lorentzian curve using a software by Massiot et al.

[93]. These values when compared to the results reported in [94], the chemical shifts of all samples (with and without CNTs) correspond to the γ

polymorphic phase. Comparison of the pure PA12 fibers before and after annealing, with literature is shown in table 4.1. Samples are annealed, heated to 160

C for 3 hrs and are allowed to cool down slowly.

For these annealed samples the chemical shift corresponds to the γ form.

In all the samples, it is observed that the α phase is completely absent, this is common for PA12. From literature, it is known that PA12 when quenched (cooled very fast) at atmospheric pressure from melt, is more prone to crystallize in the γ

phase instead of α or γ phase and on annealing above 110

C at atmospheric pressure, a γ

to γ transformation occurs [66].

Table 4.1: Chemical shifts (in ppm) of

C in pure PA12 samples.

Sample C

CH

gauche CH

all trans C

C

C = O

PA12 fiber 27.3 30.9 33.1 37.0 40.4 173.6

Annealed PA12 26.9 30.5 33.5 37.0 39.9 173.0

γ[94] ≈ 28 30.6 33.5 37.1 40.0 173.7

γ

[94] ≈ 27 30.8 33.2 37.0 40.6 173.7

α[94] 27.3 31.2 34.3 38.7 42.4 172.8

4.3 Wide angle x-ray diffraction (WAXD)

2-D WAXD is an extremely effective tool to understand the morphology

and crystalline structure of the composites, the properties of such

materials are largely dependent on structure. WAXD patterns provide

us with information about the polymorphic phases present, an estimate

of the percentage crystallinity and crystalline orientation.

Films

Polyamides are in general constituted of both α and γ polymorphic phases. However it has been previously reported for PA12, that the only phase observed in all the samples is the γ phase [60].The films are un- oriented (Fig. 4.3a), however after tensile testing, the crystalline peaks obtain a preferred orientation as seen in Fig. 4.3b. Two sets of peaks are observed in Fig. 4.3b which are termed as the meridional (inner reflec- tion) and equatorial (outer reflection) peaks. The meridional peak is from the phase γ020 which corresponds to a separation between the carbonyl oxygen atoms of individual polymer chains. The software XRD2DScan

Figure 4.3: 2-D WAXD images of a) 0.5% pristine CNT composite film b) 0.5%

pristine CNT composite film after tensile testing [2].

V4.1 [95] is used to integrate the 2-D WAXD pattern to generate a 1-

D intensity vs 2θ profile as shown in Fig. 4.4. The equatorial peak has

contributions from both the crystalline and the amorphous phases. This

peak is deconvoluted using MATLAB programming to separate the

crystalline and amorphous parts as shown in Fig. 4.4. The peaks are fit-

ted to Lorentzian and Gaussian line shapes, the Lorentzians result in a

better fit. The total fit consists of a background, an amorphous contri-

bution and a primary crystalline contribution. A small peak at 2θ ≈ 5.1

has to be included to improve the fit and it can be attributed to a second

order inter-planar spacing. The equatorial percentage crystallinity (ECI)

is calculated by [59]

ECI = area under crystalline peak

area under crystalline peak + area under amorphous peak . (4.1) For all samples of virgin PA12 and the composites, the position of the

Figure 4.4: Deconvolution of equatorial peak for pure PA12 films. The blue peak is the total fit to the original data (red) [2].

fitted equatorial crystalline peaks correspond to an inter-planar spac- ing, d = 4.15 Å to 4.22 Å. Thus, these peaks can be attributed to the γ200 plane [65]. However the ECI varied considerably for the samples.

In Table 4.2 we provide a summary of the values for 0.5 wt% nanofiller content.

An increase in the crystallinity, as observed from WAXD (Table 4.2),

may be related to the nucleation effects of the nanofillers with the

growth of new crystalline domains around them as previously observed

in literature [69]. From the higher ECI values it can be inferred that

GnPs act as the best nucleating agents whereas the presence of

surfactants with CNTs hinders the growth. However, the ECI values

do not reflect the total crystallinity as only the equatorial crystalline