ENHANCEMENTS OF MECHANICAL, THERMAL STABILITY, AND TRIBOLOGICAL PROPERTIES BY ADDITION OF FUNCTIONALIZED REDUCED GRAPHENE OXIDE IN EPOXY

Transcription

1 ENHANCEMENTS OF MECHANICAL, THERMAL STABILITY, AND TRIBOLOGICAL PROPERTIES BY ADDITION OF FUNCTIONALIZED REDUCED GRAPHENE OXIDE IN EPOXY Rakesh K. Shah, M.S. Dissertation Prepared for the Degree of DOCTOR OF PHILOSPHY UNIVERSITY OF NORTH TEXAS August 2014 APPROVED: Jose Perez, Major Professor Duncan Weathers, Committee Member Usha Philipose, Committee Member Guido Verbeck, Committee Member Christopher Littler, Interim Chair of the Department of Physics Mark Wardell, Dean of the Toulouse Graduate School

2 Shah, Rakesh K. Enhancement of Mechanical, Thermal Stability, and Tribological Properties by Addition of Functionalized Reduced Graphene Oxide in Epoxy. Doctor of Philosophy (Physics), August 2014, 83 pp., 2 tables, 45 figures, 117 numbered references. The effects of octadecylamine-functionalized reduced graphene oxide (FRGO) on the frictional and wear properties of diglycidylether of bisphenol-a (DGEBA) epoxy are studied using a pin-on-disk tribometer. It was observed that the addition of FRGO significantly improves the tribological, mechanical, and thermal properties of epoxy matrix. Graphene oxide (GO) was functionalized with octadecylamine (ODA), and then reduction of oxygen-containing functional groups was carried out using hydrazine monohydrate. The Raman and x-ray photoelectron spectroscopy studies confirm significant reduction in oxygen-containing functional groups and formation of ODA functionalized reduced GO. The nanocomposites are prepared by adding 0.1, 0.2, 0.5 and 1.0 wt % of FRGO to the epoxy. The addition of FRGO increases by more than an order of magnitude the sliding distance during which the dynamic friction is 0.1. After this distance, the friction sharply increases to the range of We explain the increase in sliding distance during which the friction is low by formation of a transfer film from the nanocomposite to the counterface. The wear rates in the low and high friction regimes are approximately 1.5 x 10-4 mm 3 /N m and 5.5 x 10-4 mm 3 /N m, respectively. The nanocomposites exhibit a 74 % increase in Young s modulus with 0.5 wt. % of FRGO, and an increase in glass transition and thermal degradation temperatures.

3 Copyright 2014 by Rakesh K. Shah ii

4 ACKNOWLEDGEMENTS I would like to greatly thank my advisor Dr. Jose Perez for his inspiring way to guide me to a deeper understanding of research and his valuable comments during my study at UNT. I would like to thank Dr. Witold Brostow for providing me an opportunity to work in his lab with his worldwide colleagues. The work would not have been possible without his help and suggestions. Further, I am highly thankful to my other committee members Dr. Duncan Weathers, Dr. Usha Philipose, and Dr. Guido Verbeck for being in my dissertation committee and for their valuable comments and suggestions in preparing this dissertation. I am also thankful to Dr. Christopher Littler and Dr. David Schultz for their support whenever it was needed from the Department of Physics. I am very grateful to all my friends who helped and encouraged me to achieve success in this work. I am particularly thankful to Dr. Yudong Mo and Dr. Joshua Wahrmund for their help and time for useful discussions. I would like to express my heartfelt thanks to my former advisor Dr. Saikat Talapatra for his invaluable comments and suggestions at different points to make a successful career. I would also like to express sincere gratitude to my parents for their unconditional love and support throughout my life. Last, but certainly not least, I like to thank my wife Nitu for her continuous support and understanding during this work. iii

5 TABLE OF CONTENTS Page ACKNOWLEDGMENTS... ii LIST OF TABLES... vi LIST OF FIGURES... vii CHAPTER 1 INTRODUCTION Introduction Composite Materials Band Structure of Graphene Graphene Oxide Functionalization of Graphene and Graphene Oxide Epoxy Resin Challenges in Polymer Nanocomposite Dispersion of Nanomaterials Cost Effectiveness Health and Environmental Issues Motivation Outline of Dissertation CHAPTER 2 EXPERIMENTAL TECHNIQUES Synthesis of Graphene Oxide Functionalization of Graphene Oxide with Octadecylamine Characterization of Graphene Oxide and Functionalized Graphene Oxide Raman Spectroscopy of Graphene and Graphene Oxide iv

6 2.3.2 X-ray Photoelectron Spectroscopy (XPS) of Graphene Oxide Synthesis of Functionalized Graphene Oxide-Epoxy Nanocomposites Characterizations of Polymer Nanocomposites Mechanical Properties: Tensile Testing Mechanical Properties: Dynamic Mechanical Analysis Thermal Stability Determination: Thermogravimetric Analysis (TGA) Tribological Properties of Polymer Nanocomposites Friction Wear Friction and Wear Measurements of Polymeric Materials: Pin-on-disk Tribometer CHAPTER 3 CHARACTERIZATION OF GRAPHENE OXIDE AND FUNCTIONALIZED REDUCED GRAPHENE OXIDE Characterization of Functionalized Reduced Graphene Oxide Using SEM Characterization of Functionalized Reduced Graphene Oxide Using Raman Spectroscopy Characterization of Functionalized Reduced Graphene Oxide Using XPS CHAPTER 4 MECHANICAL, THERMAL, AND TRIBOLOGICAL PROPERTIES OF NANOCOMPOSITES Introduction Young s Modulus Measurements of FRGO-Epoxy Nanocomposites Estimation of Young s Modulus of FRGO-Epoxy Nanocomposites Using Halpin-Tsai Model v

7 4.4 Dynamic Mechanical Analysis of FRGO-Epoxy Nanocomposites Thermogravimetric Analysis of FRGO-Epoxy Nanocomposites Tribological Properties of FRGO-Epoxy Nanocomposites Friction Results Wear Results Investigation of Wear Mechanism CHAPTER 5 CONCLUSIONS REFERENCES vi

8 LIST OF TABLES Table 4.1 Mechanical properties of the neat epoxy and the nanocomposites Table 4.2 The Young s Modulus of FRGO-Epoxy Nanocomposites Calculated Using Halpin-Tsai Model Page vii

9 LIST OF FIGURES Figure 1.1 Representation of hexagonal lattice of graphene sheet. The shaded region represents the unit cell Figure 1.2 First Brillion zone in reciprocal space where, Γ, K, K, and M are high symmetry points Figure 1.3 Band structure of graphene derived from tight band approximation Figure 1.4 Chemical structure of graphene oxide with different oxygen-containing functional groups on basal plane and around edges of a graphene layer Figure 1.5 Chemical structure of diglycidylether of Bisphenol-A (DGEBA), n=0, for the derivatives, n > Figure 1.6 Schematic of various levels of dispersion of nanomaterials in a polymer matrix Figure 2.1 A layout of GO synthesis procedure Figure 2.2 Photographs during the oxidation process of graphite powder. (a) During the second stage of oxidation reaction. (b) After the first dilution with DI water Figure 2.3 Schematic of a freeze drying system Figure 2.4 (a) Schematic of GO functionalized with ODA. (b) Schematic of functionalized reduced GO. The long chains represent ODA and is attached to the basal plane of GO Figure 2.5 Schematic of the setup used to functionalize GO Figure 2.6 Vibrational modes of (a) G peak and (b) D peak (b). The black circles represent the carbon atoms and the arrows represent the direction of vibration Figure 2.7 Ejection of a photoelectron upon irradiation by a mono-energetic X-ray Figure 2.8 Photograph of the horn sonicator used for the dispersion of FRGO Figure 2.9 A schematic representation of FRGO/epoxy nanocomposite preparation Figure 2.10 Schematic of the tensile testing equipment Figure 2.11 Schematic of a tensile testing sample Page viii

10 Figure 2.12 Response of a sinusoidal force applied to a polymeric material at a certain frequency Figure 2.13 Schematic of the driving system of the DMA equipment Figure 2.14 Schematic of the single cantilever (i.e., 2-point bending mode) system of the DMA equipment Figure 2.15 Photograph of the thermogravimetric analyzer Figure 2.16 (a) Schematic illustration of a hard body sliding over a polymeric surface. (b) Enlarged view of the region of polymeric material in contact with the hard surface Figure 2.17 Schematic of the pin-on disk tribometer and a polymeric nanocomposite sample with a wear track Figure 2.18 Photograph of the pin-on disk tribometer used for friction and wear measurements Figure 3.1 (a-c) SEM of FRGO powder at different magnifications Figure 3.2 Raman Spectra of (a) GO (b) FRGO Figure 3.3 XPS spectrum of graphene oxide. The black circles represent raw data. The black line is fitted sum and the colored lines are fitted peaks using the software OMNIC for Almega Figure 3.4 XPS spectrum of functionalized reduced graphene oxide. The black circles represent raw data. The black line is fitted sum and the colored lines are fitted peaks using the software OMNIC for Almega Figure 4.1 Load versus elongation curves of the neat epoxy, and nanocomposites containing 0.1, 0.2, 0.5, and 1.0 wt. % FRGO Figure 4.2 Young s modulus of neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % FRGO. Error bars are for three samples Figure 4.3 Raman spectra of FRGO deposited on a SiO 2 substrate Figure 4.4 Comparision of the Young s modulus estimated using the Halpin-Tsai model with the experimental Young s modulus for the nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % FRGO Figure 4.5 Plot of tan δ versus temperature for neat epoxy and nanocomposite ix

