Chapter 1 Introduction 1

Chapter 1
Introduction
1.1 Carbon nanomaterials
Carbon is a universal material that has been ever found. Single carbon atom has six electrons with 1s2, 2s2 2p2 atomic orbital configuration. Even pure carbon can have quite a few allotropes. This is because the four valence electrons can make different types of bonds with other atoms. The special electronic configuration of carbon leads to sp (e.g., Acetylene) that are separated by an angle of 180°, sp2 (e.g., fullerenes C60, carbon nanotubes, graphite, graphene) that are coplanar separated by 120°, or sp3 (e.g., diamond) hybridization forms that are separated by 109.5°. Fig. 1.1 shows different allotropes of carbon.

Fig 1.1. Allotropes of carbon (a) diamond, (b) graphite, (c) fullerenes C60, (d) carbon nanotubes, (e) graphene
Carbon Nanotubes and Graphene have received significant attention from both the academia and the industry due to their exceptional properties.

1.2 Carbon nanotubes (CNTs)
Since the accidental discovery by Sumio Iijima of the NEC Corporation in the soot of the arc-discharge method in 1991, carbon nanotubes (CNTs) have generated a great deal of research in most areas of science and engineering due to their attractive physico-chemical properties: stronger than steel, harder than diamond and electrical conductivity higher than copper .

A carbon nanotube is a hollow tube made from pure carbon in a hexagonal arrangement. In this hexagonal order, each carbon shares a sp2 covalent bond with its three neighboring atoms. The bond distance is approximately 1.5 A°. CNTs have a typical diameter of 1-5nm, about 100,000 times thinner than an average human hair and they can be several hundred microns long.
Carbon nanotubes (CNTs) exhibit a remarkable set of electrical, mechanical, optical and thermal properties that offer opportunities for materials design in many research areas and commercial products, .
1.3 Graphene
Graphene is a two-dimensional, single-layer sheet of sp2 hybridized carbon atoms. It has attracted enormous attention and research motives for its different properties. In sp2 hybridized bond, the in planer C–C bond is one of the strongest bonds in materials and the out-of-plane is p bond, which imparts to a delocalized network or array of electrons resulting electron conduction by providing weak interaction among graphene layers or between graphene and substrate. Graphene is a material with a large theoretical specific surface area (2630 m2g-1), high intrinsic mobility (200,000 cm2v-1s-1), high Young’s modulus (?1.0 TPa) high thermal conductivity (?5000 Wm-1K-1) , and good electrical conductivity (ability to stand current density of 108 A/cm2 ). Graphene is a very special material because of its unique properties, high hopes have been placed on it for technological applications in many areas .
1.3.1 Graphene oxide (GO)
During the measurement of the atomic weight of carbon, Benjamin Brodie (1859) achieved some of the primary experiments on the chemical properties of graphite . He claimed that significant oxidation of graphite by exposing it to a mixture of fuming HNO3 and KClO3 solution for several days, could be happened. Finally, at the end of his experiment, he called a light yellow color product, as “graphic acid” which we today call graphite oxide. After about a century later, Hummers and Offeman (1958) found a significantly safer synthesis of graphite oxide using a mixture of graphite with NaNO3, and KMnO4 in concentrated H2SO4 .
GO as the same as graphene has been explored in a wide range of applications, such as electronic and photonic devices, drug delivery, energy generation/storage, optical devices, clean energy, and chemical/bio sensors, electrochemical application , .

1.3.2 Reduction of graphene oxide
Different reduction methods such as thermal reduction, photo catalyst reduction, solvothermal/hydrothermal reduction, microwave and photo reduction have been reported in order to provide reduced graphene oxide (rGO) . Chemical reduction of GO is a suitable method to synthesized reduced graphene oxide (rGO) and graphene in large quantity. (see section 4.1.2)
1.4 Characterization of carbon-based nanoparticles
Carbon nanotubes and graphene due to their specific atomic structure, have special physical and chemical features. The size, structure, morphology and composition of synthesized pristine and doped carbon-based nanoparticles and their nanocomposites are characterized and analyzed by different methods such as electron microscopy observations, which includes Scanning Electron Microscopy (SEM), Transmission Electron Microscopy (TEM), and Atomic Force Microscopy (AFM). The spectroscopic analysis like X-Ray Diffraction (XRD), Raman and Infrared (IR) spectroscopies are also useful to characterize the morphology and structure of these nanoparticles. Thermal analysis such as Thermogravimetry Analysis (TGA), Differential scanning calorimetry (DSC) and Dynamic Mechanical Thermal Analysis (DMTA) are very helpful to analyze the combination of nanocomposites based on carbon nanoparticles.
1.4.1 Electron microscopy observations
Electron microscopy is one of the essential tools for characterizing any nanomaterial according to the direct shape, size, and structure observation. The local structure and surface of the carbon nanotubes can be analyzed at the nanometer level by these techniques. Boehm et al in 1962 observed the single layers and multilayers of colloidal graphite oxide by electron microscopy .

