EFFECT OF DISPERSION OF SWCNTs ON THE VISCOELASTIC AND FINAL PROPERTIES OF EPOXY BASED NANOCOMPOSITES by Cihan Uzunpinar A thesis submited to the Graduate Faculty of Auburn University in partial fulfilment of the requirements for the Degree of Polymer and Fiber Engineering, Master of Science Auburn, Alabama May 14, 2010 Carbon nanotubes, composites,dispersion Copyright 2010 by Cihan UZUNPINAR Approved by Maria L. Auad, Chair, Profesor of Polymer and Fiber Engineering Peter Schwartz, Head of Polymer and Fiber Engineering Yaser Gowayed, Profesor of Polymer and Fiber Engineering ii Abstract The properties of polymer nanocomposites depend primarily on the dispersion of nanoparticles and the interaction betwen nanoparticles and the polymer matrix. Homogeneous dispersion and good interfacial adhesion can transfer mechanical loads efectively to carbon nanotubes (CNT) and significantly improve bulk properties. However, because of the extreme inertnes of CNT, eforts to enhance the mechanical properties of CNT-polymer nanocomposites have met with only modest succes. The aim of this research is to understand the efect of the dispersion quality on the final properties (mechanical and thermo-mechanical) of single wal carbon nanotube/epoxy composites. Several dispersion techniques such as sonication, surfactant and chemical functionalization were chosen. The quality of the dispersion was evaluated through rheological studies. In addition the efect of the aspect ratio with diferent SWCNT concentration was examined by using experimental results and the Einstein analytical model. Using a combination of techniques (microscopy, thermal gravimetric analysis, rheology and dynamic mechanical analysis), an atempt was made to characterize the single wal carbon nanotube/epoxy systems. iii Acknowledgments The author would like to expres his thanks and gratitude to his advisor, Dr. Maria L. Auad, for her guidance, support and suggestions. The author also expreses his gratitude to the other members of his advisory commite, Dr. Yaser Gowayed and Dr. Peter Schwartz. I sincerely appreciate their time, efort, and critical evaluation of my work. Additionaly, I would like to acknowledge Mirna A. Mosiewicki and Roberto J. J. Williams for their great help and advices. Thanks to the Department of Polymer and Fiber Engineering for providing supportive welcoming environment. I also would like to mention that this research would not have been possible without funding from the Department of Commerce. Finaly, I would like to thank my parents and al the friends who have helped me and encouraged me over the many years. iv Table of Contents Abstract.........................................................................................................................................ii Acknowledgments ......................................................................................................................iii List of Tables..............................................................................................................................vii List of Figures............................................................................................................................viii List of Abbreviations....................................................................................................................x Chapter 1 Introduction .................................................................................................................1 1.1 Literature review ........................................................................................................1 1.1.1 Nanocomposites.....................................................................................................1 1.1.1.1 Ceramic matrix composites..............................................................................2 1.1.1.2 Metal matrix composites...................................................................................2 1.1.1.3 Polymer matrix composites..............................................................................3 1.1.2 Carbon nanotubes..................................................................................................4 1.1.2.1 Synthesis...........................................................................................................9 1.1.3 Chalenges............................................................................................................11 1.1.4 Dispersion............................................................................................................12 1.1.4.1 Ultra-sonication..............................................................................................13 1.1.4.2 Surfactant-asisted functionalization..............................................................13 1.1.4.3 Covalent functionalization..............................................................................14 1.1.5 Applications.........................................................................................................15 v 1.2 Research Objectives..................................................................................................17 References.......................................................................................................................18 Chapter 2 Dispersion of single wal carbon nanotubes epoxy suspensions................................23 2.1 Introduction................................................................................................................23 2.2 Materials....................................................................................................................24 2.3 Techniques.................................................................................................................25 2.3.1 Rheology..............................................................................................................25 2.3.2 Scanning electron micsroscopy...........................................................................26 2.3.3 Thermal gravimetric analysis...............................................................................26 2.4 Results and discussion...............................................................................................29 2.4.1 CNT/matrix dispersion........................................................................................29 2.4.2 Experimental results............................................................................................34 2.4.2.1 Efect of dispersion quality.............................................................................34 2.4.2.2 Efect of functionalization..............................................................................36 2.4.2.3 Efect of concentration....................................................................................37 2.4.2.4 Efect of aspect ratio and concentration.........................................................41 2.4.2.5 Efect of orientation........................................................................................47 2.5 Conclusion.................................................................................................................50 References .......................................................................................................................51 Chapter 3 Mechanical and thermal properties of SWCNT/epoxy composites...........................53 3.1 Introduction................................................................................................................53 3.2 Materials....................................................................................................................55 3.3 Techniques.................................................................................................................56 vi 3.3.1 Flexural experiment.............................................................................................56 3.3.2 Dynamic mechanical analysis (DMA) ................................................................56 3.4 Results and discussion...............................................................................................57 3.4.1 Mechanical properties..........................................................................................57 3.4.2 Thermo-mechanical properties............................................................................61 3.5 Conclusion.................................................................................................................65 References........................................................................................................................66 Chapter 4 Conclusion..................................................................................................................68 vii List of Tables Table 2.1 G? values at low angular frequency ...........................................................................35 Table 2.2 G? values at low angular frequency............................................................................37 Table 3.1 Properties of neat epoxy resin and nanocomposites for diferent systems of 1 wt% of SWCNTs.....................................................................................................................................58 Table 3.2 Properties of neat epoxy resin and nanocomposites for diferent SWCNT concentrations.............................................................................................................................59 Table 3.3 Properties of neat epoxy resin and nanocomposites...................................................64 vii List of Figures Figure 1.1 Alotropes of carbon....................................................................................................5 Figure 1.2 Schematic of a) SWCNT, b) DWCNT, c) MWCNT..................................................6 Figure 1.3 Schematic of chirality on graphene sheet....................................................................7 Figure 1.4 Chirality of CNTs a) arm-chair, b)zig-zag, c) chiral ..................................................7 Figure 1.5 a) Non-ionic surfactant molecule, b) CNT wrapped by surfactant molecule, c) a surfactant micele, d) CNT wrapped by surfactant miceles.......................................................14 Figure 2.1 Oxidation proces of SWCNT...................................................................................27 Figure 2.2 TGA result for pristine SWCNT and SWCNT-COOH.............................................28 Figure 2.3 SEM images of a) pristine SWCNT b) SWCNT-COOH..........................................28 Figure 2.4 Efect of the dispersion quality on the SWCNT suspension.....................................30 Figure 2.5 Efect of functionalization on the dispersion quality................................................31 Figure 2.6 Efect of the SWCNT concentration on the dispersion quality.................................33 Figure 2.7 Efect of the aspect ratio on the dispersion quality...................................................33 Figure 2.8 Efect of the pre-shear on the dispersion quality.......................................................34 Figure 2.9 Change of G?