11 containing 0.5 wt % Figure 4.6 T g for neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO Figure 4.7 Thermogravimetric analysis showing mass % versus temperature for neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO Figure 4.8 Expanded view of thermogravimetric analysis in the temperature range o C Figure 4.9 The first derivative of the thermogravimetric curve with respect to temperature for neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO Figure 4.10 Plot of the friction versus sliding distance for neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO Figure 4.11 Expanded view of the friction versus sliding distance for neat epoxy Figure 4.12 Optical microscopy images of the countersurface (i.e., tungustan carbide ball of diameter 6mm). (a) Clean surface. (b) Surface after sliding on neat epoxy showing no transfer film. Surface after sliding in the low friction regime for nanocomposites contaning (c) 0.1, (d) 0.2, (e) 0.5 and (f) 1.0 wt.% of FRGO. Transfer films are observed Figure 4.13 Optical microscopy images of the countersurface (i.e., tungustan carbide) after sliding in the high friction regime for nanocomposite containing (a) 0.1, (b) 0.2,(c) 0.5 and (d) 1.0 wt.% of FRGO Figure 4.14 Profilometer cross-sections of wear tracks taken at the end of the friction measurements shown in Figure 4.7 for the neat epoxy and the nanocomposites containing various wt. % of FRGO Figure 4.15 Wear rates of the neat epoxy and nanocomposite containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO in the high friction regimes and wear rates of various nanocomposites in low friction regime. Error bars are for three wear tracks Figure 4.16 A SEM images of wear tracks in the low friction regime of (a) and (b) Neat epoxy. (c) and (d) Nanocomposites containing 0.1 wt. % of FRGO. (e) and (f) Nanocomposites containing 0.5 wt. % of FRGO Figure 4.17 SEM images of wear tracks in the high friction regime of (a) and (b) Neat epoxy. (c) and (d) Nanocomposites containing 0.1 wt. % x

12 of FRGO. (e) and (f) Nanocomposites containing 0.5 wt. % of FRGO xi

13 CHAPTER 1 INTRODUCTION 1.1 Introduction The field of nanotechnology has attracted much attention over the last few decades due to increasing applications in many fields such as devices, sensors, and biomedical applications [1-2]. The development of these areas depends upon fabrication of nanomaterials of various sizes, shapes, and properties. The development of these nanomaterials has provided an opportunity to create new, very useful and very diverse materials which society could use for various applications such as personal protection, equipment for affordable health care and so on [3]. This can be achieved from suitable combinations of properties of various kinds of materials depending upon requirements. Polymers are one of the most commonly used materials due to impressive properties such as low cost, ease in processing, and recycling capability [3-4]. Further, the polymeric materials have provided numerous useful applications, ranging from the fabrication of materials used in daily life to aerospace components [5]. However, the intrinsic properties of these materials are not enough to meet many applications. Thus, some other kinds of suitable materials with quite different properties are combined to the polymer matrix in order to fabricate a new kind of material with improved properties [3]. In this way, a multi-component material is created with unique functional and physical properties, such as enhanced mechanical, thermal, and electrical properties [6-7]. Since the discovery of graphene by Geim and his group [8], this material has been tested for applications in many fields of science and technology due to its many interesting properties. Further, the development of techniques for the dispersion of graphene in the nanoscale range has 1

14 opened a new area in material science [9]. One of such area is polymer composite in which graphene or modified graphene is used as one of the constituent material [2, 9]. The remarkable properties of graphene such as superior mechanical strength (~130 GPa), very high Young s modulus (1 TPa), very high thermal conductivity (~5,000 W/mK), very high electrical conductivity (~600 S/cm), and high surface area (~2,630 m 2 /g) make this material suitable as a filler for the fabrication of nanocomposite materials [10-14]. 1.2 Composite Materials A composite material has quite distinct properties that are not present in any of the component materials [4]. There are three main constituents in any kind of composite material: the matrix, the reinforcement (i.e., filler) and interfacial region [4]. The interfacial region is responsible for the modification or improvement of any properties of the matrix [4]. There are various kinds of fillers such as one-dimensional fillers (e.g., fibers and nanotubes), twodimensional fillers (i.e., clay, graphite and graphene), and three-dimensional fillers (i.e., spherical particles) [3-4]. It requires suitable combination of properties of constituent materials in order to produce composite materials having improved properties instead of just mixing two or more materials [4]. Further, properties of polymer composite materials improve significantly with fillers of nano-scale dimensions because the mixing phase occurs in this scale range [3-4]. The composite material fabricated using a nano-filler is called a nanocomposite material. Most of nanomaterials have a high surface area compared to their volume. The uniform dispersion of this type of filler provides a very high interfacial area between the filler and matrix, and this differentiates nanocomposite material from traditional composite material [4]. Moreover, it requires much less of these nanofillers to fill the entire volume of the matrix [2]. Further, the 2

15 distance between the fillers is in the nano-scale range. These unique features of the fillers provide enhanced performance of the nanocomposite material compared to the neat matrix. There are various different kinds of composite materials such as a polymer matrix, cement matrix, metal matrix, carbon matrix, and ceramic matrix [15]. The polymer matrix composite is very common and is useful for light weight production of structural materials, which has many potential applications in the automotive, aerospace, construction and electronic industries [9]. The fabrication of polymer nanocomposite materials has been growing rapidly along with the development of new kinds of nanofillers. Generally, the nanocomposite materials are fabricated using nanofillers such as layered silicate clays [15], carbon nanofibres [16-17], carbon nanotubes [18-20], expanded graphite [3,21], and graphene [2,9]. These nanofillers have different chemical structures, morphologies, and aspect ratios. Among these nanofillers, graphene and modified graphene (i.e., graphene oxide, reduced graphene oxide, and functionalized graphene) offer advantages as fillers because of their planar structure that facilitates the interaction with the polymer matrix at the molecular level [22]. The significant improvement of properties of graphene-based nanocomposite materials depends upon the level of homogeneous dispersion of graphene in the polymer matrix. The uniform dispersion of these fillers into the matrix provides higher efficiency of external load transfer to the filler through strong interfacial interactions. 1.3 Band Structure of Graphene Graphene has a 2-dimensional honeycomb lattice with sp 2 bonded carbon atoms as shown in Figure 1.1 [23]. This material is the building block of all other forms of carbon material such 3

16 4 as 0-dimensional fullerenes, 1-dimensional nanotubes and 3-dimensional graphite. The unit cell of graphene consists of two non-equivalent carbon atoms A and B separated by 1.14 Å as shown in the Figure 1.1. This distance is less than its inter-planar distance, which is 3.35 Å The electronic properties of graphene can be derived from its band structure. The tight binding model is generally used in order to determine the band structure of graphene. The unit cell vectors a 1 and a 2 is given by a a a a 2 3, 2 3, 2 3, 2 3 a 2 1 a The lattice constant is given by the magnitude of each of the 2 1 a a c a The reciprocal lattice vectors b 1 and b 2 are given by the relations: ; 2 ; 2 a a a a a a a a a a a a a a a b b b where 3 a is a unit vector along a z-axis, which does not play any role in current discussion as the electronic states in x-y plane is only discussed here. The above relation gives a a a a a a a a 3 2, 3 2, 3 2, 3 2 b 2 b 1 (1.1)

17 Figure 1.1: Representation of hexagonal lattice of graphene sheet. The shaded region represents the unit cell. Figure 1.2: First Brillion zone in reciprocal space where, Γ, K, K, and M are high symmetry points. 5

18 This shows that the reciprocal lattice vectors b 1 and b 2 are rotated by 30 with respect to the real space vectors a1 and a 2, respectively. Figure 1.2 also shows high symmetry points (i.e., Γ, K, K and M) within the first Brillion zone where, K and K are the unequivalent Dirac points in the reciprocal space. The solution to the tight bending method, considering only nearest neighboring atoms, gives the following energy dispersion relation for single layer graphene. 3K K a K a xa y 2 y E( K x, K y ) 1 4cos cos 4cos (1.2) Here arises from the nearest neighbor contribution and its value is about 0.3 ev. Figure 1.3: Band structure of graphene derived from tight binding approximation. 6

19 The positive and negative parts of equation (1.2) lie above and below the Fermi level as shown in Figure 1.3, respectively. The part above the Fermi level is the conduction band and the part below is the valance band. The conduction and valance bands touch each other at singularity points (K and K ). The band structures around these points are linear and this gives that the charge carriers are massless [24]. 1.4 Graphene Oxide Graphene oxide (GO) has a two-dimensional planar structure similar to graphene but with disrupted sp 2 hybridized carbon atoms due to covalent attachment of various oxygen containing functional groups, as shown in Figure 1.4. The complete restoration of sp 2 hybridization of a single layer of GO gives a graphene sheet. Graphene is usually produced by methods such as mechanical exfoliation, chemical vapor deposition, epitaxial growth, and chemical exfoliation. Among these methods, chemical exfoliation is the route best suited for mass production of graphene [25]. This method is cheaper, simpler, more efficient, and better yielding and is suitable for industrial or commercial applications [25]. The chemical exfoliation of graphene from graphite is an indirect approach because graphene oxide (GO) is first produced and then reduced to graphene. The oxidation of crystalline graphite into GO breaks the sp 2 hybridized structure of the graphene sheets that are stacked in graphite [26]. The oxidation also increases the distance between the adjacent graphene layers from nm to 0.68 nm. The increase in separation depends upon amount of water intercalated between the stacked sheets, and this reduces the Van der Walls interaction between the sheets [27]. During the oxidation process, the oxygen molecules not only increase the separation between sheets, but also make the layers hydrophilic in nature. Low power sonication 7