Scanning Electron Microscopy (SEM), Transmission Electron Microscopy (TEM), and Atomic Force Microscopy (AFM) are useful tools to check the exfoliation of bundles and the purity of the CNTs. However TEM and SEM produce damages on the sample due to the use of the electron beam.
AFM is used to study the sample in three dimensions and facilitates forming 3D images of a sample surface with high lateral spatial resolution (0.1 – 1.0 nm), as well as atomic scale vertical resolution (0.01 nm). Any special sample preparation is required for AFM so it can be used in aqueous solution or vacuum ambient .

1.4.2 Spectroscopic analysis
1.4.2.1 X-ray Diffraction (XRD)
X-rays are electromagnetic (EM) waves with typically shorter wavelength about 1Å (1×10-10m) which were immediately after discovery (Roengton in 1895) applied to elucidate the inner structure of crystalline solids.
The utilizing of scattered X-rays from material to investigate properties mentioned above called as X-ray diffraction (XRD). XRD is a powerful non-destructive bulk technique for analyzing a wide range of materials from research to industry in both the local and global features such as lattice structure with phase morphologies and interlayer distances.

Rezni et al. in 1995 for the first time characterize CNTs structure by utilizing the powder XRD technique .
This technique is based on measuring the intensity and the angle of scattered X-rays from electrons bound to atoms (sample material) as a function of atomic position. Depending on the geometry of the crystal lattice, initial waves scattered by atoms at different positions and angles. The simplest and most frequently used application of scattering theory is the Bragg law. By considering crystals as reflection layers for X-rays, W.H. Bragg derived the following equation:
2dSin?= n?
where ? is the wavelength of the X-rays, 2? is the scattering angle, n is an integral number and d represents the distance between successive identical planes of atoms in the crystal. A visualization of Bragg law represented in Fig.1.6. Therefore, it is possible to perfectly measure the CNTs’ crystallinity by XRD.
Fig. 1.6. Diagram of Bragg’s law
1.4.2.2 X-ray photoelectron spectroscopy (XPS)
X-ray photoelectron spectroscopy (XPS) is one of the best tools which can give some information about the chemical structure of CNTs and GO due to the chemical interaction with organic compounds or gases adsorption .

The distribution and bonding of heteroatom dopants in carbon nanomaterials (CNTs, GO) can be particularly shown by appears new peaks in XPS according to the bonds formed between carbon atoms and compounds added .

1.4.2.3 Raman spectroscopy
Raman spectroscopy provides a powerful, fast and non-destructive spectroscopic characterization tool for carbon-based nanomaterials, showing different characteristic spectral features for sp3, sp2, and sp carbons. All carbon’s allotropes forms are Raman spectroscopy active . Each band in the Raman spectrum corresponds to a specific vibrational frequency of a bond within the molecule. Raman spectra in the range of 1000 to 1700 cm-1 appears for carbon nanotubes and graphene .

The Raman spectrum of SWCNT consists of three dominant Raman-allowed bands (Fig 1.7) ,:
(i) A high- frequency bunch called G-band (~1600 cm-1) corresponding to a splitting of the C-C band stretching mode of graphitic carbon atoms in the nanotubes (Carbon sp2)
(ii) Disorder line (D-band) is derived from defects (from impurities or disorders) in the carbon network appears at ~1300 cm-1
(iii) A bunch of peaks for poly-disperse samples when resonating conditions are met called as radial breathing mode (RBM) (< 200 cm-1 )
(iv) G` (or 2D) band is overtone of D band depends on the diameter of CNTs.
MWCNT have similar Raman spectra to those of SWCNT. The basic differences are the lack of RBM band (the outer tubes restrict the breathing mode) and a more obvious D-band (the multilayer configuration indicates more disorder in the structure) in Raman spectra of MWCNT .

The quality (crystallinity) of CNTs can be measured by comparing the D to G band intensity. If the relative intensity of G and D bands is greater than 9:1, the spectra indicate high purity CNTs.
Raman spectroscopy is also carried out on graphene. It is the fastest and most precise method of identifying the thickness of graphene flakes and estimating its crystalline quality. This is because graphene exhibits characteristic Raman spectra based on number of layers present .