(Pa) with frequency for diferent SWCNT-epoxy suspensions...........35 Figure 2.10 Change of G? (Pa) with frequency for pure, acidified and surfactant plus sonicated SWCNT-epoxy suspensions (1 wt%).........................................................................................36 Figure 2.11 Change of G? (Pa) with frequency for diferent SWCNT concentrations...............38 Figure 2.12 G? values at low frequency for diferent concentrations.........................................39 ix Figure 2.13 Van Gurp-Palmen plor for diferent concentration of SWCNTs............................40 Figure 2.14 Change of complex viscosity for diferent SWCNTs concentration.......................41 Figure 2.15 Efect of the sonication time on the storage modulus (G?) of 1 wt% SWCNT surfactant-epoxy suspension.......................................................................................................42 Figure 2.16 Efect of the sonication time on the complex viscosity of 1 wt% SWCNT surfactant- epoxy suspension........................................................................................................................42 Figure 2.17 Efect of the sonication time at 0.25 wt% surfactant-asisted SWCNT epoxy suspension...................................................................................................................................43 Figure 2.18 Efect of the sonication time at 0.5 wt% surfactant-asisted SWCNT epoxy suspension...................................................................................................................................44 Figure 2.19 Efect of the sonication time at 1.5 wt% surfactant-assisted SWCNT epoxy suspension...................................................................................................................................44 Figure 2.20 Storage modulus at low frequency versus sonication time for diferent concentration of surfactant asisted SWCNT-epoxy suspension......................................................................45 Figure 2.21 Initial dynamic viscosity of the suspension as a function of SWCNT volume fraction........................................................................................................................................46 Figure 2.22 Efect orientation on G? for pure SWCNT suspensions..........................................48 Figure 2.23 Efect of orientation on G? for acidified SWCNT suspensions...............................48 Figure 2.24 Efect of orientation on G? for SWCNT-surfactant suspensions............................49 Figure 2.25 Efect of orientation on G? for 5 min sonicated SWCNT-surfactant suspensions..49 Figure 3.1 The comparison of experimental and analytical results for flexural modulus of SWCNT-epoxy system...............................................................................................................61 Figure 3.2 Storage modulus (E?) vs. temperature for diferent SWCNT-epoxy system............62 x Figure 3.3 Tandelta vs. temperature for diferent SWCNT-epoxy system.................................63 xi List of Abbreviations SWCNT Single Wal Carbon Nanotube MWCNT Multi Wal Carbon Nanotube CNF Carbon Nanofiber MPa Megapascal GPa Gigapascal TPa Terapascal DSC Diferential Scanning Calorimetry TGA Thermal Gravimetric Analysis FTIR Fourier Transform Infrared SEM Scanning Electron Microscopy AFM Atomic Force Microscopy DMA Dynamic Mechanical Analysis G? Storage modulus G? Loss modulus E? Storage modulus E?? Loss modulus T g Glas transition temperature 1! ! CHAPTER ONE INTRODUCTION 1.1 LITERATURE REVIEW 1.1.1 NANOCOMPOSITES Recent technological progres has created new chalenges in the field of materials, composites in particular. Experimental work has generaly shown that virtualy most types and clases of nanocomposite materials lead to new and improved properties when compared to their macro-composite counterparts. Nanocomposites can be defined as a composite material in which at least one of the phases (mostly the filer) shows dimensions in the nanometer range. As the filers size reach the nanometer level, the interactions at the interfaces become considerable large with respect to the size of the inclusion and thus the final properties show significant change. Nanocomposites promise new applications in many fields such as mechanicaly reinforced lightweight components (automotive parts), non-linear optical devices, bateries, nano-wires, sensors, bio-electronic, etc. Moreover, advanced technologies like bone cement, filter membranes, aerogels and solar gels are significant other applications for nanocomposites. Nowadays, nanocomposites offer new technology and busines opportunities for diferent industries (Camargo et al. 2009, Hay et al. 2000). 2! ! Nanocomposites can be clasified acording to their matrix materials in three diferent categories; ceramic matrix nanocomposites, metal matrix nanocomposites and polymer matrix nanocomposites. 1.1.1.1 Ceramic matrix nanocomposites (CMNC) Although typical ceramic composites have good wear resistance and high thermal and chemical stability, they are mainly britle materials. In order to overcome this disadvantage, the incorporation of energy-disipating components, such as nanowhiskers, fibers platelets, or particles into the ceramic matrix is a good way to increase the fracture toughnes of these materials (Sternitzke et al. 1997). The most common methodologies for CMNCs preparations are spray pyrolysis, vapor techniques and chemical methods such as the sol-gel proces. Potential applications for ceramic matrix nanocomposites are wear or friction surfaces, load-bearing structural parts, automotive, aerospace, power generation applications (engines, turbines), medical devices and implants. 1.1.1.2 Metal matrix nanocomposites (MMNC) Metal matrix nanocomposites refer to materials consisting of a ductile metal or aloy matrix in which some nanosized reinforcement material is implanted. They are high strength and high modulus materials with substantial ductility and toughnes (Tjong et al. 2000). They are suitable for uses in aerospace and automotive industries. The most common techniques for MNCs are mechanical aloying, vertex proces, spray deposition liquid metal infiltration, rapid solidification, and chemical proceses such as sol-gel and colloidal (Padhi et al. 2008). 3! ! 1.1.1.3 Polymer matrix nanocomposites (PMNC) Polymers are the most widely used materials for composites because of their lightweight, low cost, and good chemical resistance. Using a polymer as the matrix facilitates procesing and structural integrity while the filers add diferent functionalities. On the other hand, they have some disadvantages such as low modulus and strength compared to metals and ceramics. Therefore, an efective method for property enhancement is the addition of nanoreinforcements to the polymer matrix. For instance, polymers have been filed with several synthetic or natural compounds in order to increase heat and impact resistance, flame retardancy and mechanical strength. (Fischer , 2003). In recent years, the use of nanofilers as reinforcement of a polymer matrix has been gaining in popularity in industrial and academic research. These materials have been used as alternatives to conventional polymers or polymer blends. In mechanical terms, because of their enormous surface area to volume ratio and high aspect ratio (length to diameter),nanocomposites difer from conventional composite materials. The nanofilers? interfacial area (103-104 m 2 /ml) is typicaly orders of magnitude greater than conventional composite materials (30 m 2 /ml) (Chawla, 1998). Thus, the mechanical properties of the final composite are significantly afected by the strong interactions created at the matrix-filer interface. Ajayan et al.(2000) reported that local chemistry, degree of thermoset curing, polymer chain mobility, polymer chain conformation, degree of crystalinity and final physical and mechanical properties of the polymer nanocomposites can vary significantly and continuously from the interface of the reinforcement to the bulk of the matrix. Therefore, a relatively smal amount of nanoscale reinforcement (1-5wt %) can have significant efect on the macro-scale properties of the composites (Schadler et al. 1998, Lu et al. 1997). Haggenmueler et al. observed ~100% increase 4! ! in elastic modulus for single wal carbon nanotube (SWCNT)/poly(methyl methacrylate) (PMA) nanocomposites. Park et al. obtained 10 decade enhancement in electrical conductivity at a nanotube loading of only 0.1 vol % for SWNT/Polyimide nanocomposites (Du et al. 2004). Many methods have been described in the literature for the fabrication of polymer nanocomposite materials (Fischer et al. 2003, Thostenson et al. 2001), such as, intercalation of the polymer (Shen et al. 2002) (clay with high density polyethylene (HDPE) or polyvinylalcohol (PVA) etc. ); in-situ intercalative polymerization (montmorilonate with epoxy or polyurethane) (Yao et al. 2002); melt intercalation (montmorilonote with polystyrene) (Vaia et al. 1997), direct mixture of polymer and particulates (CNTs with epoxy) ( Sandler et al. 1999), template synthesis (hectorite with polyacrylonitrile (PAN)) (Watkins et al. 1995), and in-situ polymerization (poly(ethylene terephthalate) with calcium carbonate (Roslaniec et al. 2003)) and sol-gel proces (polyimide with silica (SiO 2 ) (Liu et al. 2002) . 1.1.2 CARBON NANOTUBES Carbon nanotubes, which were first discovered by Sumo Iijima in 1991, are comprised of sheets of carbon in a hexagonal latice that are rolled into a cylinder (Iijima 1991). Since their discovery, carbon nanotubes have been gaining popularity in research area due to their unprecedented physical and chemical properties (Kennedy 2008). Nanotubes are members of the fullerene structural family, which also includes the spherical bucky-bals that were discovered by Smaley and coworkers at Rice University (Kroto et al. 1985). Carbon nanotubes are long, slender fullerenes with wals composed of hexagonal carbon and capped ends. Their name is derived from their size, since the diameter of a nanotube is on the order of few nanometers (approximately 1/50,000 th of the width of a human hair). 5! ! Figure 1.1: Alotropes of carbon (Hoagland 2005) Three types of carbon nanotubes exist; they are multi-wall (MWCNT), double-wall (DWCNT) and single wal carbon nanotubes (SWCNT) (Figure 1.2). Multi-wal carbon nanotubes consist of a central tube of nanometric diameter surrounded by graphitic cylinders separated by approximately 0.34 nm. MWCNTs can have diameters ranging from 2 to 100nm and lengths of microns. DWCNTs have two graphene sheets rolled up to form a cylinder. Single wal carbon nanotubes have only one central tube. Their diameter is approximately 1 nm and their length is up to few centimeters (Rao et al., 2005). The density of a SWCNT is approximately 1.33-1.40 g/cm 3 and its elastic modulus is around 1.2 TPa. 6! ! Figure 1.2: Schematic of a) SWCNT, b) DWCNT, c) MWCNT (reproduced from http:/ww.wtec.org/loyola/nano/04_03.htm) Because of the carbon-carbon bond symmetry of the cylindrical tube, CNTs have a discreet number of orientations which are caled chirality. Chirality is the way of rolling graphene into a tube and is defined by a circumferential vector (chiral vector (n and m)). Figure 1.3 shows how a hexagonal sheet of graphene is rolled to form a carbon nanotube. If n and m are equal (n=m), then it is an arm chair circumference (arm chair CNT). If (n,m)=(n,0) or (0,m) then it is a zig zag circumference (zig zag nanotubes). Al armchair SWCNTs are metalic where as SWCNTs with n-m= 3i (i being an integer diferent than 0) are semi-metalic and n-m! 3i are semiconductors (Figure 1.4). For MWCNT, the chirality is more complicated since each tube that forms the nanotube may posses a diferent chirality. 