20 in water is sufficient to disperse the GO into individual sheets from the oxidized graphite. The GO produced in this way has various oxygen containing functional groups attached to the edge and the basal plane of the graphene sheets as shown in the Figure 1.4 [2]. The major oxygencontaining functional groups attached to the edge of graphene sheets are carbonyl and carboxylic, while the major oxygen-containing functional groups attached to the basal plane are hydroxyl and epoxide [2]. Figure 1.4: Chemical structure of graphene oxide with different oxygen-containing functional groups on the basal plane and around the edges of a graphene layers. GO was used as a filler in a polymer matrix to produce nanocomposites. GO containing high amount of oxygen functional groups is analogous to two-dimensional clay sheets such as montnorilllonite. Nanocomposites based on the clay sheets have been investigated extensively for various applications [28]. The filler clay is in the range of a few microns thick whereas GO exhibits a much higher surface-to-volume ratio. In addition to the high surface area, GO has a 8

21 high dispersibility in various aqueous and inorganic solvents. Further, different oxygen functional groups attached to the basal plane and the edge of GO promote interaction between the filler and matrix [2]. This allows GO to be used as a unique filler material to produce various kinds of nanocomposite materials for different applications. The oxygen-containing functional groups can be reduced significantly before incorporating the material into the matrix. 1.5 Functionalization of Graphene and Graphene Oxide Pristine graphene flakes do not disperse in many solvents because of their tendency to aggregate due to Van der Walls interactions between the layers. This creates difficulties in producing graphene-based nanocomposites with enhanced properties. In order to overcome this problem, graphene sheets are usually functionalized with suitable organic molecules. There are various approaches to functionalization of graphene and graphene oxide. The covalent functionalization of graphene is one approach that involves attaching certain organic molecules on the surface of graphene [9,111]. This process disrupts the sp 2 hybridization of graphene and forms sp 3 hybridized structures [111]. This also helps to tune the band gap of the material that is very useful in the fabrication of electronic devices [112]. The covalent functionalization of graphene can be carried out in two different ways [111]. The first approach is to form direct covalent bonds between the sp 2 hybridized carbon structure and organic molecules that are to be attached [111]. The other approach is to attach the organic molecules covalently with oxygen containing functional groups of GO [111]. Different organic molecules such as amines, isocyanates, and diisocyanates compounds have been used for covalent functionalization of graphene and GO [84, ]. These molecules reduce the hydrophilic nature of GO. Stankovich et al. [113] have reported that isocyanate functionalization 9

22 of GO readily forms a stable dispersion of GO in a variety of organic solvents such as dimethylformamide (DMF) and N-methylpyrrolidone (NMP). Covalent attachment of octadecylamine (ODA) to the surface of GO provides a way to introduce long hydrocarbon chains that makes ODA functionalized GO disperse very well in polar solvents [111]. The ODA functionalization of GO occurs due to the nucleophilic substitution reaction between the amine group of ODA and the epoxy group of GO [111]. The ODA functionalization of GO enhances the surface roughness of GO, and the effect is significant with longer amine chains [53]. The surface roughness of ODA functionalized GO enhances the dispersion in polymers [53]. The reduction of a stable dispersion of GO using hydrazine monohydrate produces irreversible aggregation of the reduced graphene sheets. This problem can be solved by suitable functionalization of GO before reduction. Non-covalent functionalization is another technique that prevents aggregation of graphene sheets. In this type of functionalization, functional groups are attached to graphene without disrupting its sp 2 hybridized structure [ ]. Stankovich et al. [50] have reported that the reduction of exfoliated graphite oxide in the presence of Poly (styrenesulfonate) produces a stable aqueous dispersion of reduced GO. The reduction of GO using hydrazine in the presence of single-stranded DNA (ssdna) forms non-covalent functionalization of graphene with ssdna that is dispersible in water up to 2.5 mg/ml [116]. The functionalization of GO with aryl-diazonium salt followed by reduction using hydrazine monohydrate forms GO wrapped with sodium dodecyl benzene sulfonate (SDBC) functional groups [117]. This type of functionalization makes the graphene disperse easily in inorganic solvents such as DMF and NMP up to concentration of 1 mg/ml with significantly less aggregation [117]. The thermal stability of SDBC functionalized GO is significantly higher than that of GO [117]. 10

23 The functionalization of graphene and graphene oxide with both covalent and noncovalent attachment of suitable functional groups provides highly dispersible graphene sheets with high concentration in a variety of solvents. This makes the material quite useful in many different applications that include fabrication of electronic devices, energy storage devices, and nanocomposite materials. 1.6 Epoxy Resin Epoxy resin is a highly cross-linked thermosetting plastic. Epoxy resin is widely used in a variety of applications such as adhesive, paint, coating, sealant, medical implants, and electrical devices [29-30]. Epoxy is also used as the matrix for the fabrication of polymer nanocomposites, which are used in aerospace and wind-turbine industries [5]. The three dimensional structure of cured epoxy resin provides excellent physical properties [31]. The main component of the epoxy resin is diglycidal ether of bisphenol A (DGEBA) as shown in Figure 1.5. Bisphenol A is produced from the reaction of two phenols with one acetone [31]. After adding the hardener, oxygen atoms from the glycidyl groups of the epoxy resin react with the amine hydrogen atoms of the hardener in order to produce cured epoxy resin [31]. The most common hardener of epoxy resin are the polyamines, which are organic materials with two or more amine groups. The amine group consists of nitrogen atoms with one or two hydrogen atoms attached to it. The epoxy resin is referred to as Part A whereas the hardener is referred to as Part B. Figure 1.5: Chemical structure of diglycidylether of Bisphenol-A (DGEBA), n = 0, for the derivatives, n > 0. 11

24 1.7 Challenges in Polymer Nanocomposites There are many challenges in order to prepare polymer-nanocomposite (PN) materials that utilize the advantages of the filler materials. This section discusses a few of the challenges such as dispersion of nanomaterials, cost effectiveness, and health and environment issue Dispersion of Nanomaterials The uniform dispersion of nanomaterials in the matrix is one of the challenges to produce nanocomposite materials with improved performance [2]. The difference in surface charge between the filler and polymer and Van der Walls interactions between the nanomaterials often causes agglomeration of nanomaterials in the matrix [32]. Figure 1.6 shows three different types of composites according to the level of dispersion. The phase separated polymer composite material is formed when a polymer matrix is unable to penetrate in between the layers of filler material [5, 33]. In this situation, the actual potential of the filler material is not completely utilized. The properties of the composite material improve slightly or can become worse than the neat polymer [34]. In the case of intercalated composite materials, as shown in Figure 1.6, the polymeric chains are intercalated in between the layers of nanomaterials, but not the latter are dispersed fully [3]. This increases the performance of the composite material compared to phase separated materials. The third type of dispersion is exfoliated. In this case, the layered fillers such as graphene or reduced graphene are completely and uniformly dispersed in the polymer matrix [3]. This allows significant improvement in the properties of the composite material. Further, one can modify the properties of composite materials by optimizing the interfacial bonding between the filler and polymer [5]. 12

25 Figure 1.6: Schematic of various levels of dispersion of nanomaterials in a polymer matrix Cost Effectiveness Nanomaterials having a high aspect ratio, such as carbon nanotubes and graphene, are highly desirable for the fabrication of nanocomposites with enhanced mechanical, electrical, and thermal properties. Specifically, carbon nanotubes are quite expensive to produce. So, it is important to reduce the cost of nanofillers so that they can be utilized in the manufacturing of nanocomposite. This can often be achieved by choosing a filler without sacrificing the properties of the nanocomposite. 13

26 1.7.3 Health and Environment Issues The use of nanocomposite materials is increasing yearly due to their enhanced performance compared to the neat material. As a result of increased use, there is more concern about the impact of these materials on health and the environment [3]. Nanomaterials can easily reach inside the body through inhalation, skin contact or by ingestion. As most of the nanomaterials have a high surface area, the atoms of nanomaterials can react easily with the atoms of tissue of the human body [3]. This may cause severe health effects. For example, carbon materials, which have a significant impact on health, are widely used for the fabrication of nanocomposites. The carbon materials cause skin diseases and respiratory problems [35]. Some of the nanoparticles can be inhaled easily and reach deep in lung tissue. Similarly, nanoparticles formed during the combustion process such as forest fires and industrial wastes have severe effects on health [36]. Therefore, a better understanding of toxicity of nanomaterials and nanocomposite materials is essential to develop materials having little impact on health and on environment. 1.8 Motivation Epoxy resin is an important class of thermoset materials used in a wide variety of applications such as structural applications [31]. The wide variety of applications comes from its characteristics, including high chemical and corrosion resistance, good mechanical, thermal and electrical properties, and easy processability [29-30]. However, these materials have a high friction coefficient and low wear resistance [37]. Thus, it is very important to reduce the friction and wear rate of the epoxy as these properties are associated with durability. The reduction in friction and wear rate results from the reduction of adhesion to the counterface and the 14