In addition, the 2D (D`) peak is common to all graphite samples, and its width, intensity and location can be used to find out the number of sample layers .
Fig. 1.7. General Raman spectra of carbon base nanoparticles
1.4.2.4 Fourier transform infrared spectroscopy (FTIR)
Fourier transform infrared spectroscopy (FTIR) is a complementary technique to Raman spectroscopy to characterize the graphitic nature of the carbon based nanomaterials. FTIR is sensitive to the vibrations and polar bonds of hetero-nuclear functional group and normally used to determine impurities remaining from synthesis or molecules capped on the CNTs surface . This method is not a suitable technique to distinguish SWCNT from MWCNT, because the IR-active modes in these nanoparticles are the same as the frequencies of graphite at 868 and 1590 cm-1 .

CNT-IR absorbance spectra (Fig 1.8) clearly show IR-active peaks centered at :
(i) At 1584cm?1 (G band) corresponding to the graphite-like E1u mode originating from the sp2-hybridized carbon
(ii) A double structured absorbance peak at ?1200cm?1 is a disorder-induced one phonon absorbance band related to the sp3 bonded carbon (D band).
(iii) The presence of the several peaks at ?3000cm?1 range is an evidence for strong structures near the graphite modes .

Fig 1.8. A typical CNT-IR absorbance spectrum
1.4.3 Thermal analysis
1.4.3.1 Thermogravimetric analysis (TGA)
Thermogravimetric analysis (TGA) is an experimental technique that measures the weight changes which occur during the heating the sample. This method can be used to determine the thermal stability of a material as well as the presence of volatile components in a sample. TGA can be also used to measure the purity of carbon nanotubes as these materials undergo complete loss at elevated temperatures.
In addition, many studies have been observed the importance of studying the thermal properties particularly with TGA of nanocomposites containing CNTs and graphene oxide since a significant enhancement in thermal stability of the polymeric matrices filled with the carbon nanomaterial, compared to unfilled ones .

1.4.3.2 Differential scanning calorimetry (DSC)
Differential scanning calorimetry (DSC) is the most broadly used of the thermal analysis techniques in different research areas, development, and quality inspection and testing. By performing DSC over a large temperature range, thermal effects can be quickly identified. The relevant temperatures and the characteristic caloric values can be determined using substance quantities of only a few mg. The measured properties by DSC during a thermal scan of nanofillers (CNTs or graphene) include heat capacities, heats of transitions, kinetic data, sample purity, the degrees of crystallinity, and temperatures of glass transitions (Tg) .
1.4.3.3 Dynamic Mechanical Thermal Analysis (DMTA)
DMTA was frequently used in nanocomposites characterization since it allows the measurement of the dynamic storage modulus (E?), dynamic loss modulus (E?), mechanical loss angle tangent (tan ?= E?/E?). The tan ? illustrates the macromolecules mobility as well as the phase transition in the polymers. Furthermore, DMTA is one of the most used techniques for measuring the glass transition temperature of carbon based polymeric materials (Tg), since it is fast and suitable for quality control applications .

1.5 Outline of the thesis
In this dissertation, we will explore synthesize of doped carbon nanotubes and graphene-based nanomaterials and application of synthesized nano-compound in different fields.
The dissertation is divided into five core chapters. Chapter 2 will be reviewed the synthesis and characterization of pristine carbon nanotubes (CNT), oxygen doped carbon nanotubes (COx) and nitrogen doped carbon nanotubes (CNx) via CVD method. In the chapter 3 based on Paper I, preparation method and characterization of epoxy nanocomposites based on synthesized carbon nanotubes (CNT, COx, CNx) will be shown and flame retardant properties of prepared nanocomposites will be discussed.
Chapter 4 will be represented the microwave assisted copper decoration of graphene oxide in different solvents and characterization techniques for the products. Study about the performance of copper salts decorated graphene oxide in photodegradation application will be explained, respectively. In the chapter 5, our investigation about catalytic application of copper decorated graphene oxide in different organic reactions will be discussed.