7! ! Figure 1.3: Schematic of chirality on graphene sheet (Moniruzzaman and Winey 2006) Figure 1.4: Chirality of CNTs a) Arm-chair b) Zig-zag c) Chiral (Khare et al. 2005) 8! ! Carbon nanotubes exhibit exceptional properties that are a consequence of the covalent sp" bonds formed betwen the individual carbon atoms. They are one of the best performance filers that have low density (1.3 to 1.4 g#cm $3 ). Their mechanical properties have been reported by diferent research groups. The first measurements were made by Treacy et al. with MWCNT which were produced by arc discharge method. While measuring the amplitude of the intrinsic thermal vibration with transmision electron microscopy (TEM), Treacy and others found that the Young?s modulus ranged from 0.41-4.15 TPa (Treacy et al., 1996). The first direct measurement was made by Wong et al. in 1997 using atomic force microscopy (AFM) where the Young?s modulus of single wal nanotubes was found to be 1.28 TPa. Additionaly, the strength of the carbon nanotubes was estimated at around 14 GPa (Wong et al., 1997). Yu et al. performed stres-strain measurements for individual MWCNTs inside an electron microscope. They reported modulus values of 0.27-0.95 TPa, fracture strain of 12%, strengths in the range 11-63 GPa and toughnes of 1240 J/g (Yu et al., 2000). Overney et al. calculated the Young?s modulus of SWCNT to be 1.5 TPa similar to graphene (Overney et al. 1997, Lu et al. 1997). The measurement made by Salvetat et. al. using AFM method showed a tensile modulus of 1 TPa for smal diameter bundles. Intertube slippage within bundles is a serious limitation for SWCNTs mechanical properties, since efective modulus and strength for bundles are far below to the expected values. (Salvetat et al. 1999). Additionaly, the tubes can undergo plastic deformation under excesive tensile strain. This deformation begins at strains of approximately 5% and increases the maximum strain which the tubes undergo before fracture by releasing strain energy (Qian et al. 2002). CNTs are not nearly 9! ! as strong under compresion. Because of their hollow structure and high aspect ratio, they tend to undergo buckling when placed under compresive, torsional or bending stres (Yao et al. 1998). The electric current carrying capacity is estimated to be around 10 9 amp/cm 2 (Song et al. 2005). The thermal conductivity of SWCNT is predicted to be 6000W/mK at room temperature (Penumadu et al. 2003). Additionaly, they are stable up to 2800 ?C in vacuum and 750 ?C in air. (Thostenson et al. 2005) 1.1.2.1 Synthesis Nanocomposites found great atention in the research and industry areas; however, they are stil not wel-industrialized materials. The main reason is that the synthesis of carbon nanotubes and purification proceses increased the cost of these materials. In order to reduce costs asociated with the production of these nanofilers, diferent fabrication methods have been explored. Some of these methods are arc discharge, laser ablation and chemical vapor deposition. The arc method was first used for the production of C60 fullerenes, and single and multi- waled nanotubes are also produced by this proces. A high current electric arc pases through graphite electrodes separated by approximately 1mm in the presence of metal catalyst particles. This is the easiest technique to produce CNTs; however, the tubes have a high amount of impurities which requires further purification steps. Additionaly, the sample yield achieved with this method is not high, 30 percent by weight. . It is possible to produce nanotubes with lengths of up to 50 micrometers with few structural defects (Wilson et al., 2002). 10! ! Dr. Richard Smaley and co-workers at Rice University were the first to synthesize CNTs using a laser ablation method (Kroto et al. 1985). They used composite graphite and metal catalyst particles to create multi-wall carbon nanotubes and single-wal carbon nanotubes. The laser ablation method yields around 70% and produces primarily single-waled carbon nanotubes with a controllable diameter determined by the reaction temperature. However, it is more expensive than either arc discharge or chemical vapor deposition. Laser vaporization of graphite rods with a 50:50 catalyst mixture of Cobalt and Nickel at 1200?C in flowing argon is usualy used to prepare CNTs. This proces is followed by a heat treatment in a vacuum at 1000?C to remove fullerenes and impurities. Arc-discharge and laser vaporization are currently used for obtaining smal quantities of high quality CNTs. However, both methods suffer from industrial level production and also high amounts of impurities. Extra purification proceses are required, which increases the asociated costs of these techniques. Carbon vapor deposition method (CVD) has been used for producing carbon fibers and filaments for over 20 years. CVD is more suitable for controlling the length and structure of produced nanotubes and is also conducive for achieving high quantities, which make nanotubes available for industrial fabrication. In CVD technique, a gaseous carbon source such as hydrocarbon is decomposed catalyticaly, and the nanotubes are deposited on a substrate or grow from a substrate which is prepared with a layer of metal catalyst particles, most commonly nickel, cobalt, iron, or a combination of them. The diameters of the nanotubes are related to the size of the metal catalyst particles that are used. The diameter of the carbon nanotubes can be controlled by paterned (or masked) deposition of the metal, annealing, or by plasma etching of a metal layer. The substrate is heated to approximately 700?C. To initiate the growth of 11! ! nanotubes, two gases are bled into the reactor: a proces gas (such as amonia, nitrogen or hydrogen) and a hydrocarbon source. Nanotubes grow at the sites of the metal catalyst; the hydrocarbon gas source is broken apart at the surface of the catalyst particle, and the carbon is transported to the edges of the particle, where it forms the nanotubes. This proces requires a lower preparation temperature and simpler equipment (Poulsen et al., 2000). Using this methodology, wel-aligned or entangled SWCNTs and MWCNTs can be produced in high quantities. 1.1.3 CHALLENGES Recently, extensive research has been carried out on the incorporation of diferent types of carbon nanotubes as reinforcements into polymeric materials to produce high performance composites with high mechanical strength, electrical and thermal conductivity (Cui et al., 2003). However, there are some chalenges to create high performance carbon nanotube composites. Due to the sp 2 bonds and the high intermolecular atraction betwen the bundles, the dispersion of these filers in the host matrix and organic solvents is usualy an isue. CNTs tend to agglomerate and are dificult to disperse in the host polymer. In most of the early work, untreated CNTs or only purified CNTs were used and no significant mechanical improvement was observed. At present, it is wel recognized that in order to maximize the final performance of materials, a uniform dispersion of CNTs into the polymer matrix and good interfacial adhesion are two critical isues in the fabrication of CNT-polymer nanocomposites. The non- covalent functionalization approach of CNTs has been demonstrated to be capable of improving the dispersion of CNTs in epoxy resin. On the other hand, the covalent bonding betwen CNTs 12! ! and polymers appear to be a very useful technique in order to achieve the maximum interfacial shear strength (Cheng et al., 2007). Many of the applications of CNT composites are limited by the current cost of CNTs. While production costs are expected to fal in the future with increasing demand, costs asociated with CNT procesing such as purification, dispersing and functionalization wil remain a substantial part of the end cost of CNTs nanocomposites. 1.1.4 DISPERSION The key chalenge for nanocomposite materials is to obtain a homogenous dispersion of nanotubes in the polymer matrix. Because of their smal size, the atractive interactions betwen nanotubes are very strong and this strong interaction causes agglomeration of CNTs. Especialy, SWCNTs aggregate to form bundles that are very dificult to disperse and these bundles are entangled with one another. Several methods have been developed to prepare uniformly dispersed nanotube reinforced composites. These include the use of surfactants (Gong et al., 2000, Vaisman et al., 2006), high shear mixing (Gojny et al., 2004, Andrews et al., 2002), sonication (Qian et al., 2000), in situ polymerization of monomers in the presence of carbon nanotubes (Velasco-Santos et al., 2003), and chemical functionalization of the outside wal of the tubes (Dyke et al., 2004, Chen et al., 2006). 13! ! 1.1.4.1 Ultra-Sonication Sonication is a simple technique that uses high-intensity acoustic energy to mix and disperse materials. During sonication, nanotube entanglements are reduced. However, this technique damages and degrades the structure of the filers which in turn could eventualy reduce the length of the nanotubes. The sonication efect has been evaluated by Suave et al. They observed that the tensile strength of CNTs systems increased after sonication at high amplitude for a short period of time (20min). They also found that the best results were observed for low power sonication (165 W) performed over a longer period of time (40min) (Suave et al., 2009). 1.1.4.2 Surfactant-assisted functionalization Surfactants are amphiphilic compounds that have a hydrophilic (polar head) and a hydrophobic (tail group) parts that are jointed by a short hydrocarbon chains. They can be clasified as cationic, anionic, nonionic or zwiterionic (Geng et al., 2008). Surfactants have been extensively used in procesing stable colloidal dispersions since they are adsorbed at the interface of super-molecular structures. Chemical characteristics of the particles, surfactant molecules and the type of the solvent are key parameters for adsorption of surfactants onto inorganic and organic surfaces. The adsorption of a surfactant on the CNT?s surface lowers the surface tension of the CNT. This prevents the formation of aggregates (Figure 1.5). Furthermore, due to the presence of electrostatic repulsive forces promoted by the surfactant treatment, CNTs overcome the Van der Wals atractions which are responsible for the agglomerations (Geng et al., 2008). J. Liu et al. observed more than a 30 % increase in elastic modulus for non-ionic surfactant-asisted carbon nanotubes (Liu et al., 2000). Y. Geng et al. also observed significant improvements in 14! ! mechanical properties, thermo-mechanical properties, and electrical conductivity of MWCNT/epoxy systems (Geng et al., 2008). Figure 1.5: a) Non-ionic surfactant molecule b) CNT wrapped by surfactant molecule c) a surfactant micele d) CNT wrapped by surfactant miceles (reproduced from Geng et al. 2008) 1.1.4.3 Covalent Functionalization Covalent functionalization and surface chemistry of carbon nanotubes are important aspects for nanocomposite materials. Since carbon nanotubes have a relatively smooth graphene surface, interfacial bonding betwen a polymer matrix and carbon nanotubes is weak. As previously mentioned, nanotubes are typicaly held together as bundles in a polymer matrix which causes poor nanotube dispersion. The covalent functionalization of carbon nanotubes is expected to aid in dispersion, which improves the final properties of nanocomposites. However, 15! ! during the functionalization proces, the interaction betwen CNTs and the organic chemical used for the grafting proces causes some defects on the nanotube surface. Chemical functionalization methods are not atractive for applications requiring high conductivity since defects on the nanotube surface decrease the electrical conductivity and tensile strength of nanotubes (Gojny et al., 2004). 