27 enhancement of mechanical properties such as Young s modulus, tensile strength, and hardness of the nanocomposites [3]. This improvement can be accomplished by the addition of certain kind of fillers into the epoxy matrix. Previously, fillers containing fluorine atoms such as fluoropolymers were used to enhance the tribological properties of epoxy [38]. Unfortunately, this filler reduces wear resistance [38]. Fillers such as carbon fibers and carbon nanotubes have been used to enhance the tribological properties [39-40]. Dong et al., have observed that incorporation of 1.5 wt. % multi-walled carbon nanotubes into the epoxy matrix reduces the friction coefficient from ~0.32 to ~ 0.2 and the wear rate from to mm 3 /Nm [39]. Although fillers like carbon nanotubes have the capability to improve tribological properties, they also have a high tendency to agglomerate in the epoxy matrix. This makes it difficult to produce a nanocomposite with enhanced properties. Furthermore, the high cost of fabrication of carbon nanotubes limits their use in the fabrication of nanocomposite materials. On the other hand, chemically derived graphene is very suitable for mass production and potentially very useful for the fabrication of nanocomposites. Previous studies have shown that the addition of graphene in various polymers results in significant improvement in mechanical, thermal, electrical, and tribological properties [9, 32, 41-45]. Recently, Kanudanur et al., reported that the addition of graphene in polytetrafluoroethylene reduced the wear rate significantly [44]. Further, Pan et al., have also reported improvement in the tribological properties of nylon matrix with the addition of very low concentrations of octadecylamine functionalized graphene oxide [45]. These results on chemically derived graphene have led me to study the mechanical, thermal, and tribological properties of graphene-epoxy resin nanocomposites. 15

28 1.9 Outline of Dissertation This dissertation concerns the fabrication, characterization, and properties of epoxy nanocomposite reinforced with various weight percentages of functionalized reduced graphene oxide (FRGO). Chapter 1 deals with a description of polymer nanocomposite materials, epoxy resin, the band structure of graphene, graphene oxide, challenges in polymer nanocomposites, and the motivation for the research. Chapter 2 discusses the experimental techniques for the synthesis of graphene oxide, functionalization of graphene oxide, and synthesis of FRGO-epoxy nanocomposite materials. This chapter also discusses the methods used to characterize FRGO and the nanocomposite materials. In Chapter 3, the characterization of FRGO using scanning electron microscopy, Raman spectroscopy, and X-ray photoelectron spectroscopy is presented. Chapter 4 discusses the enhancement of Young s modulus, tensile strength, glass transition temperatures, friction, and wear properties of nanocomposite materials fabricated with various weight percentages of FRGO filler. Chapter 5 gives conclusions about the properties of FRGO and the FRGO-epoxy nanocomposite materials. 16

29 CHAPTER 2 EXPERIMENTAL TECHNIQUES 2.1 Synthesis of Graphene Oxide The general method to produce large quantities of graphene is to begin with the oxidation of graphite to graphite oxide (GO). The procedures for the synthesis of GO were developed a few decades ago by Brodie [46], Staudenmeire [47], and Hummers et al.,[48], and these procedures are still in use, with and without modifications. The first two methods are very time consuming and generate ClO 2, which must be handled with caution due to high toxicity and tendency to decompose in air and explode [46, 49]. Thus, modified Hummer s method was used for the preparation of GO, since it is less hazardous compared to the other two methods. In this method, GO preparation starts from bulk graphite or graphite powder. The graphite powder is oxidized in a water-free medium by treating with concentrated sulfuric acid (H 2 SO 4 ), sodium nitrate (NaNO 3 ), and potassium permanganate (KMnO 4 ). The active oxidizing agent is dimanganese heptoxide (Mn 2 O 7 ), which is formed from the reaction of KMnO 4 with H 2 SO 4 as shown in reaction 2.1. KMnO 3H SO 3 4 MnO MnO K Mn O 2 MnO H O HSO 4 (2.1) The oxidation of graphite can be accomplished using various commercially available graphite powders. However, graphite flake is widely used for making oxidized graphite for various applications. The GO synthesized using this process not only contains oxidized graphite, but also non-oxidized heavy graphite particles. The non-oxidized graphite particles can be 17

30 removed using centrifugation and sedimentation. The prolonged dialysis removes salt and ion impurities from the oxidation process. A layout of synthesis of GO is shown in Figure 2.1. Figure 2.1: A layout of GO synthesis procedure. The entire reaction process takes less than two hours and is carried out below 45 C with comparatively little danger. In our preparation of GO, 5 g of graphite powder (325 mesh, Southwestern Graphite) and 2.5 g of NaNO 3 were added to 115 ml of concentrated H 2 SO 4 at room temperature and stirred for 15 minutes. The mixture was transferred to an ice bath and 15 g of KMnO 4 was added while stirring vigorously. The mixture was then transferred to a water bath to maintain the temperature within the range of C for half an hour as shown in Figure 2.2 (a). After half an hour, 235 ml of deionized (DI) water was added to slow the reaction and the 18

31 mixture was stirred for another 15 minutes. At this stage, the color of the sample becomes dark brown as shown in Figure 2.2 (b). An additional 830 ml of water was added followed by the slow addition of H 2 O 2 (30%). The mixture was then repeatedly filtered and washed with HCl (1:10) aqueous solution. The filtered material was dispersed in water using horn sonication, and centrifuged at 3500 rpm in order to remove any non-oxidized graphite. Residual acids and salt impurities were removed using dialysis for about 10 days using cellulose permeable membrane. Finally, the suspension was freeze dried using the setup shown schematically in Figure 2.3 to get a powdered GO. In the freeze drying process, the water from the GO dispersion sublimates at low pressure. The water-vapor produced during this process is condensed to water and is trapped by a refrigerator. Figure 2.2: Photographs during the oxidation process of graphite powder. (a) During the second stage of the oxidation reaction. (b) After the first dilution with DI water. 19

32 Figure 2.3: Schematic of a freeze drying system. 2.2 Functionalization of Graphene Oxide with Octadecylamine Pristine graphene samples are not suitable for dispersion in polymers because of their tendency to agglomerate due to π-π interactions, which is difficult to undo using sonication [50]. The agglomeration of sheets can be reduced significantly by functionalization before reduction [9]. The functional group attached to the graphene sheets can be small molecules of polymer chains. The chemical functionalization of graphene is very useful because it increases the solubility in organic polymers with enhanced interactions [51-52]. In this study GO was functionalized using octadecylamine (ODA). The octadecylamine functionalization not only facilitates uniform dispersion of graphene in the epoxy, but also prevents the re-aggregation of graphene sheets [9, 51-52]. The ODA functionalized GO exhibits increased surface roughness with a higher hydrophobicity [53]. The ODA functionalization was carried out by dispersing 300 mg of GO powder in 300 ml of ethanol by sonication for 2 h and then adding 450 mg of ODA in 45 ml of ethanol. The mixture was refluxed for 24 h at 90 C using the setup shown schematically in Figure 2.5, and then repeatedly filtered and rinsed with ethanol to remove excess ODA. ODA functionalization 20

33 occurs by nucleophilic substitution reactions between the amine groups of ODA and the epoxide groups of GO [54]. Therefore, ODA functionalization partially reduces GO; the presence of other oxygen-containing functional groups, such as hydroxyl and carbonyl groups, remains as shown in Figure 2.4 (a). These groups can be reduced by reaction with hydrazine monohydrate without significant effects on ODA functionalization [55]. For reduction, the functionalized GO powder was dispersed in ethanol, and hydrazine monohydrate was then added. The mixture was refluxed at 90 C for 24 h, and the final product was repeatedly filtered and washed with ethanol and then dried in a vacuum oven. The chemical structure of functionalized reduced GO is shown in Figure 2.4 (b). Figure 2.4: (a) Schematic of GO functionalized with ODA. (b) Schematic of functionalized reduced GO. The long chains represent ODA and is attached to the basal plane of GO. 21

34 Figure 2.5: Schematic of the setup used to functionalize GO. 2.3 Characterization of Graphene Oxide and Functionalized Reduced Graphene Oxide Raman Spectroscopy Raman spectroscopy is one of the most important tools for the study of carbon based nanostructures due to its non-destructive nature and fast acquisition of data [56]. More importantly, Raman analysis provides structural and electronic information about the carbon structure [56]. Raman spectroscopy gives detailed information on the crystal structures of graphene oxide, reduced graphene oxide, and functionalized graphene oxide. Raman spectroscopy also provides an indication of the degree of oxidation of GO and FRGO. 22

35 Raman spectra of graphene consists G and 2D peaks at 1580 cm -1 and 2700 cm -1, respectively. The G peak is due to in-plane bond stretching of pairs of sp 2 carbon atoms, as shown in Figure 2.6 (a). The 2D peak is due to second order zone boundary phonons [57]. In the Raman spectra of defected graphene there is a D peak located at about 1350 cm -1 due to first order zone boundary phonons, which is absent in the case of defect free graphene [56]. The D peak is similar to a breathing mode which requires out-of-plane translational motion induced by a sp 3 hybridized structure, as shown in Figure 2.6 (b). Figure 2.6: Vibrational modes of (a) G peak and (b) D peak. The black circles represent the carbon atoms and the arrows represent the direction of vibration. The G and D peaks of GO are broad compared to graphite because of disorder due to extensive oxidation. Further, the G peak shifts towards higher frequency, which is attributed to isolated double bonds that resonate at a frequency higher than that of graphite [56]. The upward shift of the G peak is due to a reduction in size of sp 2 hybridized carbon [58]. Moreover, the full width at half maximum of the 2D peak of GO is very wide (i.e., about 200 cm -1 ) compared to 23