Finally, in the chapter 6, some future plans for continue this work will be introduced.
Chapter 2
Synthesis of Oxygen and Nitrogen doped carbon nanotubes
Background
2.1. Classifications of CNTs
Nanotubes are considered as nearly one-dimensional structures according to their high length to diameter ratio. There are two types of CNTs, called single walled CNTs (SWCNT) and multi walled CNTs (MWCNT) as shown in Figure 2.1 The first-discovered carbon nanotube (Iijima, 1991) was in a multi walled form and single walled tubes were found two years later . A SWCNT with their well-defined atomic structure, their high length to diameter ratio, and their chemical stability constitute one-dimension molecules can be considered as a hollow cylindrical structure of pure carbon with a diameter ranges from 0.5 nm to about 10 nm and a length from a few microns to centimeters. Multi-walled carbon nanotubes (MWCNTs) has multiple concentric cylindrical walls with spacing between walls equivalent to the interlayer spacing of graphene sheets (~3 Å) and are typically larger than SWCNT (Fig 2.1). The length and diameter of these structures differ a lot from those of SWNTs and, of course, their properties are very different.
Fig 2.1. (a) Single walled carbon nanotube (SWCNT), (b) Multi walled carbon nanotube (MWCNT)
Graphene is a poly-aromatic monoatomic hexagonally arranged layer of carbon atoms with hybridization of sp2. SWCNT consists of rolling graphene sheets into cylinders with fullerene cap in suitable diameter at their ends . Depends on the rolling-up the graphene sheets and how the hexagons are orientated along the axis of the tube, different designs of SWCNT could be existed (Fig. 2.2).
Three different categories are known as armchair, zigzag and chiral form, which show different electronic properties. An armchair form with conductivity better than copper is the most commonly occurring SWNT . The chiral and zigzag forms of nanotubes are similar to semiconductors in sharing electrical properties.
Fig. 2.2. Illustration of three different SWCNT designs (a) armchair (b) zig-zag, and (c) a helical (chiral) type carbon nanotubes
2.2 Synthesis of carbon nanotubes
Nowadays, CNTs can be synthesized in large quantities by three principal methods: arc discharge, laser ablation and chemical vapor deposition (CVD). In the following, the three CNTs growth methods are described in detail.

2.2.1 Arc-discharge method
Arc discharge was originally used for producing fullerenes. Iijima performed a closer study of the carbonaceous soot deposited on the cathode, and discovered cylindrical fullerenes in the deposit . A schematic diagram of the arc-evaporation apparatus for producing CNTs, which is a vacuum chamber filled by Helium at a pressure of 0.5 atmosphere, is shown in Fig. 2.3 A pair of electrodes (at least one electrode is made of graphite) are located vertically in range of 1-2 mm distance connected to the fairy low voltage supply (10– 30 V). The arcing leads to an evaporation of the graphitic anode moreover, as the carbonaceous material cools down, it precipitates in the form of nanotubes along with the other carbon byproducts like fullerenes and soot, either on the cathode, anode or on a special collector. When pure graphite electrodes are used, the carbon nanotube product will be MWCNTs . If the carbon anode is enriched with catalytic transition metal (e.g. Ni, Co, Fe …), the product also contains SWCNT 35. The most advantage of this method is selectivity between SWCNT/MWCNT productions . Furthermore, the nanotubes synthesized by this means have high crystallinity and yield, as well. The drawback of the mentioned method is the resulting randomly oriented powder with significant amounts of carbonaceous side products that require subsequent purification .
Fig. 2.3. Schematic diagram of CNT formation apparatus by the arc-discharge method
2.2.2 Laser ablation
Laser ablation or pulsed laser vaporization method had been used in 1952 as a source of clusters and ultrafine particles . In 1996, Thess et al. reduced high yields (>70%) of CNTs by means of the laser ablation of graphite rods Smalley’s group developed the laser ablation/ vaporization to produce fullerene and carbon nanotubes production with small amounts of Ni and Co. at 1200? for the first time . Fig. 2.4 shows a typical laser- furnace apparatus.

This allows the tube to terminate with a fullerene like tip or with a catalyst particle. In this method, a target consisting of a mixture of graphite and some transition metal such as Fe, Ni or Co is bombarded (ablated/ vaporized) using a high-energy laser light at high temperatures (?1200°C) inside a tube furnace. The flowing Ar gas transfers the vaporized carbon to the end of the tube . As the vapor condenses, CNTs are formed. The CNTs diameter distribution can be differed by modifying temperature of ablation and changing in the composition of catalyst and Ar flow rate. In this method the same as arc discharge method, side products such as fullerenes, amorphous carbon, graphite particles, and graphitic polyhedrons included metal particles are also formed
Extensive purification of the products is required to produce pure CNTs materials and the process yields only small batch quantities. 41.

Fig. 2.4. Schematic diagram of the laser-furnace apparatus 41
2.2.3 Chemical vapor deposition (CVD)
Chemical vapor deposition (CVD) is thermally decomposed in the presence of a metal catalyst, which is of interest for this study, was discovered for nanotube synthesis for the first time in 1996 . The method is also known as thermal or catalytic CVD to distinguish it from the many other kinds of CVD used for various purposes. CVD is capable of controlling the growth direction on the substrate and synthesizing a large quantity of nanotubes .