1.1.5 APLICATIONS Nanocomposites present many benefits such as increased mechanical properties, high thermal and chemical stability, etc. However, the usage of nanocomposites in industry is not as high as expected, but the number of application areas has been growing in recent years. There are possible applications such as aerospace structural panels, sporting goods, and ultra- lightweight thin-waled space structures (Malick 1993, Rao et al., 2005, Thostenson et al., 2005). The automotive sector is one of the leading application areas for nanocomposite materials. Nanocomposites may be potentialy utilized as mirror housings on various vehicle types, door handles, engine covers and intake manifolds and timing belt covers. Additionaly, safety isues and environmental policies push manufacturers to find new materials that met expectations. CNT-based nanocomposites are now available for diferent car models. For example Audi, Evobus and Diamler Chrysler automobiles have nano-coatings on their car?s parts. Moreover, fuel cels, porous filters and energy conversion components are other popular applications. Additionaly, light weight bodies made of polymer based nanocomposites are another significant area for the car industry. 16! ! Other application areas are microelectronics, dental applications, optical devices, composite membranes for gas separation applications, abrasion resistant coatings, beverage container applications, power tool housing, mobile phones, food packaging applications, etc. (Camargo et al., 2009). 17! ! 1.2 RESEARCH OBJECTIVES There are two main objectives asociated with this research: The first objective of this work is to understand and evaluate the problems asociated with the dispersion of carbon nanotubes in polymeric systems. Diferent dispersion techniques such as sonication, surfactants and covalent functionalization wil be evaluated and the quality of the sample dispersion wil be asesed by the constancy of the storage modulus with frequency in the low-frequency range. The second objective is to obtain a basic understanding of the thermal, mechanical and morphological properties of single wal carbon nanotubes (SWCNT)/epoxy composites. 18! ! REFERENCES 1. Ajayan PM, Schadler LS, Giannaris C, Rubio A., Single-Waled Carbon Nanotube-Polymer Composites: Strength and Weaknes, Adv. Mater.12, No. 10, 2000 2. Carlos Velasco-Santos, Ana L. Mart??nez-Herna?ndez, Frank T. Fisher, Rodney Ruoff, and Victor M. Castan, Improvement of Thermal and Mechanical Properties of Carbon Nanotube Composites through Chemical Functionalization, Chem. Mater., Vol. 15, No. 23, 2003 3. Chang Dae Han, Rheology in polymer procesing, Academic Pres, New York,1976 4. Christopher A. Dyke, James M. Tour, Covalent functionalization of Single-waled carbon nanotubes for materials applications, J. Phys. Chem. A. Vol, 108 No. 51, 2004 5. CNT Rao, FRS and A. Govindaraj, Nanotubes and Nanowires, RSC publishing, 2005 6. D. Qian, E. C. Dickey, R. Andrews and T. Rantel, Load transfer and deformation mechanisms in carbon nanotube-polystyrene composites, Appl. Phys. Let., Vol. 76, No. 20, 2000 7. Dalton AB, Coolins S, Munoz E, Razal JM, Ebron VH, Ferraris JP. Super tough carbon ? nanotube fibers; these extraordinary composite fibers can e woven into electronic textiles, Nature 423(6941):703-703, 2003 8. E.W. Wong, P.E. Sheehan and C.M. Lieber, Nanobeam mechanics: elasticity, strength, and toughnes of nanorods and nanotubes, Science 277, 1971?75, 1997 9. Erik T. Thostenson, Zhifeng Ren, Tsu-wei Chou, Advance in the science and technology of carbon nanotubes and their composites: a review, Composites Science and Technology 61,1899?1912, 2001 10. F.H. Gojny, M.H.G. Wichmann , U. Kopke, B. Fiedler , K. Schulte, Carbon nanotube- reinforced epoxy-composites: enhanced stifnes and fracture toughnes at low nanotube content, Composites Science and Technology 64, 2363?2371, 2004 11. Fangming Du, Robert C. Scogna, Wei Zhou, Stijn Brand, John E. Fischer and Karen I. Winey, Nanotube networks in polymer nanocomposites: Rheology and electrical conductivity, Macromolecules, 37, 9048-9055, 2004 19! ! 12. Fischer H. Polymer nanocomposites: from fundamental research to specific applications, Materials science and engineering; C.8 (1):31-37,2003 13. Fischer H., Polymer nanocomposites: from fundamental research to specific applications, Materials science and engineering; C 23 (6-8),763-772,2003 14. Fuyong Cheng, Dr., Alex Adronov, Noncovalent functionalization and solubilization of carbon nanotubes by using a conjugated Zn-Porphyrin polymer, Chem. Eur. J. 12, 5053? 5059, 2006 15. H.W Kroto, J.R Heath, S.C O?Brien, R.F Curl and R.E Smaley, Nature 318,162, 1985 16. H.W. Goh, S.H. Goh, G.Q. Xu, K.P. Pramoda, W.D. Zhang, Dynamic mechanical behavior of in situ functionalized multi-waled carbon nanotube/phenoxy resin composite, Chemical Physics Leters 373,277?283, 2003 17. J. Liu, A. G. Rinzler, H. Dai, J. H. Hafner, R. K. Bradley, P. J. Boul, A. Lu, T. Iverson, K. Shelimov, C. B. Huffman, F. Rodriguez-Macias, Y. Shon, T. R. Le, D. T. Colbert, and R. E. Smaley, Fullerene Pipes, Science, 280, 1253, 1998 18. J.P. Salvetat, G.A.D. Briggs, J.M. Bonard, R.R. Bacsa, A.J. Kulik and T. Stockli et al., Elastic and shear moduli of single-waled carbon nanotube ropes, Phys Rev Let 82 (5), 944? 947,1999 19. Jaqueline Suave, Luiz A.F. Coelho, Sandro C. Amico, Sergio H. Pezin, Efect of sonication on thermo-mechanical properties of epoxy nanocomposites with carboxylated-SWNT, Materials Science and Engineering A 509,57-62, 2009 20. Jian Ping Lu, Elastic properties of carbon nanotubes and nanoropes, Phys. Rev. Let.,79, 1297 ? 1300,1997 21. Linda Vaisman, H. Daniel Wagner, Gad Marom, The role of surfactants in dispersion of carbon nanotubes, Advances in Colloid and Interface Science, 128-130,37-46,2006 22. Liu J. Gao Y, Wang F, Li D, Xu J, Preparation and characteristic of a new clas of silica/polyimide nanocomposites, Journal of materials science, 37 (14), 3085-3088,2002 23. Lu JP. Elastic properties of single and multilayered nanotubes, J Phys Chem Solids 58 (11);1649-52, 1997 24. Luqi Liu, H. Daniel Wagner, Rubbery and glasy epoxy resins reinforced with carbon nanotubes, Composite science and technology 65,1861-1868, 2005 25. M Wilson, Nanotechnology: Basic Science and Emerging Technologies, 2002 20! ! 26. M. Yu, O. Lourie, M.J. Dyer, T.F. Kely and R.S. Ruoff, Strength and breaking mechanism of multiwaled carbon nanotubes under tensile load, Science 287, pp. 637?640, 2000 27. Malick, P. K. Fiber-reinforced composites, Marcel Dekker, New York, 1993 28. Mohamad Moniruzzaman, Karen I Winey, Polymer nanocomposites containing carbon nanotubes, Macromolecules, Vol. 39, No. 16, 2006 29. Nan Yao, Vincenzo Lordi, Young's modulus of single-waled carbon nanotubes, J. Appl. Phys. 84, 1939, 1998 30. O. Meincke, D. Kaempfer, H. Weickmann, C. Friedrich, M. Vathauer, H. Warth, Mechanical properties and electrical conductivity of carbon-nanotube filed polyamide-6 and its blends with acrylonitrile/butadiene/styrene, Polymer 45, 739, 2004 31. Overney G, Zhong W, Tomanek D., Structural rigidity and low frequency vibrational modes of long carbon tubules. Phys At Mol Clusters 27 (11): 1649-52,1997 32. Payodhar Padhi, S. C. Panigrahi, and Sudipto Ghosh, A new method for preparation of metal matrix nanocomposites,!AIP Conf. Proc.Volume 1063,371-375, 2008 33. Pedro Henrique Cury Camargo, Kestur Gundappa Satyanarayana, Fernando Wypych, Nanocomposites: Synthesis, structure, properties and new application opportunities, Materials research vol 12 No: 1, 1-39, 2009 34. Rakesh K. Gupta, Polymer and composite rheology,2 nd edition, Marcel Dekker INC., New York, 2000 35. Richard C. Hoagland, http:/ww.enterprisemision.com/moon5.htm, 2005 36. Rodney Andrews, David Jacques, Mickael Minot, Terry Rantel, Fabrication of Carbon Multiwal Nanotube/Polymer Composites by Shear Mixing, Macromol. Mater. Eng. 287, No. 6, 2002 37. Roslaniec Z, Broza G, Schulte K., Nanocomposites based on multiblock polyester elastomers (PE) and carbon nanotubes (CNT), Composite interfaces, 10 (1):95-102, 2003 38. Rumiana Kotsilkova, Thermoset nanocomposites for engineering applications, Smithers Rapra Tech., 2007 39. Rupesh Khare, Suryasarathi Bose,Carbon Nanotube Based Composites- A Review, Journal of Minerals & Materials Characterization & Engineering, Vol. 4, No.1, 31-46, 2005 40. S. Cui, R. Canet, A. Derra, M. Couzi and P. Delhaes, Characterization of multiwal carbon nanotubes and influence of surfactant in the nanocomposite procesing, Carbon volume 41, Issue 4, 797-809, 2003 21! ! 41. S. Iijima, Helical microtubules of graphitic carbon, Nature, 354, 56-8, 1991 42. Sad A. Khan, Joseph R. Royer, Srinivasa R. Raghavan, Rheology: Tools and methods, National academy of sciences, 2000 43. Sandler J, Shafer MSP, Prase T, Bauhofer W, Schulte K, Windle AH, Development of a dispersion proces for carbon nanotubes in an epoxy matrix and the resulting electrical properties, Polymer 40 (21), 5967-5971,1999 44. Schadler LS, Giannaris SC, Ajayan PM, Load transfer in carbon nanotube epoxy composites, Applied Physics Leters 73, 3842-3844,1998 45. Sternitzke M., Review: structural ceramic nanocomposites, Journal of European Ceramic Society, 17 (9);1061-1082,1997 46. Thostenson ET, Ren Z, Chou TW, Advances in the science and technology of carbon nanotubes and their composites: a review, Composite science and technology 61 (13); 1899- 1912, 2001 47. Tjong SC, Wang GS, High-cycle fatigue properties of Al-based composites reinforced with in situ TiB 2 and Al 2 O 3 particulates, Materials Science and Engineering:A , 386 (1-2):1887- 81, 2000 48. Treacy MMJ, Ebbesen TW, Gibson JM. Exceptionaly high Young?s modulus observed for individual carbon nanotubes, Nature 381 (6584), 678-80, 1996 49. Vaia RA, Giannelis EP, Latice of polymer melt intercalation in organicaly modified layered silicates, Macromolecules 30 (25), 7990-7999, 1997 50. Watkins JJ, Mcarthy TJ, Polymer/metal nanocomposite synthesis in supercritical CO 2 , Chemistry of materials 7 (11), 1991-1994, 1995 51. Wei Chen, Maria L. Auad, Roberto J. J. Wiliams, Steven R. Nutt, Improving the dispersion and flexural strength of multiwaled carbon nanotubes-stif epoxy composites through %- hydroxyester surface functionalization coupled with the anionic homopolymerization of the epoxy matrix, European polymer journal 42, 2765-72, 2006 52. Xiaoyi Gong, Jun Liu, Suresh Baskaran, Roger D. Voise, James. S Young, Surfactant- Asisted Procesing of Carbon Nanotube/ Polymer Composites, Chem. Mater.,12 (4), 1049- 1052, 2000 53. Yan Geng, Ming Yang Liu, Jing Li, Xiao Mei Shi, Jang Kyo Kim, Efects of surfactant treatment on mechanical and electrical properties of CNT/epoxy nanocomposites, Composites: Part A 39, 1876-83, 2008 22! ! 54. Yao KJ, Song M, Hourston DJ, Luo DZ. Polymer/layered clay nanocomposites: 2- polyurethane nanocomposites, Polymer 43(3), 2685-2688, 2002 55. Zhihang Fan, Suresh G. Advani, Rheology of multiwal carbon nanotube suspensions, J.Rheol. Volume 51, Issue 4, 585-604, 2007 23! ! CHAPTER TWO DISPERSION OF SINGLE WALL CARBON NANOTUBES EPOXY SUSPENSION 2.1 INTRODUCTION Carbon nanotubes (CNTs) as an important component in polymer nanocomposites have atracted considerable interest due to their unusual electrical, thermal and mechanical properties (Alaoui et al. 2002). Recent works have demonstrated a remarkable enhancement in the heat transfer eficiency of CNT-polymer nanocomposites (Thostenson et al. 2005) and succes in mechanical reinforcement of polymers. In the early work, untreated CNTs or only purified CNTs were used and no significant mechanical improvement was observed, however at present, it is wel recognized that to maximize the final performance of materials, uniform dispersion of CNTs in polymers and good interfacial adhesion are two critical isues in the fabrication of CNT- polymer nanocomposites (Uchida et al. 2005). Pristine CNTs are generaly incompatible with organic molecules including polymers and their extremely high aspect ratio (up to 1000) makes them easy to entangle with one another and create filer aggregates in suspensions due to strong Van der Wals atractions, which cause considerable dificulties in preparing CNT-nanocomposites. Advani et al. defined ?a good dispersion? as uniformly distributed CNT aggregates in the suspending fluid (Fan et al. 2007). Thus, diferent procedures such as high shear mixing, sonication, surfactants and CNTs chemical functionalization have been used to aid in dispersion. The non-covalent functionalization of 24! ! CNTs has been demonstrated to be capable of improving the dispersion of CNTs in epoxy resin (Zhu et al. 2003) however, the covalent bonding betwen CNTs and polymers appear to be preferable in order to achieve the maximum interfacial shear stres (Suhr et al. 2005, Zinoviev et al. 1994). In this chapter, sonication, surfactant-asisted dispersion and acid functionalization, have been evaluated. The quality of the dispersion was evaluated via diferent characterization techniques such as rheology and scanning electron microscopy. 