36 that of mechanically exfoliated graphene (i.e., 30 cm -1 ). The overall Raman peak intensities are reduced after reduction, which suggests the loss of carbon during the reduction process [59]. Reduced GO has a structure that is different from that of GO due to the removal of oxygen-containing functional groups and carbon atoms. The area under the D and G peaks of the Raman spectrum is a measure of the size of the sp 2 clusters in a networks of sp 2 and sp 3 carbon. According to the Tuinstra-Koenig relation, the intensity ratio of G and D peaks (i.e., G/D) is proportional to the average size of in-plane sp 2 hybridized carbon [58] X-ray Photoelectron Spectroscopy (XPS) of Graphene and Graphene Oxide XPS was used to characterize graphene oxide, reduced graphene oxide, and functionalized graphene oxide in order to understand the structural and electronic properties of these materials. Since GO consists of many oxygen-containing functional groups, it is important to know the concentrations and types of functional groups. In XPS, the electrons are ejected from the sample when the sample is irradiated with mono-energetic soft x-rays as shown in Figure 2.7. The identification of the sample can be accomplished by analyzing kinetic energies of the ejected photoelectrons according to equation 2.2. Similarly, one can determine the relative concentration of the elements from the intensities of the ejected photoelectrons. The kinetic energy of ejected electrons can be calculated from the energy conservation equation: E E ( E ) (2.2) KE h BE 24

37 where, E is the kinetic energy of the photoelectron, E h is the energy of the incoming X-ray KE (e.g., ev for Al, Kα X-rays). Similarly, function of the spectrometer. EBE is the binding energy and is the work Figure 2.7: Ejection of a photoelectron upon irradiation by a mono-energetic X-ray. 2.4 Synthesis of Functionalized Reduced Graphene-Epoxy Nanocomposites Graphene-polymer nanocomposites are commonly prepared using techniques such as solution-blending, melt mixing, and in-situ polymerization. The functionalized reduced graphene-epoxy nanocomposites were prepared using a solution-blending method. This is the most popular technique in which a solvent for both fillers and epoxy is used. Initially, the filler material is dispersed in a solvent and then the mixture is added to the polymer in the same solvent. Later, after mixing the mixture of fillers and polymer, the solvent is removed by evaporation. 25

38 The process of synthesis of FRGO-epoxy nanocomposites is shown schematically in Figure 2.9. First, FRGO was dispersed in acetone (100 mg of FRGO in 100 ml of acetone) using horn-sonication (shown in Figure 2.8) for 2 hours in an ice bath in order to prevent any increase in temperature. Varying amounts of epoxy resin (System Three Resin, Inc.) were then added and the mixture was sonicated for 1 hour. The acetone was evaporated by heating the mixture at 70 C using a hot-plate. Residual acetone was removed by placing the mixture in a vacuum oven for 12 hours at 70 C. After cooling to room temperature, a low viscosity slow curing agent (System Three Hardener Part B, Number 3) was added. The mixture was poured into silicone molds of different shapes to make samples for various types of characterizations. The samples were cured for 24 h at room temperature. All the samples with varying concentration of FRGO were fabricated in a similar way. The characterization of all samples was carried out after a week of curing. Figure 2.8: Photograph of the horn sonicator used for the dispersion of FRGO. 26

39 Figure 2.9: A schematic representation of FRGO-epoxy nanocomposite preparation. 27

40 2.5 Characterizations of Polymer Composite Materials Mechanical Properties: Tensile Testing The study of the mechanical properties of polymer composite materials is one of the most important studies in the field of material science and engineering. This gives insight into the ability of the material to withstand various load levels [60]. It is important to know the viscoelastic properties of the material in order to understand its mechanical properties [61]. The viscoelastic properties depend both on internal and external conditions. The internal conditions refer to the elastic, viscous or a combination of both properties of the material, whereas the external conditions refer to the temperature and pressure [60]. The viscous properties can be observed when a force is applied to a polymeric material for an extended period of a time at a lower shear rate. During this time, the polymeric chains respond to the force, and the material flows along the direction of the force [60]. When the force is applied for a short duration of time, then the molecular chains do not have sufficient time to flow along the direction of the force [60]. In this case, the cross-linking and entanglements of the polymers are responsible for the elastic properties [60]. There are various kinds of tests used to study the mechanical properties of polymeric materials. The tensile test and the dynamic mechanical analysis are examples of static and dynamic mechanical tests, respectively. Tensile testing is a widely used method used to determine the Young s modulus, tensile strength, Poisson s ratio, and strength at break. This consists of a constant speed movement crosshead and a fixed crosshead, both of which have threaded sample grips, as shown in Figure The movable crosshead moves at a constant speed (i.e., tensile rate) with respect to the fixed crosshead, and this produces elongation or compression in the sample. The tensile rate can be varied from 1 mm/min to 500 mm/min depending on the type of material. The samples are 28

41 usually in the form of a dog-bone shape as shown in Figure In the tensile testing measurement, stress at increasing strain is measured at a constant tensile rate until the sample breaks due to application of the load. Figure 2.10: Schematic of the tensile testing equipment. Figure 2.11: Schematic of a tensile testing sample. 29

42 Stress is defined as F (2.3) A 0 where is the tensile stress, F is the applied force, and A0 is the initial cross-sectional area of the sample. The maximum value of the stress is called tensile strength of the material. The strain is given by the relation ( L0 L) L (2.4) L L 0 0 where L0 is the initial length of the sample and L is the length after a certain amount of strain (i.e., the current length). The Young s modulus or elastic modulus can be calculated using the equation E (2.5) E is calculated within the elastic limit i.e., the initial linear portion of the stress versus elongation curve Mechanical Properties: Dynamic Mechanical Analysis Dynamic mechanical analysis (DMA) is used to characterize the viscoelastic behavior of polymeric materials [61]. The viscoelastic property is a combination of properties of elastic solids and Newtonian fluids. In DMA, a periodic force is applied to the material, and its response 30

43 is measured in terms of strain at various temperatures. The frequency of the periodic force can be varied depending on the type of material and range of measurement [61]. The strain of a viscoelastic material is out phase with respect to the applied stress by an angle, as shown in Figure The phase change occurs due to the excess time necessary for molecular motions and relaxations to occur. The amplitude of deformation at the peak of sine wave and the lag between the stress and strain give quantities such as the modulus, viscosity, and damping. Figure 2.12: Response of a sinusoidal force applied to a polymeric material at a certain frequency. The dynamic stress and the strain are given by the relations sin ( t ) (2.6) 0 sin ( ) (2.7) 0 t where is the angular frequency. Now, from equations (2.6) the stress can be expressed as 31

44 0 sin ( t)cos ( ) 0 cos ( t)sin ( ) (2.8) The above equation can be expressed as ' " E sin ( t) E cos ( ) (2.9) 0 0 t where ' 0 cos 0 E and E " sin. 0 0 The complex modulus is given by the relation E * 0 e 0 i 0 (cos i sin ) E 0 ' ie " (2.10) The above equation shows that the complex modulus obtained from a DMA consists of real and imaginary parts. The real part of equation (2.10) is the storage modulus i.e., the ability of the material to store and release potential energy upon deformation. The imaginary part of equation (2.10) is the loss modulus, and is associated with energy dissipation in the form of heat energy. The phase angle is given by " E tan (2.11) ' E The glass transition temperature is another interesting parameter which can be obtained from the DMA measurements. The glass transition temperature is obtained from the peak position of tan versus temperature curve. 32

45 The DMA was carried using a PerkinElmer DMA8000 apparatus. A schematic of the equipment is shown in Figure The measurements were performed in the single cantilever bending mode as shown in Figure Figure 2.13: Schematic of the driving system of the DMA equipment. Figure 2.14: Schematic of the single cantilever (i.e., 2-point bending mode) system of the DMA equipment. 33

46 2.6 Thermal Stability Determination: Thermogravimetric Analysis (TGA) Thermogravimetric analysis (TGA) of polymeric materials is carried out by measuring mass as a function of increasing temperature at a constant rate in a controlled environment [62]. The controlled environment can be an inert or reactive gas. The TGA measurement gives information about the physical and chemical properties by recording changes in mass with increasing temperature [63]. Most polymers exhibit mass loss with increasing temperature due to the removal of unstable components such as moisture, residual solvents or low molecular mass additives or oligomers that generally evaporate between room temperature and 300 C [62]. Another source of mass change may be due to absorption, adsorption, desorption, dehydration, decomposition, oxidation, and reduction [63]. A TGA system consists of a sample pan made from either ceramic or platinum that is connected to a precision micro-balance. The chamber is purged with a gas in which the sample is to be analyzed. In most cases, inert gases such as nitrogen, argon or helium are used to study the thermal degradation behavior. Figure 2.15 shows the TGA system used in our studies. 34

47 Figure 2.15: Photograph of the thermogravimetric analyzer. 2.7 Tribological Properties of Polymer Nanocomposites Tribological properties of polymers refer to the friction coefficient and wear resistance of the materials. The tribological properties are not only a material property as they depend on the environment as well [64]. It is important to study fiction and wear properties of epoxy as they are related to durability and performance. Further, epoxy resin has high friction and wear rate [37]. Thus, such studies provide a measure of improvements in friction and wear resistance. 35