CVD synthesis (Fig 2.5) is achieved by taking a hydrocarbon species in the gas phase and using an energy source, such as a plasma, piezoelectric, or a resistively heated coil, to impart energy to a gaseous carbon molecule. Common hydrocarbon sources are methane, ethylene, carbon monoxide and acetylene. The energy source is used to crack the molecule into a reactive radical species. These reactive species then diffuse down to the substrate, which is heated and coated on the catalyst surface (usually a first row transition metal such as Ni, Fe, or, Co). CVD allows the scientist to avoid the process of separating nanotubes from the carbonaceous particulate that often accompanies the other two methods of synthesis. Excellent alignment, as well as positional control on the nanometer scale, can be achieved by the use of CVD. Control over the diameter and simultaneously the growth rate of the nanotube can also be maintained. Reported temperatures for the synthesis of nanotubes by CVD vary somewhat, but are generally within the 650-800 °C range. It has been shown that the size and material of the catalyst particles play a vital role during synthesis.
Fig. 2.5. Schematic diagram of a CVD process for CNTs synthesis
In comparison with arc-discharge and laser methods, CVD is a simple and economic technique for synthesizing CNTs in a single step process, without further puri?cation at low temperature and ambient pressure without prior preparation of substrates . It is versatile in that it harnesses a variety of hydrocarbons in any state (solid, liquid, or gas), enables the use of various substrates, and allows CNTs growth in a variety of forms, such as powder, thin or thick films, aligned or entangled, straight or coiled, or even a desired architecture of nanotubes at predefined sites on a patterned substrate. It also offers better control over growth parameters. Another important advantage of CVD, which is important in our work, is the ease for different gas composition to change the combination of synthesized CNTs and prepare doped CNTs with nitrogen and oxygen.

The chief drawback of CVD which is not important in our research is the high defect density of the obtained CNTs owing to low synthesis temperatures, compared with arc discharge and laser ablation. As a result, the tensile strength of the CNTs synthesized by CVD is only one-tenth of those made by arc discharge .

2.3 Carbon based nanocomposites
Using of carbon nanotubes or graphene without any supporting medium is not applicable. In the other word, many of the noticeable properties of carbon nanotubes (CNTs) can be the best obtained by incorporating the nanotubes into some form of matrix to make different composites. Carbon based nanocomposites are made up of a polymer matrix and carbon nanotubes (CNTs), carbon fiber, graphene, graphene oxide (GO) or reduced graphene oxide (rGO) as the fillers.
Due to the unique structure of CNTs and graphene, they have incredible potential to be used as nanofillers for many structural and functional materials, particularly in modifying the electrical, thermal, mechanical, optical, photoelectrical … properties of carbon based nanocomposites , .
Because of the nanometer size and high aspect ratio, the Van der waals attraction between each set of tubes, carbon nanoparticles have high surface energy and a significant tendency for agglomeration and bundling.
In addition, pristine CNTs with special characteristics like highly hydrophobic surface area, week binding capability and low solubility in organic solvent is almost nonreactive in chemical reactions.
To employ CNTs and graphene as effective reinforcement in polymer nanocomposites, proper dispersion and appropriate interfacial adhesion between the carbon nanoparticles and polymer matrix have to be guaranteed.

2.3.1 Distribution vs. Dispersion
Uniform distribution and a good dispersion of the CNT and graphene is needed for the high performance of hybrid composites. Dispersion refers to the allocation of the nanoparticles within the matrix, whereas the distribution indicates the breaking of the aggregates into small sizes. A sufficient distribution does not inevitably implies a good dispersion, and vice versa. Fig. 2.6 (a) illustrates a case of poor distribution and poor dispersion, (b) good dispersion but poor distribution, (c) good distribution but poor dispersion and (d) good distribution and good dispersion .

Usually, it is difficult to achieve a good dispersion of CNTs due to their large surface area; however, there are techniques available to break up these agglomerations.
Fig. 2.6. shows of (a) poor distribution and poor dispersion, (b) poor distribution but good dispersion, (c) good distribution but poor dispersion and (d) good distribution and good dispersion.

2.4 Methods for dispersing CNTs
Some different techniques to disperse the CNTs aggregates into finer structure to enhance the desired properties of the composites belongs to two main categories: physical and chemical techniques.

The most convenient physical approaches for dispersing are ball milling, high shear mixing and ultrasonication. Physical or mechanical methods can only break up agglomerates into smaller parts or single-agglomerates so the dispersion quality is often inadequate .
Chemical methods can satisfactory provide separated individual tubes in a stable suspension includes functionalization and doping. In the following, more detail will come about these methods.
2.4.1 Functionalization of Carbon nanotubes
The CNTs functionalization could be an effective way to replaced modified surface nanotubes with original one to improve their various applications.