2.2 MATERIALS SWCNTs were purchased from Shenzen Nanotech Port. Co. Ltd., China. The diameter range for the SWCNTs was < 2 nm; the length range was 5-15 ?m; the purity of CNTs was > 90 %. A diglycidyle ether of bisphenol A epoxy (DGEBA, EPON 828 Miler-Stephenson Chemical Co. Inc.) was used as a matrix. The sample was dried in a vacuum oven before utilization. Non-ionic surfactant, polyoxyethylene octyl phenyl ether (Triton X-100) was used as a surfactant agent. Tetrahydrofuran (THF), sulfuric acid (H 2 SO 4 ), nitric acid (HNO 3 ) and hydrochloric acid (HCl) were purchased from Fisher Scientific. Al the materials were used as received. 25! ! 2.3 TECHNIQUES 2.3.1 Rheology Rheology, defined as the study of deformation and flow characteristic of materials, is an effective method to understand the micro-structural changes and the filer/matrix interaction in a filer reinforced polymeric system. Rheology is a useful tool for studying viscoelastic materials, which show viscous and elastic properties (Han et al. 2007). These properties of viscoelastic materials can be explained with complex modulus which (G* = G?+iG??, (i =!") ). G? is the elastic modulus or the storage modulus that provides information about the elastic nature of the materials and shows the deformational energy stored in the system. On the other hand, G?? is the viscous modulus or the loss modulus characterizes the viscous nature of the materials and shows the disipated energy during the viscous deformation of the system (Gupta et al. 2000). The rheological properties of the epoxy liquid suspensions were determined using a commercial strain controlled rheometer (AR-G2, TA Instruments). A cone and plate geometry (diameter 60 mm and 27 &m truncation) was used to register the viscoelastic response of the CNTs/epoxy suspensions. This geometry offers several advantages, such as smaler amount of sample, easier loading and cleaning condition and a homogenous deformation throughout the sample (Khan et al. 2000). Viscoelastic properties of al samples were measured at 25?C temperature. Dynamic strain sweps at a constant frequency (1Hz) were used to find the linear viscoelastic region in which G? and G?? were independent of strain amplitude. For al samples, 1% strain was chosen in the linear viscoelastic region. A linear viscoelastic response characterized the entire range of frequency amplitudes used in this work (100 to 0.01 rad/s). Five 26! ! minute rest time was used to eliminate the efect of loading history. For the orientation of CNTs in the epoxy suspension, 1/20 s shear rate was applied for 30 minutes. 2.3.2 Scanning Electron Microscopy Scanning electron microscopy (SEM, Cambridge 360) was used to observe dispersed samples deposited in carbon grids. Samples were sputter-coated with gold prior to SEM observation. 2.3.3 Thermal Gravimetric Analysis (TGA) Thermal gravimetric analysis (TGA Q500, TA Instruments) was employed to quantify the weight percentage (wt %) of the organic component in functionalized-CNTs. Samples were heated to 800 ? C at 10 ? C /min under nitrogen. Preparation of SWCNT Epoxy suspensions The SWCNT powder was added directly into the solvent (THF) and mixed for one wek. After one wek of mixing, epoxy was added into the solution and mixed for 2 hours. The solvent was evaporated using a vacuum oven at 80?C. Al experiments were made under room temperature. Oxidation of SWCNTs In this study, CNTs were oxidized based on a proces developed by Liu et al. (61). Acording to the procedure (Figure 2.1), 1 g of SWCNT was dispersed in 250 mL of 3:1 H 2 SO 4 (98 %) / HNO 3 (70 %). This solution was sonicated for 1 h followed by 3 h of stirring at room temperature. These conditions were selected to achieve moderate oxidation and maintain a higher aspect ratio. After exhaustive washes with deionized water, HCl was added to facilitate 27! ! the formation of carboxylic acid groups ?COOH (Figure 2.1). The solution was washed again with deionized water until a constant pH value around 5-6 was achieved. The acid-treated CNTs were obtained after centrifugation of the aqueous suspensions and dried overnight in a vacuum oven at 90?C. Figure 2.1: Oxidation proces of SWCNT TGA was employed to quantify the weight percent (wt %) of the organic group in the functionalized CNTs. This technique gives a basic idea about the mas fraction of organic modifier that can be atached to the CNTs. The TGA results for pure and functionalized SWCNTs are shown in Figure 2.2. Pure CNTs mas loss (1.7 wt %) indicates the presence of amorphous carbon and impurities (Figure 2.2). The mas loss of organic compounds was observed at 600?C. The shape of the curves for the acidified SWCNT in Figure 2.2 shows that there was a fast weight loss proces ending at about 300?C, followed by a slowly and continuous weight loss proces that is active in the whole temperature range. Most of the graft COOH groups in the SWCNT were lost at about 300?C and this was followed by a degradation of CNT promoted by the previous acidification treatment. The mas fraction of the organic groups eliminated at 600?C was calculated. The amount of organic acid groups (-COOH) atached to the structure of the filers is 8.6 wt%. 28! ! Figure 2.2: TGA result for pristine SWCNT and SWCNT-COOH Scanning Electron Microscopy (SEM) was used to observe the morphology of SWCNTs before and after acidification. While the pristine SWCNT shows entangled structure and bundles, the oxidized SWCNTs (SWCNT-COOH) show les bundles and a les entangled structure (Figure 2.3). Figure 2.3: SEM images of a) Pristine SWCNT b) SWCNT-COOH 29! ! 2.4 RESULTS and DISCUSION 2.4.1 CNT/Matrix Dispersion The final properties of polymer nanocomposites strongly depend on the dispersion of nanoparticles and the interaction betwen nanoparticles and the polymer matrix. Because of carbon nanotubes nanometric size and strong Van der Wals atraction betwen the tubes, eforts to enhance the mechanical properties of CNT-polymer nanocomposites have met with only modest succes (Dyke et al. 2004). Therefore, CNTs stay together in the matrix which causes agglomeration. This agglomeration weakens structural integrity and the physical and mechanical properties of the composite. To achieve a beter dispersion, diferent methods were used in this work, such as mechanical dispersion methods (ultra-sonication and mixing) and surface modification methods which include oxidation of CNTs and non-ionic surfactant. During sonication, the agglomerated CNTs are mechanicaly separated. The use of surfactants lowers the surface tension of CNTs and this prevents the formation of aggregates (Geng et al. 2008). For both covalent functionalization and non-covalent modification, electrostatic repulsive forces are promoted. These repulsive forces help CNTs to overcome the Van der Wals atraction which is the mean reason for CNTs segregation and aggregation. In order to obtain beter understanding of structural characteristics of matrix/filer interaction such as the dispersion quality and filer alignment, rheological analysis is an efective tool. In this analysis, the rheological response of epoxy matrix without a curing agent can be characterized as a Newtonian liquid, thus, al of the viscoelastic and non-linear behaviors are caused by the addition of SWCNTs. 30! ! Effect of Dispersion Quality The quality of dispersion determines how uniformly nanotubes are distributed in the suspension. A good dispersion of the tubes is particularly important for final properties. On the other hand, bad dispersion causes a phase separation due to the strong surface interactions betwen the tubes. Carbon nanotubes are present in the matrix in two forms: aggregated and separated nanotubes. Figure 2.4 shows a cartoon of the CNTs/epoxy suspension before and after an efective dispersion technique. Before an eficient dispersion technique is applied, the SWCNTs stay agglomerated forming bundles that act as spherical particles of considerable size and very low aspect ratio (micron size). The CNT network structure contains micro CNT aggregates connected by separated CNTs. After an efective dispersion technique, the filers become beter dispersed in the matrix. The entire tube interfacial area becomes efective transferring load betwen the filer and the matrix. Both of these CNT network structures exhibit distinct viscoelastic behavior, which can be observed as a Newtonian plateau of the storage modulus (G?) developed at low frequencies during frequency swep tests and considerable higher values of G? for the wel dispersed sample. Figure 2.4: Efect of the dispersion quality on the SWCNT suspension 31! ! Effect of Functionalization One way to acomplish uniform dispersion of nanotubes is to chemicaly modify the surface of the CNTs. Functionalized carbon nanotubes have the potential to form stable and uniform dispersions in organic solvents. Atached organic groups on the CNTs surface help the CNTs separate from each other and therefore improve the dispersion (Figure 2.5). Improved dispersion of CNTs and interfacial adhesion betwen CNTs and polymer matrices has been extensible reported in the literature (Zhu et al 2003, Koratkar et al. 2002). Figure 2.5: Efect of functionalization on the dispersion quality Effect of Filer Concentration The concentration of carbon nanotubes in a nanocomposite system is an important parameter which afects the procesing and final properties of material. The quality and the degree of dispersion are also influenced by the concentration of the filers. As concentration increases, the distance betwen SWCNTs wil be smaler and there wil be stronger entanglements and interaction betwen them. In general, two types of interactions are observed in nanocomposite systems which are filer-filer interactions and polymer-filer interaction (Kotsilkove et al. 2007). 32! ! At relatively low nanotube contents, clusters and individual flocks occur due to the ability of nanotubes to difuse and stick together. Such structure can be defined as the first rheological threshold (flocculation) and the system behaves like viscous liquids with non-terminal zone behavior of viscoelasticity where G?G?). Percolation concentration is simply afected by the quality of the dispersion, the size of the filers (aspect ratio) and the degree of alignment of the components. These ideas from the percolation theory are commonly applied to the properties of suspensions and composites of impenetrable particles. Equation 1 has been widely used in the literature to model the behavior of such systems (Du et al. 2004). Equation 1 Where, ! is the volume fraction of the particles in the suspension, " is the critical exponent, P is a property (i.e. electrical and thermal conductivity, shear modulus, dielectric constant, transport properties, such as the viscosity of fluid suspensions of rigid particles, etc) and ! * is the threshold concentration. 