48 2.7.1 Friction Friction is a force that always resists the motion of an object when it slides or rolls on the surface. This force always acts tangentially along the direction opposite to the motion, as shown in Figure 2.16 (a). There are two types of frictional forces: static and dynamic friction. The static friction acts before the beginning of motion of an object, while the dynamic friction acts when the object is in motion. These forces are measured from the ratio of the applied force F to the weight W of the object. The frictional property is not completely a material property [64]. It depends on the condition such as temperature, pressure, atmosphere (i.e., air or nitrogen or vacuum), material roughness, roughness and cleanness of the counterface, lubrication, and speed. Further, the friction of polymeric materials is the result of the interfacial and the cohesive work done on the surface [3] as shown in Figure 2.16 (b). It is assumed that the counterface is hard compared to the polymeric material, and undergoes little or no elastic deformation [3]. The interfacial work depends on the adhesive interaction between the counterface and the polymeric material and that varies with factors such as hardness, surface roughness, glass transition temperature, and electrostatic-chemical interaction between polymer and counterface [3]. The other factor affecting the interfacial work is the cohesive interaction within the polymeric material, and it results in plowing of the polymer by the counterface [3]. Further, the energy required for the plowing depends upon factors like the Young s modulus, tensile strength, and geometry of the counterface [3]. 36

49 (a) (b) Figure 2.16: (a) Schematic illustration of a hard body sliding over a polymeric surface. (b) Enlarged view of the region of polymeric material in contact with the hard surface Wear Wear is an outcome of friction when an object slides over the surface of another [85]. This is also not completely a material property. It depends on operating conditions such as speed, contact pressure, and surface roughness of the counterface. Most of materials having high friction also have high wear rate. There are various kinds of wear mechanisms for polymeric materials such as interfacial, cohesive, abrasive, and adhesive [3]. The factors affecting the wear rate are elastic modulus, tensile strength, and hardness of the polymer. Generally, polymers having higher elastic modulus have higher wear resistance [3] Friction and Wear Measurements of Polymeric Materials: Pin-on-disc Tribometer There are various methods for determining the friction and wear of polymeric materials. The pin-on-disk method is one of the widely used methods for simultaneous determination of friction and wear. A schematic of a pin-on-disk tribometer is shown in Figure 2.17, and a 37

50 photograph of the equipment used for measurements is shown in Figure In this type of measurement, the pin (i.e., the counterface) slides over the surface of the polymeric material in a circular path, and it penetrates the sample as measurement are made. The depth through which the pin penetrates into the sample depends upon wear resistance. The depth through which the pin penetrates is measured using a linear differential transduction device (LVDT). The sliding speed, radius of circular track, and type of counterface (i.e., pin) can be varied depending upon choice of measurement. The wear of a composite material using the pin-on-disk method can be measured by determining the dimension of the groove produced by the pin. The dimension of the groove can be measured using a profilometer. If A is the cross-sectional area of a worn track and R is the radius of the track, then the volume V of the groove is given by V 2 R A (2.12) The wear rate is given by W V (2.13) N x where N is the normal load and x is the sliding distance. The rate is usually measured in units of mm 3 /N m. The wear mechanism is generally studied using optical and scanning electron microscopy (SEM) of the surface, debris formed during abrasion, and film transferred from the sample to the counterface. 38

51 Figure 2.17: Schematic of the pin-on disk tribometer and a polymeric nanocomposite sample with a wear track. Figure 2.18: Photograph of the pin-on disk tribometer used for friction and wear measurements. 39

52 CHAPTER 3 CHARACTERIZATION OF GRAPHENE OXIDE AND FUNCTIONALIZED REDUCED GRAPHENE OXIDE This chapter deals with the characterization of functionalized reduced graphene using scanning electron microscopy (SEM), Raman spectroscopy, and X-ray photoelectron spectroscopy. 3.1 Characterization of Functionalized Reduced Graphene Oxide Using SEM The surface morphology of FRGO powder was characterized using an FEI Nova 200 NanoLab SEM. Figure 3.1 shows SEM images at various magnifications. The higher magnifications images (i.e., Figure 3.1 (b) and (c)) of FRGO powder show that most of the layers are separated. Further, Figures 3.1 (b) and (c) show that surface and structure of graphene sheets are not damaged by chemical modification. The planar structure of graphene is completely preserved even after the chemical reaction. Furthermore, it can be observed from the SEM images that that the lateral dimensions of the graphene flakes are in the range of a few microns. Note: Most of this chapter is duplicated from the accepted paper with the permission of Maney Publishing: Effects of Functionalized Reduced Graphene Oxide on Frictional and Wear Properties of Epoxy Resin, R. Shah, T. Datashvili, T. Cai, J. Wahrmund, B. Menard, K. P. Menard, W. Brostow, J. Perez, Material Research Innovations, References and figures numbers are changed to accommodate the dissertation. 40

53 Figure 3.1: (a-c) SEM images of FRGO powder at different magnifications. 41

54 3.2 Characterization of Functionalized Reduced Graphene Oxide Using Raman Spectroscopy Micro-Raman spectroscopy was carried out using a Thermo Electron Almega XR spectrometer using a green laser (i.e., λ= 532 nm) as the excitation source. Figure 3.2 shows Raman spectra of the GO and FRGO powders. The spectra are normalized so that the G peaks of the GO and FRGO have the same height. The G peak is due to first order scattering of E 2g phonons (in plane optical mode) of sp 2 hybridized carbon atoms close to the Γ point [65]. As shown in Figure 3.2 (a), the G peak in GO is at about 1593 cm -1. The G peak is shifted from its value in graphite of 1581 cm -1 due to oxidation. Carbon materials also exhibit a D peak at about 1340 cm -1 due to defect-induced zone boundary phonons [66]. The D peak in GO is due to a reduction in size of sp 2 hybridized domains due to oxidation [65]. The Raman spectrum of FRGO has G and D peaks at 1587 cm -1 and 1346 cm -1, respectively, as shown in Figure 3.2 (b). In FRGO, the G peak shifts towards the position of the G peak in graphite due to restoration of sp 2 hybridized domains [67]. The ratio of the intensities of the D and G peaks in FRGO is greater than that in GO due to a decrease in size of in-plane sp 2 hybridized domains in FRGO due to reduction [66]. The Raman results are in good agreement with previous reports on GO and FRGO [68]. Figure 3.2: Raman Spectra of (a) GO (b) FRGO. 42

55 3.3 Characterization of Functionalized Reduced Graphene Oxide Using XPS The GO and FRGO was further characterized using a VersaProbe X-ray Photoelectron Spectroscopy (XPS) system from Physical Electronics. The XPS spectra of GO and FRGO are shown in Figures 3.3 and 3.4, respectively. The spectrum for GO indicates covalently attached hydroxyl (C-OH), epoxide (-C-O-C-), and carboxylic (-O-C=O) oxygen groups at ev, ev, and ev, respectively, along with sp 2 hybridized C-C bonds at ev. The XPS spectrum of FRGO shows that the intensity of oxygen-containing functional groups is significantly reduced and a new functional group appears at ev, corresponding to C-N bonds that appear due to the functionalization of GO with ODA. These results are also in good agreement with previous reports [68]. Figure 3.3: XPS spectrum of graphene oxide. The black circles represent raw data. The black line is fitted sum and the colored lines are fitted peaks using the software OMNIC for Almega 7. 43

56 Figure 3.4: XPS spectrum of functionalized reduced graphene oxide. The black circles represent raw data. The black line is fitted sum and the colored lines are fitted peaks using the software OMNIC for Almega 7. 44

57 CHAPTER 4 MECHANICAL, THERMAL, AND TRIBOLOGICAL PROPERTIES OF NANOCOMPOSITES 4.1 Introduction Graphene is a material with a two dimensional honeycomb lattice with sp 2 bonded carbon atoms [1, 8, 23, 69-70]. Graphene has high mechanical strength and high electrical and thermal conductivities making it potentially useful in the fabrication of polymer nanocomposites with enhanced mechanical, tribological, thermal and electrical properties [10-13, 32, 41]. Earlier publications dealt with various carbon materials including carbon black, carbon nanofibers, exfoliated graphite and carbon nanotubes (CNTs) as fillers in polymers to improve these properties [7, 71-76]. CNTs have been shown to be particularly effective due to their high mechanical strength, and high electrical and thermal conductivities [7, 77-78]. It has been reported that the addition of 2 weight percentage (wt. %) of CNTs in epoxies doubles the Young s modulus [71]. The addition of 0.1 wt. % of CNTs increases the electrical conductivity from 10-9 Sm -1 to 10-2 Sm -1 [72], and 1.0 wt. % increases the thermal conductivity by 80% [73]. In addition, CNTs have been reported to increase the glass transition temperature (T g ) of epoxies [78]. CNTs functionalized with maleic anhydride and amino groups increase T g by 10 C and 14 C, respectively [79]. CNTs modified with nonionic surfactants increase T g by 25 C [75]. The dispersion and adhesion at the molecular level, altering the chain dynamics [80-81]. In contrast, Note: Most of this chapter is duplicated from the accepted paper with the permission of Maney Publishing: Effects of Functionalized Reduced Graphene Oxide on Frictional and Wear Properties of Epoxy Resin, R. Shah, T. Datashvili, T. Cai, J. Wahrmund, B. Menard, K. P. Menard, W. Brostow, J. Perez, Material Research Innovations, References and figures numbers are changed to accommodate the dissertation. 45