The most advantages of the surface modification with a variety of functional groups include large surface area, high bundling tendency, and good ability to disperse, which lead to potentially considerable CNTs as building blocks for hybrid nanomaterials . The CNTs functionalization methods involves two main strategies: covalent or non-covalent interactions.

2.4.1.1 Non-covalent functionalization of CNTs
The non-covalent functionalization of CNTs could be achieved by ?–? stacking interactions between conjugated molecules and the graphitic sidewall of CNTs . The non-covalent modifications of CNTs can preserve their desired remarkable properties and improving their solubility, simultaneously. Interaction between CNTs and some polymers (wrapping with polymers) such as polyphenyl ethers (PPE) , polyethylene glycol (PEG) , and polyvinylpyrrolidone (PVP) are good examples to affect dispersion and stability of nanotubes in polymer matrix.

The non-covalent interaction methods of CNTs do not destroy the intrinsic sp2-hybridized conjugated system of the nanotubes walls, so the properties of the final structure of the material can be preserved. This method has another important feature that makes it even more interesting, which is the reversibility. Actually, it is possible to recover the bare CNTs from the CNT-coated both in aqueous or organic solvents .

2.4.1.2 Covalent functionalization of CNTs
In this group, functionalization is based on the covalent chemical bonding between carbon atom of CNTs surface and functional groups via chemical reactions. This method in compare to the non-covalent strategies is more effective, and is better to control.

Because of the extra strain in the cap region of CNTs, the carbon atoms in the caps are highly reactive than atoms on the sidewall . Therefore, according to the location of functional groups, covalent modification can be classified into two main groups including to the sidewalls and defect functionalization of the CNTs ends (caps).

I) Sidewall functionalization
A change in hybridization from sp2 to sp3 and a loss of p-conjugation system on graphene layer simultaneously resulting from sidewall functionalization.
The first sidewall functionalization by exposing pristine CNTs to a fluorine containing gas at room temperature reported . Since C-F bonds in the fluorinated CNTs are weaker than those in alkyl fluorides, the functionalized CNT can provide substitution sites for additional functionalization. Zhang et al. (2004) have reported successful replacements of the fluorine atoms by and hydroxyl groups
Appearing a large number of defects on the CNTs sidewalls during the functionalization reaction and changing the carbon hybridization from sp2 to sp3 are the major drawbacks of this method.
II) Defect functionalization
Oxidative damaging of CNTs is one of the most common and effective example of this series. Oxidant can open the ends and convert the capped CNTs into open fullerene pipes. Finally the free ending radicals will be stabilized by carboxylic acid (-COOH) or hydroxylic (-OH) groups bonding .
Typical functionalization could be occurred by exposing CNTs framework to a strong acid, mixture of acids like HNO3, H2SO4 , strong oxidants such as KMnO4 , ozone , hydrogen peroxide and so on. Functionalization of CNTs by this method trend to open the tubes and to generate oxygenated functional groups such as carboxylic acid, alcohol, ketone and ester groups that serve too many different types of chemical moieties onto the ends and defect sites of these tubes.

The CNTs produced by this method possess many functional groups such as polar or non-polar groups so the solubility of the functionalized nanotubes in various organic solvents is good.
2.5. Doping
According to the definition, doping is the phenomena of introducing either non-carbon atoms or molecules and compounds (impurities) into the layered sp2 carbon nanosystems in different manners at small concentrations (from parts per million to small weight percentages) . There are three main categories of doping: exohedral (or intercalation), endohedral (encapsulation or filling) and inplane (or substitutional) doping (Fig. 2.7).
Fig 2.7. Schematic molecular model of (a) endohedral; (b) exohedral, and (c) in-plane doping in MWCNT bundles
Doping is one of the effective approach that allows for the intrinsic modification to the electrical and chemical properties of graphene and CNTs . The first doping reactions by K and Rb were performed on MWCNTs prepared by the electrical arc-discharge method . So far today, several chemical modification and doping strategies have been reported for CNTs and graphene.

Iodination or bromination are known to specifically affect CNT, since such doping can produce ?-type semiconductors .
Nitrogen doped CNTs with bamboo-type structure in a low nitrogen concentrations were generated via pyrolysis of pyridine and methylpyrimidine for the first time .