33! ! Figure 2.6: Efect of the SWCNT concentration on the dispersion quality Effect of Aspect Ratio The aspect ratio of CNTs is an important parameter which afects the rheological behavior of the suspension. High aspect ratio of CNTs is expected to increase the viscosity since the interaction betwen nanotubes increased. During the sonication proces, nanotubes are fractionated, and the aspect ratio decreases (Figure 2.7). In order to evaluate the efect of CNTs aspect ratio, 2, 5, 10 and 30 min sonication times were used. SWCNTs were sonicated after dispersed in the solvent (THF). Figure 2.7: Efect of the aspect ratio on the dispersion quality 34! ! Effect of Orientation The last approach of this chapter is the efect of the orientation of carbon nanotubes on the suspension behavior. In order to observe the orientation efect, CNTs/epoxy suspensions were pre-sheared. This pre-shear proces orients the nanotubes in the shear direction (Figure 2.8). It is expected that oriented filers have a lower viscosity than randomly dispersed suspension. This alignment usualy aids to increase electrical conductivity or mechanical properties in the shear direction. Figure 2.8: Efect of the pre-shear on the dispersion quality 2.4.2 Experimental Results 2.4.2.1 Effect of Dispersion Quality Figure 2.9 shows the change of G? vs. frequency for the 1 wt% of pristine, surfactant asisted, and surfactant-sonication asisted SWCNTs/epoxy suspensions. This figure shows how a beter dispersion of SWCNTs produces considerable changes in the storage modulus particularly at lower frequencies (terminal regime). The sonicated-surfactant asisted SWCNT suspension exhibited the higher dynamic and loss modulus. Table 2.1 shows the storage modulus (G?) values at low frequency (' = 0.01 s -1 ). The lowest G? value was observed for pristine SWCNT, and the highest one was observed for the sonicated-surfactant asisted SWCNT. 35! ! Surfactants atached onto the SWCNTs surface providing repulsion forces which separated the SWCNT in the suspension. It also needs to be mentioned that using surfactants helps in avoiding the harmful efect of strong acids and long sonication time. It is also observed that sonication- surfactant asisted SWCNT behaved like a solid compared to the other dispersions, showing higher plateau at lower frequencies. Figure 2.9: Change of G? (Pa) with frequency for diferent SWCNT-epoxy suspensions Table 2.1: G? values at low angular frequency (' = 0.01 s -1 ) The Type of the filer G? (Pa) Pure SWCNT 0.3864 Surfactant-Asisted SWCNT 1.755 Sonication+Surfactant-Asisted SWCNT 38.86 36! ! 2.4.2.2 Effect of Functionalization Figure 2.10 shows the efect of the functionalization on the storage modulus (G?). Acidification increased the G? since the organic acid groups atached onto the SWCNT surface provide a repulsion force betwen the CNTs which separate the SWCNT in the suspension. Table 2.2 shows the G? values at low frequency and it indicates that the acidified SWCNT/epoxy suspension has higher G? value than pure SWCNT, however this plateau is lower than surfactant- sonicated SWCNT/epoxy suspension. Figure 2.10: Change of G? (Pa) with frequency for pure, acidified and surfactant plus sonicated SWCNT-epoxy suspensions (1wt %). 37! ! Table 2.2: G? values at low angular frequency (' = 0.01 s -1 ) The Type of the filer (1 wt%) G? (Pa) Pure SWCNT 0.3864 Acidified SWCNT 1.357 Sonication+Surfactant-Asisted SWCNT 38.86 2.4.2.3 Effect of Concentration Figure 2.11 shows the change of G? with the frequency at diferent concentration of SWCNT/epoxy samples prepared by the aid of surfactant. As the concentration is increased, the storage modulus increases due to the reducing mobility and the increased resistance to flow. From figure 2.11 is clear that after 0.5 wt% of SWCNTs a rubbery plateau is developed which indicates the formation of interconnecting network (change form a liquid to solid like behavior at low frequencies). The neat resin presents relaxed terminal behavior at low frequencies where G? is proportional to the square of frequency (' 2 ). The addition of smal amount of carbon nanotubes (more than 0.5wt %) makes a qualitative change in the liquid suspension and causes disappearing of the terminal behavior observed in the neat epoxy sample. At higher concentrations, the dependence of G? on frequency decreases until a plateau is clearly formed at low frequencies. The sudden change in G? indicates that the system has reached the rheological percolation. At high frequencies the efect of filers on the rheological behavior is weaker and the structures break down into individual particles. 38! ! Figure 2.12 shows the storage modulus values at 0.01s -1 frequency for diferent concentrations. As it is mentioned before, the G? increased sharply after the system reached the percolation concentration (~ 0.5 wt %). Figure 2.11: Change of G? (Pa) with frequency for diferent SWCNT concentrations 39! ! Figure 2.12: G? values at low frequency (' = 0.01 s -1 ) for diferent concentrations Beside the change of G?, complex modulus (G*) versus phase angle (() is also shown in Figure 2.13. This type of plot is caled Van Gurp-Palmen plot (Meincke et al.2004). Acording to this, the pure epoxy, which shows Newtonian liquid behavior, is expected to be completely out of phase (( = 90?). For the diferent concentration of SWCNT suspension, it is expected to observe viscoelastic behavior in which the phase angle is nearly 90? at high frequencies but decreases at lower frequencies with the concentration of filer. The Van Gurp-Palmen plot does not clearly diferentiate betwen liquid viscoelastic and solid viscoelastic behavior; however, it gives idea about the range of the concentration where the change occurs (Marcovich et al. 2006). From this picture it is observed that the system change from a liquid (( ~ 90?) to a solid (( ~ 0?) behavior at approximately 0.5 wt% of SWCNTs. 40! ! Figure 2.13: Van Gurp-Palmen plot for diferent concentration of SWCNTs The figure 2.14 shows the complex viscosity at diferent concentration of SWCNT. As the concentration increases, the mobility is reduced and the suspension?s resistance to flow is increased as wel. For higher concentrations beyond the threshold, it is observed that complex viscosity increased sharply at low frequencies. It is also found that at low concentrations below threshold, the suspension behaved more like a simple dispersion system where the nanofiler interaction can be almost neglected. 41! ! Figure 2.14: Change of complex viscosity for diferent SWCNTs concentration 2.4.2.4 Effect of Aspect Ratio and Concentration By changing the sonication time, the efect of carbon nanotubes aspect ratio is evaluated. The sonication proces aid to the dispersion of the tubes and also shortens the carbon nanotubes. In this study, surfactant asisted SWCNT/epoxy suspensions with diferent amounts of CNTs concentrations (0.25, 0.5, 1, 1.5 wt %) were sonicated for diferent periods of time (2, 5, 10, 30 min.). Acording to the Figure 2.15, the storage modulus (G?) was increased as the sonication time is increased. For the sample with 1wt% of SWCNT, at 5 min., the storage modulus reaches the highest value. Our hypothesis is that the sonication helps to separate the CNT aggregates into single tubes. It is also expected that the suspension without sonication present considerable amount of aggregates. However after 5 min. sonication, G? started to decrease. The reason is that the entanglement of CNTs was separated using longer sonication time which caused weaker interconnection betwen the carbon nanotubes and a possible decrease at the aspect ratio. The 42! ! same results were observed for the complex viscosity ()*) at the same frequency range. It means that both G? and )* are sensitive to CNT aspect ratio (Figure 2.16). Figure 2.15: Efect of the sonication time on the storage modulus (G?) of 1 wt% SWCNT surfactant-epoxy suspension Figure 2.16 Efect of the sonication time on the complex viscosity of 1 wt% SWCNT surfactant- epoxy suspension 43! ! The same procedure was followed for 0.25, 0.5 and 1.5 wt % surfactant asisted SWCNT suspensions with diferent sonication times. The results are shown in Figures 2.17, 2.18 and 2.19, respectively. It is observed that as the concentration increased, the ultimate sonication time, where the SWCNTs stil have a good interconnection, increased. It is clear that at high concentrations, the SWCNTs are closer to each other and more bundles occurred. Therefore, in order to separate the SWCNTs and build a good network take more sonication time for higher concentrations. Figure 2.17: Efect of the sonication time at 0.25 wt% surfactant-asisted SWCNT epoxy suspension 44! ! Figure 2.18: Efect of the sonication time at 0.5 wt% Surfactant-asisted SWCNT epoxy suspension Figure 2.19: Efect of the sonication time at 1.5 wt% Surfactant-asisted SWCNT epoxy suspension 45! ! In addition, Figure 2.20 shows the change in the storage modulus at low concentration with diferent sonication times for each weight fraction of SWCNTs. Figure 2.20: Storage modulus at low frequency versus sonication time for diferent concentration of surfactant asisted SWCNT-epoxy suspension Analytical Model An analytical model was used to estimate the aspect ratio of the SWCNTs in the epoxy suspension. Aspect ratio and volume fraction of non-spherical particles are considered by many simple models for describing the viscosity of particle in a suspension. Einstein was the first in develop one of the simplest model to describe this behavior (Einstein 1956). Einstein models take into acount the hydrodynamic efects of the particles on the flow field. The mathematic equation that respects this model is showed in equation 2. )/) m = 1+ 0.67* ? + 1.62 * 2 ? 2 Equation 2 0! 5! 1 0! 1 5! 2 0! 2 5! 3 0! 0! 2 0! 4 0! 6 0! 8 0! 10 0! 12 0! 1.5 wt%! 1 wt%! 0.5 wt%! 0.25 wt%! Sonication time (min) 46! ! where ) is the viscosity of the suspension, ) m is the viscosity of the matrix, * is the volume fraction of the particles, and ? is the aspect ratio of the filer (length/diameter). (Volume fraction of the SWCNT was calculated by using the nanotube density 1.1 g/ml) Based on the Figure 2.21, the experimental data and analytical model show good agreement using aspect ratio betwen 75 and 100. These values are consistent with observations of the SWCNTs after they were broken up by the 10 min sonication. These values are consistent with data reported in the literature (M.D Frogley et al. 2003) Figure 2.21: Initial dynamic viscosity of the suspension as a function of SWNCT volume fraction. (The solid curves are the analytical results with diferent aspect ratios (?).) 47! ! 2.4.2.5 Effect of Orientation The use of high shear flow aligns the nanotubes in the shear direction and this aligned suspension shows a lower viscosity than randomly orientated suspension because the resistance to flow decreases with the orientation of the filer. In addition, high shear flows may break down the 3D network formed by the filer and decrease the viscosity. In order to study the efect of the orientation, al samples were tested before and after being shear at a shear rate of 20 s -1 for 30 min. Figures 2.23, 2.24 and 2.25 show the behavior of the 1 wt% of SWCNTs/epoxy suspensions before and after shear flow. It is observed that G? decreased after the shear flow for acidified SWNCT, surfactant-asisted SWCNT and 5min sonicated surfactant-asisted SWCNT. The reason is that the shear flow provided an alignment for CNTs and this makes the interconnection weaker. The diference for sonicated sample was considerable because the sonication separates the nanotubes and the shear proces increased the orientation more than the other samples. On the other hand, for pure SWCNT, G? increased after the shear flow compared to the no shear (Figure 2.22). Since the amount of the SWCNT bundles is higher for pure SWCNT system, the shear flow proces possible helped to disentangle the flocculated bundles of nanotubes that exist for the pristine samples. 