58 increase in T g has been attributed to the nano-scale dimensions of CNTs that result in good macroscopic fillers such as carbon fiber and graphite do not significantly affect T g [60, 82]. In comparison to CNTs, graphene is less expensive to produce and more miscible due to its large surface area [76]. It has been reported that the addition of 0.4 wt. % of amine functionalized graphene to epoxy increases the Young s modulus by 60% [83]. The addition of 4 wt. % of graphene and 2.5 wt. % of surface-modified graphene to epoxy increases T g by 8 C and 14 C, respectively [81]. The addition of 1 wt. % of functionalized graphene to poly (acrylonitrile) increases T g by as much as 40 C [78]. Other studies on graphene-epoxy nanocomposites have reported the effects of graphene oxide on curing [59], graphene platelets on fracture properties [81], functionalized graphene oxide on hardness, electrical conductivity and thermal properties [42], and organosilane functionalized graphene on thermal degradation and tensile strength [43]. Tribological studies of graphene-polymer nanocomposites have shown that the wear rate of polytetrafluoroethylene (PTFE) is significantly reduced by the addition of 10 wt. % of thermally reduced graphene platelets [44]. The friction and wear rate of nylon are lowered by modified graphene oxide [45]. Epoxy resin is used in various aerospace and automotive applications due to its moldability and good mechanical and thermal properties [85-87]. It would be of interest to study the tribological properties graphene-epoxy nanocomposites with the goal of lowering friction and wear rates to extend their lifetime [64, 88-90]. Here, the effects of octadecylamine-functionalized reduced graphene oxide (FRGO) on the friction and wear properties of diglycidyl ether of bisphenol A (DGEBA) epoxy are reported. The effects on the Young s modulus, T g and thermal stability of the epoxy are also reported. 46

59 4.2 Young s Modulus Measurements of FRGO-Epoxy Nanocomposites A MTS system was used to measure Young s modulus at a tensile rate of 10 mm/min with the high load limit set at 5000 lbf. Testing was performed at a temperature of approximately 23 C. Figure 4.1 shows load versus elongation curves for the neat epoxy and nanocomposites containing various wt. % of FRGO. The Young s modulus, tensile strength and strain at break for the various FRGO concentrations are shown in Table 1. Figure 4.2 shows that the Young s modulus increases with the addition of FRGO for concentrations from 0.1 to 0.5 wt. %. For 0.5 wt. % of FRGO, the Young s modulus increases by 74% from 1.23 to 2.14 GPa, while the tensile strength increases by 68%. Similar increases in the Young s modulus and tensile strength of graphene-epoxy nanocomposites have been reported and attributed to good dispersion of graphene and the strong interfacial interactions between graphene and the epoxy matrix [41, 68, 91] that results in good transference of applied stress from the matrix to the FRGO [68,80, 91]. With the addition of 1.0 wt. % of FRGO, the Young s modulus and the tensile strength decrease by 6.7% and 13 %, respectively, compared to their values at 0.5 wt. %. Such a decrease has been previously reported [91-93] and attributed to the aggregation of FRGO at high concentrations that weakens the adhesion of graphene to the matrix. It was explained that the weak adhesion reduces the stress transfer capability, what ultimately reduces the Young s modulus and tensile strength [91-93]. As shown in Table 1, a decrease in strain at break occurs as FRGO is added. A similar observation at higher concentrations of filler has been reported and attributed to the lower susceptibility of deformation of graphene compared to the matrix; thus, the filler reinforces the matrix such that it deforms less [91, 93-94]. The decreased deformation indicates an increase in 47

60 strength and stiffness, and is consistent with the tensile strength and the Young s modulus measurements [94-96]. A comparison of the experimental value of Young s modulus of nanocomposites with the theoretical results is presented in the next section. Figure 4.1: Load versus elongation curves of the neat epoxy, and nanocomposites containing 0.1, 0.2, 0.5, and 1.0 wt. % FRGO. 48

61 Table 4.1. Mechanical Properties of The Neat Epoxy and The Nanocomposites. FRGO Content Young Modulus Tensile Strength Strain at Break (wt. %) (GPa) (MPa) (%) Figure 4.2: Young s modulus of neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % FRGO. Error bars are for three samples. 49

62 4.3 Estimation of Young s Modulus Using Halpin-Tsai Model The Young s modulus of graphene-epoxy nanocomposite can be estimated from the Halpin-Tsai model, which is developed for fiber reinforced composite materials. The model can be applied for graphene-epoxy nanocomposites by assuming that graphene sheet is a fiber with rectangular cross-sectional area of width (w), length (l), and thickness (t). The Halpin-Tsai equation [96-98] for fiber reinforced composite materials is given by E c lv lV eff, fib eff, fib wv 8 1 V w eff, fib eff, fib E m (4.1) where, l E eff, fib E E m eff, fib E m Eeff, fib 1 1 Em, w, Ec is the Young s modulus of the composite material, Eeff, fib 2 E m V eff, fib is the volume fraction of fiber, eff fib E, is the Young s modulus of the fiber, and Em is the Young s modulus of the matrix. The constant depends on the geometry of the fiber, and is given by w l (4.2) t The volume fraction of the fiber is given by V eff, fib volumeof fiber volumeof composite m fib fib m fib fib (1 m m fib ) m fib m fib (1 m fib ) fib m (4.3) 50

63 assumingv Equations (4.1) and (4.3) can be applied to the graphene-epoxy composite material by can be written as V E eff, fib frgoand eff fib frgo, E. In this case, with the values of l and w, equation (4.1) E c 1 1 V V frgo frgo V V 2 frgo frgo E m (4.4) where, E E frgo m is the ratio of the Young s moduli of FRGO and the epoxy matrix. This equation can be used to estimate the theoretical Young s modulus of graphene-based nanocomposites. Similarly, the volume fraction of functionalized reduced graphene using equation (4.2) and is given by V frgo m V frgo can be calculated m frgo (4.5) frgo m frgo ( 1 m frgo) where, m frgo is the mass fraction of FRGO, frgo is the density of FRGO, and m is the density of epoxy matrix. The density of graphite is 2.25 g/cm 3 with an interlayer separation of nm. The interlayer separation for the graphene oxide is~ 0.68 nm. Therefore, frgo =1.1 g/cm 3. From the SEM observations, width (w), length (l), and thickness of FRGO are ~2.5 μm, ~2.0 μm, and ~3.4 nm (for 5 FRGO layers), respectively. The aspect ratio of the material is ~1,180. The number of layers was estimated from the ratio of the intensities of G and Si peaks of the Raman spectrum of FRGO deposited on a SiO 2 substrate shown in Figure 4.3 [99]. Similarly, 51

64 m =1.07 g/cm 3. The theoretical Young s modulus, calculated using Halpin-Tsai model (i.e., equation 4.4), for various concentrations of FRGO-epoxy composite is shown in Figure 4.4. This shows that the experimental value of Young s modulus is in agreement with the theoretical value estimated using the Halpin-Tsai model for FRGO-epoxy nanocomposites up to 0.5 wt. % of FRGO in the epoxy matrix. The decreases in the experimental value of the Young s modulus beyond 0.5 wt. % of FRGO is attributed to agglomeration of FRGO at higher concentrations in the epoxy matrix [91-93]. Agglomeration is not included in the Halpin-Tsai model. Table 4.2. The Young s Modulus of FRGO-Epoxy Nanocomposites Calculated Using Halpin-Tsai Model. FRGO Content Theoretical Young s Modulus (GPa) wt. % vol. % 3 layers 5 layers 7 layers

65 Figure 4.3: Raman spectra of FRGO deposited on a SiO 2 substrate. 53

66 Figure 4.4: Comparision of the Young s modulus estimated using the Halpin-Tsai model with the experimental Young s modulus for the nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % FRGO. 54

67 4.4 Dynamic Mechanical Analysis (DMA) of FRGO-Epoxy Nanocomposites DMA was carried out using a PerkinElmer DMA8000 apparatus. The measurements were performed in the single cantilever bending mode. Three frequencies were applied, namely 0.1, 1.0 and 10.0 Hz. The differences between the sets of results for different frequencies were not significant, hence results for 1.0 Hz are reported below. All deformations were 50 microns so that strain was well below 1%. DMA measurements provide the storage modulus E (representing the elastic energy, that is the solid-like behavior) and the loss modulus E (representing liquid-like behavior). From these one obtains tan δ = E /E. In the glass transition region (i.e., the temperature at which a substance changes from hard or brittle state to rubbery state), E dramatically decreases while E shows a maximum. The width of the glass transition region varies while for convenience that region is often represented by a single number called the glass transition temperature T g. It should be remembered that representation of a region by a single number is a large simplification [61, ]. The onset of the drop in storage modulus as a function of temperature gives the best agreement with the peak of tan δ. The peak in the loss modulus is often weak or absent in certain materials and not in general use [61,101]. The peak of tan δ was used for the location of T g. This method is in common use and gives peaks with good visibility, reproducibility and minimal dependence on the analyst [61, ]. Figure 4.5 shows tan δ for neat epoxy and epoxy with 0.5 wt. % FRGO. As shown in Figure 4.6, T g shows a maximum at 0.5 wt. % FRGO. An increase in T g implies an increase in interaction between the filler and matrix. The increase in T g is consistent with previous reports using graphene platelets as fillers [81], and indicates good dispersion of FRGO in the epoxy. The decrease in T g with further addition of FRGO may be due to the agglomeration of FRGO sheets or a reduction in cross-linking density that results in a less 55

68 stiff material [103]; a similar conclusion from the Young s modulus measurements have been already formulated above. Figure 4.5: Plot of tan δ versus temperature for neat epoxy and nanocomposite containing 0.5 wt % of FRGO. 56