There are currently several methods of doping CNT with various substances: nanotubes doping during their growth (i.e. in situ methods); gas or liquid phase incapsulation in the cavity of preliminary formed CNT and chemical modification of CNT surface (i.e. ex situ methods). In situ methods include CNT modification during arc-discharge synthesis and CVD method.
In this chapter, CVD method used to grow doped CNT is described in closer detail. The effect of three types of precursor combination on produce CNTs powder by characterization of the synthesized nanotubes are explained. According to the characterization methods, (XRD, XPS, SEM, TEM, Raman spectroscopy, and TGA), pristine carbon nanotube, oxygen doped carbon nanotubes and nitrogen doped carbon nanotubes that are abbreviated to CNT, COx and CNx, respectively were well defined.

Experimental procedures
2.6 Synthesis of carbon nanotubes by CVD
All materials and the solvents were purchased from Aldrich chemical company and used as received.
2.6.1 Synthesis of CNT
MWCNT synthesized via CVD method by starting from a solution consisting of 95% Toluene and 5% Ferrocene, The precursor initially sonicated for 20 minutes. Argon gas with a flow of 2.5 L.min-1 for 30 minutes at 850 ºC were used in our single furnace CVD system.

Total obtained weight was in the range 0.9 – 1.7 g in each synthesis.

2.6.2 Synthesis of COx
In the case of COx, a different solution containing 94% Toluene, 1% Ethanol and 5% Ferrocene was used. The time and temperature in CVD system was the same as CNT processing. Finally, 0.8 – 0.95 g total weight was obtained in each synthesis.

2.6.3 Synthesis of CNx
The procedure to obtain CNx was the same as previous ones but using an initial solution containing 5 wt.% of ferrocene (FeCp2) in benzylamine (C7H9N). Benzylamine used as a nitrogen defect source. An approximate total weight of 1.2 g was obtained per each synthesis
All of these synthesized nanotubes, COx, CNx and CNT were purified and improved their epoxy dispersibility by H2O2 –UV method described in our previous work .

2.7 Characterization of nanotubes
2.7.1 X-Ray Diffraction (XRD) and Raman spectroscopy
XRD was collected in an automatic X’Pert Philips diffractometer using a Cu source. By testing different step size and counting time, we decided to collect the data in the 2? range from 10° to 90° in step-scanning mode with a step size of 0.02° and a counting time of 2 s per step.

Raman measurements were carried out via Renishaw confocal microscope based Raman spectrometer using the 514.5 nm laser excitation. For each sample, various spectra were recorded in different places in order to verify the homogeneity of the sample (Alcala University).

Raman characterization and XRD analysis (Fig. 2.8. (a) and (b) were used to know the quality of crystallinity and defect of synthesized carbon nanotubes. According to these results, CNTs has a higher degree of crystallinity in comparison with COx and CNx which is in agreement with TGA results (section 2.7.4).

Fig.2.8. (a) Raman spectroscopy and (b) XRD of CNT, COx and CNx
2.7.2 X-ray Photoelectron Spectroscopy (XPS)
XPS data were recorded with an Omicron spectrometer equipped with an EA-125 hemispherical electron multichannel analyzer and an unmonochromatized Mg K? X-ray source operating at 150 W with pass energy of 50 eV. The recorded spectra were analyzed using CASAXPS software, and RSF database by peak fitting after Shirley background correction.
Based on the XPS results (Fig. 2.9 and Table 2.1), 99% of carbon atoms of CNT are either aromatic or aliphatic and present the highest percentage of C=C bonding compared to doped nanotubes. For COx and CNx the fraction of oxidized carbon atoms is about 15% being the amount of N atoms in the structure of CNx of 1.9%. All these doping provide great reactivity and solubility which play an important role in the preparation of polymer nanocomposites and to achieve enhanced interfacial interaction between nanotubes and polymers (chapter 3).

Fig. 2.9 XPS of: CNT, COx and CNx
Table 2.1. XPS analysis of CNT, COx and CNx
2.7.3 SEM and TEM
SEM was performed using a FEI equipment, with a voltage of 10 kV and a secondary electron detector. The nanocomposites samples were pre-sputter-coated in gold to the observation.

Fig. 2.10. shows SEM images of different types of CNTs with their typical tubular structure. Dimensions (length and diameter) were determined over 500 nanotubes for each sample; histograms of lengths and diameters are presented in Figs. 2.11 and Fig. 2.12, respectively. The averages are summarized in Table 2.2. The lengths of the nanotubes were ?136 m, the diameters ranged between 60 and 90 nm and COx nanotubes presented the highest aspect ratio near 3000.

It can be observed the tendency of nanotubes to present a notably higher length to diameter ratio. This is not surprising since it has already been reported that oxygen in the gas feed (ethanol) helps to keep the Fe catalyst clean from carbon agglomeration and active for longer periods, thus resulting in longer nanotube .