48! ! Figure 2.22: Efect of orientation on G? for pure SWCNT suspensions Figure 2.23: Efect of orientation on G? for acidified SWCNT suspensions 49! ! Figure 2.24: Efect of orientation on G? for SWCNT-Surfactant suspensions Figure 2.25: Efect of orientation on G? for 5 min sonicated SWCNT-Surfactant suspensions 50! ! 2.5 CONCLUSION In this chapter, the dispersion quality of the diferent SWCNT/epoxy suspensions was discussed. The efect of the SWCNT modification on dispersion quality was examined. The amount of the organic groups atached onto the SWCNT after the acid functionalization was observed with TGA experiment and it is found that 8.6 wt% acid groups were atached onto the SWCNT. In addition, the efect of diferent sonication time, SWCNT concentration, and nanotube orientation were studied using rheological experiments. The highest storage modulus (G?) value was observed for sonicated-surfactant asisted SWCNT suspension. It was observed that as the nanotube concentration increased, the suspension showed a noticeable transition betwen a liquid to solid like behavior. The percolated threshold concentration was estimated to be approximately at 0.5 wt% of SWCNTs, where G?, G? and viscosity increased dramaticaly. It is also observed that the G? value increased as the sonication time increased. For 1 wt% SWCNT suspension, the highest G? was found at 5 min sonication time. It was noted that for 1 wt% SWCNT after 5 min sonication time, the SWCNTs built a good 3-D network. However at longer sonication time storage modulus (G?) started to decrease. The reason is that the entanglement of CNTs were separated by the longer sonication time and causing weaker interconnection betwen the carbon nanotubes. Therefore, in order to separate the SWCNTs and build a good network longer sonication times are required for higher concentrations. The efect of the orientation was also discussed. The orientation of SWCNTs in the suspension decreased the resistance to flow and thus decreased the storage modulus were found. 51! ! REFERENCES 1. A. Einstein, Investigation on the theory of Brownian Movement, Dover, New York, 1956 2. Alaoui A, Bai S, Cheng HM, Bai JB, Mechanical and electrical properties of a MWNT/epoxy composites. Composite science and technology 62,1993-1998, 2002 3. Chang Dae Han, Rheology in polymer procesing, Academic Pres, New York, 1976 4. Christopher A. Dyke, James M. Tour, Covalent functionalization of Single-waled carbon nanotubes for materials applications, J. Phys. Chem. A. Vol, 108 No. 51, 2004 5. Erik T. Thostenson, Chunyu Li, Tsu-Wie Chou, Nanocomposites in context- Composites Science and technology 65 491-516, 2005 6. J. Suhr, W. Zhang, P.M. Ajayan and N. Koratkar, Viscoelasticity in carbon nanotube composites, Nat Mater 4, pp. 134?137, 2005 7. J. Suhr, W. Zhang, P.M. Ajayan and N.A. Koratkar, Temperature-activated interfacial friction damping in carbon nanotube polymer composites, Nano Let 6,219, 2006 8. Mark D. Frogley, Diana Ravich, H. Daniel Wagner, Mechanical properties of carbon nanoparticle-reinforced elastomers, Composite Sci. and Tech. 63,1647-1654, 2003 9. N. Koratkar, B. Wei and P.M. Ajayan, Carbon nanotubes films for damping applications, Adv Mater 14 (13) pp. 997?1000, 2002 10. N.E Marcovich, M.L Auad, N. E Belesi, S.R. Nutt, M. I Aranguren, Celulose micro/nanocrystals reinforced polyurethane, J. Mater. Res., Vol. 21, No. 4, 2006 11. P.A. Zinoviev and Y.N. Ermakov, Energy disipation in composite materials, Technomic Publishing Company, PA, 1994 12. Rakesh K. Gupta, Polymer and composite rheology,2 nd edition, Marcel Dekker INC., New York, 2000 13. Rumiana Kotsilkova, Thermoset nanocomposites for engineering applications, Smithers Rapra Tech. 2007 52! ! 14. Uchida T, Kumar S, Single wal carbon nanotube dispersion and exfoliation in polymers, journal of applied polymer science, 98,985-989, 2005! ! 15. Yan Geng, Ming Yang Liu, Jing Li, Xiao Mei Shi, Jang Kyo Kim, Efects of surfactant treatment on mechanical and electrical properties of CNT/epoxy nanocomposites, Composites: Part A 39, 1876-83, 2008 16. Zhihang Fan, Suresh G. Advani, Rheology of multiwal carbon nanotube suspensions, J.Rheol. Volume 51, Issue 4, pp. 585-604, 2007 17. Zhu J, Kim JD, Peng H, Margrave JL, Khabashesku VN, Barrera EV, Improving the dispersion and integration of single waled carbon nanotubes in epoxy composites through functionalization, Nanoleter 3:1107-13, 2003 18. Zhu J, Peng H, Rodriguez-Macias F, Margrave JL, Khabashesku VN, Imam AM. , Reinforcing epoxy polymer composites through covalent integration of functionalized nanotubes, Adv. Funct. Mater, 14;643-8, 2004 . 53! ! CHAPTER 3 MECHANICAL AND THERMAL PROPERTIES OF SWCNT/EPOXY COMPOSITES 3.1 INTRODUCTION Carbon nanotubes have been atractive for reinforcement in composite materials due to their high aspect ratio, fiber-like structure, low density, and extraordinary mechanical and barrier properties (Cantournet et al. 2007, Jonathan et al. 2006). Particularly, carbon nanotubes and carbon nanofibers have shown to have a unique combination of high modulus (640 GPa to 1 TPa) and strength (150-180 GPa) (Krishnan et al. 1998). These filers offer the possibility of enhancing thermal, electrical, damping and mechanical properties with relatively lower amounts (1-5wt %) and with minimal degradation of the original systems (Schadler et al. 1998). Zhu et al. have shown that using of 1 wt% of SWCNT the Young?s modulus of epoxies was increased up to 30% and the tensile strength up to 18% (Zhu et al 2003). They are excelent candidates to substitute or complement the conventional filers in the fabrication of multifunctional polymer composites. At present, it is wel recognized that to maximize the final performance of these materials, uniform dispersion of CNTs in polymeric system and good interfacial adhesion are two critical isues during the fabrication of CNT-polymer nanocomposites. (Song et al. 2005) 54! ! Dispersion is the key isue in nanocomposite fabrication. The high surface area of nanotubes is an advantage for load transfer; however it also induces strong atractive forces betwen the tubes that cause agglomeration which causes bad dispersion (Penumadu et al. 2003). Various dispersion methods (sonication, intensive shear mixing, aid of surfactants, and covalent functionalization) were reported to provide beter dispersion and separate the agglomerated bundles into individual tubes. Another chalenge for SWCNT/polymer composites is the interfacial adhesion betwen the CNTs and the matrix. In order to provide an eficient structural reinforcement with CNTs, an efective stres transfer from matrix to the tubes is required. The interfacial bonding betwen CNTs and matrix can be improved by chemical functionalization of CNTs. The covalent bond betwen the polymer and the nanotubes improves the interfacial stres transfer and also provides beter dispersion as mentioned before. The aim of this chapter is to compare the influence of diferent dispersion techniques and diferent concentrations of SWCNTs on the mechanical and thermo-mechanical properties of epoxy-based nanocomposites. Sonication plus surfactant modification and acid functionalization were used in order to achieve beter dispersion and improve the stres transfer mechanism betwen the epoxy matrix and SWCNTs. Simple theoretical model where use in order to study the efect of the carbon nanotubes concentration and aspect ratio on the final modulus of the composite. 55! ! 3.2 MATERIALS The SWCNTs were purchased from Shenzen Nanotech Port. Co. Ltd., China. The diameter range for the CVD grown SWCNTs was < 2 nm; the length range was 5-15 ?m; the purity of CNTs was > 90 %. A diglycidyle ether of bisphenol A epoxy (DGEBA, EPON 828 Miler-Stephenson Chemical Co. Inc.) was used as a matrix and a polyether triamine (Jefamine T-403, Huntsman) with a molecular weight of 440 was used as hardener. The cured product had a glas transition temperature located above room temperature. Non-ionic surfactant, polyoxyethylene octyl phenyl ether (Triton X-100) was used as a surfactant agent. Tetrahydrofuran (THF), sulfuric acid (H 2 SO 4 ), nitric acid (HNO 3 ) and hydrochloric acid (HCl) were purchased from Fisher Scientific. Al the materials were used as received. Composite Preparation Epoxy nanocomposites were prepared with 0.25, 0.5 and 1 wt% of surfactant asisted SWCNT; additionaly samples with 1 wt% of SWCNT-COOH and 1 wt% pure SWCNT were also prepared. The corresponding amount of modified SWCNTs was dispersed in THF by mixing one wek. After one wek, epoxy resin was added into the mixture and mixed two hours. Finaly, THF was evaporated at 50 o C in a vacuum oven overnight. The corresponding hardener, Jefamine T-403 was added to the mixture at an epoxy/hardener weight ratio of 100/42. Further mixing (5 minutes) and degasification (15 minutes) were performed. The blend was cast into Teflon mold. The curing cycle was 3 hours at 80 o C, followed by another 3 hours at 125 o C. 56! ! 3.3 TECHNIQUES 3.3.1 Flexural Experiment The mechanical properties of diferent type of SWCNT epoxy nanocomposites were evaluated. Samples were tested with 3-point bending using a dynamic mechanical analysis machine,(RSA3, TA Instruments) tool and the flexural modulus was calculated by using equation of E = ((L 3 +m) / (4+b+d 3 )) (ASTM D790M) where E is the modulus of elasticity in bending (MPa), L is the support span length which is 10mm in this study, b is the width of the beam tested (mm) and m is the slope of the tangent to the initial straight-line portion of the load- deflection curve (N/m of deflection). 3.3.2 Dynamic Mechanical Analysis (DMA) Dynamic mechanical analysis (RSA3 TA instruments) was carried out to ases the storage modulus (E?), loss modulus (E??) and tan delta () of CNTs-epoxy composites. Epoxy/T- 403 samples were performed with 3-point bending fixture. For the T-403 samples, the dynamic temperature ramp test was performed at 0.1% strain at a frequency of 1Hz from 25 ? C to160 ? C with a heating rate of 5 ? C/min. 57! ! 3.4 RESULT and DISCUSION 3.4.1 Mechanical Properties Table 3.1 summarizes the efect of the SWCNT dispersion on the flexural modulus of the epoxy nanocomposites. The neat epoxy matrix that was used in this study exhibited a flexural modulus of 2.2 GPa. It is observed that the addition of pristine SWCNT increased (15 % increase) the modulus of the composite under the given procesing conditions. Moreover, the addition of acid functionalized SWCNTs increased modulus 20 %. It is clear that the acid groups located on the surface of SWCNT aid in dispersion and the filer. In addition they could react with the epoxy matrix as reported by other authors (Chen et al. 2006) and form covalent bond with epoxy, thus, inducing the interfacial adhesion. In addition, surfactant asisted SWCNT/epoxy systems (without sonication) has flexural modulus higher than pristine SWCNT system. The surfactant provides beter dispersion of SWCNTs in the epoxy matrix which activates efective mechanisms of load transfer. It is also noted that surfactant plus sonication asisted SWCNT samples have the highest flexural modulus (48% increase compared to the matrix). Sonication proces aids to separate the bundles of SWNCT providing beter dispersion. As it is reported in the previous chapter, the surfactant plus sonication asisted SWCNT samples showed the highest storage modulus (G?) that indicates a beter degree of dispersion compared to the other systems. 58! ! Table 3.1: Properties of neat epoxy resin and nanocomposites for diferent systems of 1 wt% of SWCNTs Filler Flexural Modulus (GPa) % improvement Matrix 2.2 ? 0.289 - Pristine SWCNT 2.53 ? 0.105 15 Acidified SWCNT 2.62 ? 0.074 19 SWCNT-Surfactant 3.05 ? 0.102 39 SWCNT Surfactant- sonication 3.26 ? 0.135 48 Table 3.2 shows the efect of carbon nanotubes concentration on the mechanical properties of SWCNT epoxy system. The sample modulus increased with increasing filer content. The addition of 1 wt% SWCNT produced a 50% increase in the flexural modulus with respect to the neat system . 