69 Figure 4.6: T g for neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO. 57

70 4.5 Thermogravimetric Analysis of FRGO-Epoxy Nanocomposites The thermal stability of the nanocomposites was studied using a Perkin Elmer TGA 7 thermogravimetric analyzer (TGA). The measurement was performed in a range from 20 to 600 C at a heating rate of 20 C/min under N 2 gas environment. Figure 4.7 shows a TGA thermogram of the neat epoxy and several nanocomposites and Figure 4.8 shows expanded TGA from 330 ºC to 420 ºC. The first derivatives of the thermogravimetric curves shown in Figure 4.7 with respect to temperature are given in Figure 4.9. The weight loss for the neat epoxy and nanocomposites appears to occur in two stages as shown in Figure 4.9. In the first stage, which takes place from 180 to 330 C, both the neat epoxy and nanocomposites lose their weight by about 12 %. This is attributed to a loss of adsorbed water and oligomers [92]. In the second stage, which occurs between 330 and 530 C, a significant amount of weight loss is observed and attributed to thermal decomposition of the epoxy [42, 104]. In this region, the decomposition temperature of nanocomposite increases with the addition of FRGO and attain a maximum value for the nanocomposite containing 0.5 wt. % of FRGO. The onset temperature for decomposition is about ~16 C greater for nanocomposites with 0.5 wt. % FRGO than for neat epoxy. The improvement in thermal degradation temperature attributed to good dispersion of FRGO at molecular level in the epoxy that result in strong interfacial interaction between FRGO and the epoxy which is consistent with the observed increases in the Young s modulus [41, 68, 91]. Wang et al. have reported a similar enhancement in the thermal stability of organosilanefunctionalized graphene-epoxy nanocomposites [43]. The thermal degradation temperature of the nanocomposite containing 1.0 wt. % of FRGO decreases with respect to that for 0.5 wt. % of FRGO. The decrease at higher wt. % of FRGO is due to aggregation of FRGO sheets in the epoxy which weakens the adhesion of FRGO to the epoxy [91,93]. 58

71 Figure 4.7: Thermogravimetric analysis showing mass % versus temperature for neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO. Figure 4.8: Expanded view of thermogravimetric analysis in the temperature range o C. 59

72 Figure 4.9: The first derivative of the thermogravimetric curve with respect to temperature for neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO. 4.6 Tribological Properties of FRGO-Epoxy Nanocomposites The frictional and wear properties of the nanocomposites were studied using a Nanovea tribometer from Micro Photonics, Inc. A tungsten carbide ball with a diameter of 6 mm was used as the counter surface. All measurements were performed in air using a 15 N normal load, 80 rpm rotational speed and a circular track having a radius of 2 mm. Further details are provided below Friction Results Plots of friction versus sliding distance for the neat epoxy and several nanocomposites are shown in Figure The neat epoxy exhibits an initial friction of less than or approximately 60

73 equal to 0.1 during the first 1.5 meters of sliding distance, as shown in the expanded view in Figure After this, the friction sharply increases to In contrast, the nanocomposites exhibit sliding distances during which the friction is, again, less than or approximately equal to 0.1 that are more than an order of magnitude greater. As shown in Figure 4.10 for 0.1 wt. % of FRGO, the friction is less than or approximately 0.1 for about 44 m before increasing to For 0.2, 0.5 and 1.0 wt. % of FRGO, the friction is less than or approximately 0.1 for about 55, 61 and 93 m, respectively, before increasing to 0.43, 0.44 and 0.45, respectively. The increase in sliding distance during which the friction is low is attributed to a transfer film from the nanocomposite to the counter surface. It is well known that transfer films reduce the friction by providing interfacial sliding between the surface and counter surface [105]. Figure 4.12 (a) shows an optical microscopy image of the counter surface taken using a tungsten light source before any sliding. Figure 4.12 (b) shows an image of the counter surface after sliding on the neat epoxy after a distance of 90 m; there was a clean surface with no transfer film. Figures 4.12 (c)-(f) show images of the counter surface after sliding only in the low friction regime of nanocomposites with 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO, respectively. The images were taken after sliding a distance equal to approximately 50% of the distance at which the friction sharply increases to The transfer films was observed, some of which have fringes due to interference between the incident light and the reflected light from the counter surface. Figure 4.13 (a)-(d) show images of the counter surface after sliding in the high friction regime of nanocomposites with wt. % FRGO of 0.1, 0.2, 0.5 and 1.0, respectively. These images were taken after completion of the runs shown in Figure Transfer films are also observed in these cases, although less coverage of the counter surface is seen. 61

74 Figure 4.10: Plot of the friction versus sliding distance for neat epoxy and nanocomposites containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO. 62

75 Figure 4.11: Expanded view of the friction versus sliding distance for neat epoxy. 63

76 Figure 4.12: Optical microscopy images of the countersurface (i.e., tungustan carbide ball of diameter 6mm). (a) Clean surface. (b) Surface after sliding on neat epoxy showing no transfer film. Surface after sliding in the low friction regime for nanocomposites containing (c) 0.1, (d) 0.2, (e) 0.5 and (f) 1.0 wt.% of FRGO. Transfer films are observed. 64

77 Figure 4.13: Optical microscopy images of the countersurface after sliding in the high friction regime for nanocomposite containing (a) 0.1, (b) 0.2, (c) 0.5 and (d) 1.0 wt.% of FRGO. 65

78 4.6.2 Wear Results The wear rates of the nanocomposites in the low and high friction regimes were calculated by measuring the depth of the wear track using a Veeco Dektak 150 profilometer. The wear behavior of polymers can be significantly affected by fillers [39, 64, 90, ]. For the low friction regime, the wear track corresponding to each wt. % of FRGO was measured after a sliding distance equal to approximately 50 % of the sliding distance at which the friction sharply increases. For the high friction regime, the wear track was measured at the end of the run. The depth profiles of worn surface of various FRGO/epoxy nanocomposites are shown in Figures The worn volume V was calculated using relation V =2πR A, where R is radius of the wear track and A is the average cross-sectional area of the worn track obtained from the profilometry measurement. The wear rate W (in mm 3 /N m) was calculated using the relation W = V/(N x), where N is the normal load and x is the sliding distance. The wear rate as a function of FRGO concentration in the low and high friction regimes is shown in Figure The wear rate in the low friction regime is about 5 times lower than that in the high friction regime. This is attributed to the transfer film, which is also known to reduce the wear rate by isolating the surface from the counter surface and reducing frictional stresses [64, 90]. In the low and high friction regimes, there is a reduction in wear rate of approximately 33 % and 13 %, respectively, at 1.0 wt. % of FRGO. 66

79 Figure 4.14: Profilometer cross-section of wear tracks taken at the end of the friction measurements shown in Figure 4.7 for neat epoxy and nanocomposites containing various wt. % of FRGO. 67

80 Figure 4.15: Wear rates of the neat epoxy and nanocomposite containing 0.1, 0.2, 0.5 and 1.0 wt. % of FRGO in the high friction regimes and wear rates of various nanocomposites in low friction regime. Error bars are for three wear tracks. 68

81 4.6.3 Investigation of Wear Mechanism In order to study the wear in more detail, SEM images of the wear tracks were taken. Figures 4.16 (a) and (b) show SEM images of the wear track of the neat epoxy near the end of the low friction regime after a sliding distance of approximately 3 m. Figure 4.16 (a) shows that there are areas where the surface of the wear track has roughened. Figure 4.16 (b) shows a magnified view of a roughened area showing the formation of wear particles. Therefore, the roughening starts near the very beginning for the neat epoxy, and this facilitates the sharp increase in friction after a very short sliding distance. Figure 4.16 (c) and (d), and (e) and (f) show SEM images of the wear tracks in the low friction regime for nanocomposites containing 0.1 and 0.5 wt. % of FRGO, respectively; the images were obtained after sliding distances of approximately 20 and 35 m, respectively. For the nanocomposites, the wear tracks are significantly smoother than that of the neat epoxy even though the sliding distances are about an order of magnitude longer and wear particles are not observed. It has been reported that a transfer film diminishes the formation of wear particles [104]; this is consistent with our observations. Figures 4.17 (a) and (b), (c) and (d), and (e) and (f) show wear tracks of the neat epoxy and nanocomposites with 0.1 and 0.5 wt. % of FGRO, respectively, in the high friction regime at the end of the run. All of the wear tracks are rough, consistent with the significantly higher friction and wear rate in this regime. The wear track of the neat epoxy is very rough and ploughed with loosely bound chunks of debris. Similar observations have been reported for epoxy [39, 105]. The debris has been attributed to the formation and expansion of surface and subsurface cracks due to repeated loading [44, 105, 109]. As shown in Figures 4.17 (c)-(f), when 0.1 and 0.5 wt. % of FRGO is added, the wear tracks becomes progressively smoother with less 69

82 debris. Other studies [105] have reported that the addition of fillers in an epoxy matrix forms a protective layer that reduces the wear rate. Dang et al. have reported similar observations with CNTs in an epoxy matrix and proposed that these fillers diminish the adhesion between the matrix and the counter surface; the ploughing phenomenon is thus reduced, and this results a relatively higher wear resistance [39]. Figure 4.16: SEM images of wear tracks in the low friction regime of (a) and (b) Neat epoxy. (c) and (d) Nanocomposites containing 0.1 wt. % of FRGO. (e) and (f) Nanocomposites containing 0.5 wt. % of FRGO. 70

83 Figure 4.17: SEM images of wear tracks in the high friction regime of (a) and (b) Neat epoxy. (c) and (d) Nanocomposites containing 0.1 wt. % of FRGO. (e) and (f) Nanocomposites containing 0.5 wt. % of FRGO. 71