59! ! Table 3.2: Properties of neat epoxy resin and nanocomposites for diferent SWCNT concentration The amount of SWCNT Plus surfactant (wt %) Flexural Modulus (GPa) % Improvement 0 2.2 ? 0.289 - 0.25 2.71 ? 0.013 23 0.5 2.83 ?0.013 29 1 3.26 ? 0.135 48 Mechanical Model: One of the simplest mechanical models commonly used for filed polymer is the modified Halpin-Tsai theory (Agarwal et al.1980, Chen et al 2009)"!The Halpin?Tsai equations are a set of empirical relationships that enable the property of a composite material to be expresed in terms of the properties of the matrix and filer together with their proportions and geometry (Kim et al. 2009). For low loadings of rod-like filers which are much stifer than the matrix, it reduces to; E long = E m ( (1+2?,V f )/(1-,V f )) Equation 3.1 E trans = E m ( (1+2?-V f )/(1-V f )) Equation 3.2 , = ((E f /E m )-1)/(E f /E m +2?) Equation 3.3 - = ((E f /E m )-1)/(E f /E m +2) Equation 3.4 60! ! E c = 3/8 E long + 5/8E trans Equation 3.5 where E long and E trans are the longitudinal stifnes and transverse stifnes of the composite, respectively. V f and ? are the volume fraction and aspect ratio of the nanotubes, respectively and E c is the stifnes of the composite which fibers are randomly oriented in three dimensions. The model was developed for short fiber composites and considers that there is no interaction betwen the fibers (Frogley et al. 2002).Furthermore, Radosta suggested that this equations are versatile enough to be used also for flexural modulus (Radosta, 1975,Nielsen 1970). Al results were also compared to Reuss and Voight model (Bhagwan et al. 2006, Tibbetsa et al. 1999) where; E long = E f V f +E m V m ( Reuss model) Equation 3.6 1/E trans = V f /E f + V m /E m (Voight model) Equation 3.7 Figure 3.1 shows the experimental and analytical results for SWCNT-epoxy system with diferent SWCNT aspect ratios. Each experimental point represents the average of five test values. The predicted composite modulus based on Halpin-Tsai micromechanics were compared with the experimental values of the SWCNT-epoxy composites. For these calculations, a modulus of 1 TPa and aspect ratio of 100, 75 and 50 were asumed and the flexural modulus for epoxy matrix was taken as 2.2 GPa. Reuss and Voight models (equation 3.6 and equation 3.7 respectively) gave us the upper and lower boundaries of the mechanical properties of nanocomposites. 61! ! Figure 3.1: The comparison of experimental and analytical results for flexural modulus of SWCNT-epoxy system 3.4.2 Thermo-mechanical Properties The viscoelastic behavior of SWCNT-epoxy nanocomposites was explained with DMA technique. In DMA technique the linearized stres-strain relation can be expresed as . = (E? + jE?)/ where the E? is the in-phase component which determines the storage or elastic modulus and E?? is the out-phase component which shows the loss modulus. In this study, the efect of temperature on dynamic-mechanical properties is summarized on Table 3.3 and can be visualized in Figure 3.2 for diferent SWCNT epoxy systems. In Figure 3.2, it was observed that at the glasy state (below the glas transition of the composite), surfactant plus sonication asisted SWCNTs system has the highest storage modulus (E?). At room temperature, the storage 62! ! modulus of the matrix increased 249 % after adding surfactant plus sonication asisted SWCNTs. Moreover, adding pure SWCNTs and acidified SWCNTs increased the storage modulus of the matrix. It is also noted that the elastic properties of the epoxy system was improved at elevated temperatures. This behavior can be explained in terms of an interaction betwen the CNTs and the epoxy due to the high surface area. The interfacial interaction betwen the CNTs and epoxy reduces the mobility of the epoxy matrix around the nanotubes and leads to the observed increase in thermal stability. Figure 3.2 Storage modulus (E?) vs. temperature for diferent SWCNT-epoxy system 63! ! Figure 3.3 shows that the T g of the SWCNTs/epoxy resins nanocomposite increases 4?C and the width of tan ( peaks becomes broader, with the presence of SWCNT. The increased T g and broadened loss peak observed in polymer nanocomposites could been atributed to the constraint efect of nanoparticles on the segmental motion of polymers. Similar results were observed by Goh et al. (Goh et al. 2003). Acording to the coupling model theory suggested by Ngai et al. as the more segments are restricted due to the presence of nanofiler, the activation threshold for the motion of some segments becomes higher. (Ngai et al. 1996). As a consequence, the glas transition trends to become wider with increasing restriction efect. For pristine and acidified SWCNT systems, a decrease in T g was observed. It could be due to possible partitioning mechanism of the monomers at the interphase of the epoxy-filer. Figure 3.3 Tandelta vs. temperature for diferent SWCNT-epoxy system 64! ! Table 3.3 Properties of neat epoxy resin and nanocomposites Type of composite (1 wt%) E? (GPa) at 30? C E? (GPa) at 150? C T g (C?) Matrix 0.55 0.009 94 Pure SWCNT/Epoxy 1.75 0.028 75 Acidified SWCNT/Epoxy 1.4 0.028 71 Surfactant+ Sonication/Epoxy 1.92 0.035 97 65! ! 3.5 CONCLUSION In this chapter, the mechanical and thermo-mechanical properties of SWCNT epoxy nanocomposites were investigated. When carbon nanotubes are properly dispersed in the matrix, carbon nanotube reinforcement influenced properties of the matrix. It was observed that sonication plus surfactant asisted SWCNT epoxy system has the highest flexural modulus. It was also reported in chapter 2 that this system has beter dispersion quality compared to the other systems. Furthermore, the efect of the concentration on the flexural modulus was analyzed and it was noted that as the concentration increases, the flexural modulus of the composite system increases. The experimental results were also compared with the analytical results. It was noted that the experimental results were betwen the upper and lower boundaries results which were derived from Reuss and Voight model. In addition, the mechanical and thermo-mechanical properties of SWCNT-epoxy system were studied. It was concluded that wel dispersed SWNCTs into the epoxy system increased the storage modulus and the glas transition of the matrix. 66! ! REFERENCES 1. Agarwal, B. D., Broutman, L.G. Analysis and performance of fiber composites, Wiley, New York, 1980 #" Bhagwan D. Agarwal and Lawrence j. Broutman, Analysis and Performance of Fiber Composites, John Wiley & Sons, 2006 ! 3. Cantournet S, Boyce MC, Tsou AH. Micromechanics and macromechanics of carbon nanotube-enhanced elastomers. J Mech Phys Sol , 55, 1321-1339, 2007 4. Chen, W., Auad, M.L, Wiliams, R.J.J., Nutt, S.R., Improving the Dispersion and Flexural Strength of Multiwaled Carbon Nanotubes-Epoxy Composites through #-Hydroxyester Surface Functionalization Coupled with the Anionic Homopolymerization of the Epoxy Matrix European Polymer Journal, 42,2765?2772, 2006 5. Florian H. Gojny Malte H.G. Wichmann, Bodo Fiedler and Karl Schulte Influence of diferent carbon nanotubes on the mechanical properties of epoxy matrix composites- a comparative study- Florian H. Gojny et al. Composites Science and Technology Volume 65, Issues 15-16, 2300-2313, 2005 6. Gary G. Tibbetsa and John J. McHugh, Mechanical properties of vapor-grown carbon fiber, J. Mater. Res., Vol. 14, No. 7, 1999 7. H.W. Goh, S.H. Goh, G.Q. Xu, K.P. Pramoda, W.D. Zhang, Dynamic mechanical behavior of in situ functionalized multi-waled carbon nanotube/phenoxy resin composite, Chemical Physics Leters 373,277?283, 2003 8. J.A. Radosta, 34th SPE-Antec Conference, 1975 9. Jonathan N. Coleman, Umar Khan, Werner J. Blau, Yurii K. Gun?ko, Smal but strong: A review of the mechanical properties of carbon nanotube-polymer composites, Carbon 44 1624-1652,2006 10. K. L. Ngai, A. K. Rizos, Parameterles Explanation of the Non-Arrhenius Conductivity in Glasy Fast Ionic Conductors, Phys. Rev. Let. 76, 1296,1996 11. Krishnan A., Dujardin E., Ebbesen TW. Young modulus of single-waled nanotubes. Phys. Rev. B. 1009;58,14013-14019, 1998 67! ! 12. Lawrence E. Nielsen, Generalized equation for the elastic moduli of composite materials, Journal of App. Physics, 41,11,1970 13. Myungsoo Kim, Young-Bin Park , Okenwa I. Okoli, Chuck Zhang, Procesing, characterization, and modeling of carbon nanotube-reinforced multiscale composites, Composites Science and Technology 69, 335?342, 2009 14. S. Cui, R. Canet, A. Derra, M. Couzi and P. Delhaes, Characterization of multiwal carbon nanotubes and influence of surfactant in the nanocomposite procesing, Carbon volume 41, Issue 4, 797-809, 2003 15. Tyrone l. Vigo and Barbara J Kinzig, Composite applications the role of matrix, fiber and interface, Wiley-VCH, 1992 16. Young Seok Song, Jae Ryoun Youn, Influence of dispersion states of carbon nanotubes on physical properties of epoxy nanocomposites, Carbon 43, 1378-1385, 2005 68! ! CHAPTER 4 CONCLUSION Using a combination of techniques (microscopy, thermal gravimetric analysis, rheology and dynamic mechanical analysis), an atempt was made to characterize single wal carbon nanotube/epoxy systems. The aim of this research was to understand the efect of the dispersion quality on the final properties (mechanical and thermo-mechanical) of single wal carbon nanotube/epoxy composites. Several dispersion techniques were chosen such as oxidation of CNTs, surfactant asisted systems and ultra sonication. The quality of the dispersion was evaluated through rheological studies. The highest storage modulus (G?) value, which indicates beter dispersion, was observed for sonicated-surfactant asisted SWCNT suspension. The efect of carbon nanotubes concentration on the final dispersion was also studied. It was observed that as the carbon nanotube concentration increased, the suspension showed a noticeable transition betwen a liquid-like to solid-like behavior (percolation concentration). The percolated threshold concentration was estimated to be approximately at 0.5 wt% of SWCNTs, where the storage (G?) and loss (G?) modulus and the viscosity increased several orders of magnitude compared to the unreinforced system. It was also observed that the sonication time produced considerable efect on the final dispersion of the filers. The G? value increased as the sonication time increased indicating a 69! ! beter disentanglement of the SWCNTs bundles. However at longer sonication time storage modulus (G?) decrease causing weaker interconnection betwen the carbon nanotubes and possible degradation of the final structures. The efect of the orientation on the storage modulus was also discussed. The orientation of SWCNTs in the suspension decreased the resistance to flow and thus decreased the storage modulus of the suspensions. Finaly, the mechanical and thermo-mechanical properties of SWCNT epoxy nanocomposites were investigated. It was observed that the highest flexural modulus was obtained for sonication plus surfactant asisted SWCNT epoxy system which has beter dispersion quality compared to the other systems. Furthermore, the efect of the concentration on the flexural modulus was analyzed and it was noted that as the concentration increases, the flexural modulus of the composite system increases. The experimental results were also compared with analytical results. It was noted that the experimental results were betwen the upper and lower boundaries results which were derived from Reuss and Voight model. The thermo-mechanical properties of SWCNT-epoxy system were also studied. It was concluded that the addition of sonicated SWNCTs into the epoxy system increased the storage modulus and glas transition of the matrix. These efects were produced by the restrictions introduced by the stif tubes in the local motion of the matrix.