Fig. 2.10. SEM image of: a) CNT, b) COx and c) CNx at two different magnifications
Fig. 2.11. Histograms of lengths of: a) CNT, b) COx and, c) CNx
Fig. 2.12. Histograms of diameters of: a) CNT, b) COx and, c) CNx
Table 2.2. Dimensions of CNT, COx and CNx.

TEM images show layers of CNT, COx and the characteristic stacked bamboo-like tubules for CNx are presented in Fig. 2.13. It should be noted the surface roughness found for COx.

Fig. 2.13. TEM image of: a) CNT; b) COx; and c) CNx
2.7.4 Thermogravimetric Analysis (TGA)
TGA was performed in a TGA Q50 (TA Instruments) system heating from room temperature to 800 ºC. Approximately 5 mg of sample were heated in an open Pt crucible at a rate of 10 ºC min-1 under N2 (90 mL min-1). Experiments in air atmosphere were also run for more survey.

Fig. 2.14. Thermal stabilities of CNT, COx and CNx in air and N2 atmospheres
Table 2.3. % residual mass at 700 ºC, temperature at maximum loss weight rate, temperature at 5% loss weight and maximum weight loss rate (air)
The thermal stability of synthesized nanotubes was checked by TGA in air and N2 atmospheres (Fig. 2.14 and Table 2.3). CNT presented the highest thermal stability and the lower values of the degradation temperature for COx and CNx were attributed to the presence of defects induced by doping and considerably more edge plane sites 75 that facilitate high temperature oxidation. Moreover, among the doped nanotubes, COx presents the highest weight reduction rate at the lowest temperature reflecting that this kind of nanotubes are much more reactive than CNT or CNx; intuitively, the surface roughness and the surface functionalities of these tubes may contribute to increase its reactivity. However, these doped nanotubes are more thermally stable than oxygen and nitrogen functionalized nanotubes obtained by chemical treatments of pristine CNT either in solution or in gas phase Paper I.

Chapter 3
Study the effect of CNT, CNx and COx on flame retardancy of epoxy nanocomposites
3.1 Introduction
Epoxy resin (EP) is a one of the most important thermosetting polymers (EP) with low-molecular-weight containing more than one epoxide group, which are cured using a wide variety of curing agents (amines, anhydrides, etc.).
Due to the significant characteristics of EPs such as high tensile and impact strength, good fatigue resistance, micro cracking resistance, chemical and corrosion resistance, excellent electric insulation and low manufacturing cost, they are widely used as advanced matrix in different fields of electrical, electronic, aerospace industries and anticorrosion laminate coatings .

The very high flammability property of EP has greatly limited the development and application of epoxy based. Therefore improving the flame retardant (FR) properties of EPs has developed automatically and has become one of the most attractive subject between researchers in advanced application. Halogenated compounds (also known as organohalogen FR) containing Cl or Br bonded to C, were traditional filler to improve the FR properties of EP. Chemicals with Organohalogens are considered Persistent Organic Pollutants (POPs) and present significant risks to human health and environment. Nowadays, a wide variety of chemicals compounds containing N, P, B and Si, have attracted much attention to replace traditional toxic halogenated FR to improve flame retardancy of polymer materials , . The high loadings needed of these new additives is the major problem, which cause degradation of the mechanical properties of the final composites.
The growing up of the nanocomposite technology has proposed nano-fillers in a very low loading (<10 wt.%) as a new replacement flame retardant polymer materials. Nano-materials such as clays, montmorillonite, double layered hydroxides (LDH), etc. have been introduced into epoxy aiming to increase FR , .

According to the amazing properties of carbon nanoparticles (CNTs, graphene, GO), they have attracted special attention by researchers to utilize as FR in carbon based nanocomposites. Kashiwagi et al. reported improving the flame retardancy of by adding CNT for the first time .
The effects of CNTs will be affected by aggregation of the nano-fillers due to the strong Van der Waals forces. Therefore, high dispersion state of the CNTs constituted a key point for enhancement of the ?re retardancy . Up to now, different strategies have been reported to increase the dispersion degree of CNTs such as ultrasonication, pre-dispersed CNTs into polymer, chemical modification through functionalization, etc (section 2.4).

The second experimental part is related to the study about the effect of synthesized carbon nanotubes as filler on flame retardancy of Diglycidyl ether bisphenol-A (DGEBA) as a good example of thermosetting polymer. The structure of nanocomposites were characterized by SEM and TEM. Flame retardancy research were carried out through microscale combustion calorimetry (MCC), limiting oxygen index (LOI) determination and thermogramitry analysis (TGA). These results were originally presented in Paper I.
Mechanical properties of the nanocomposite based on carbon nanotubes, glass transition temperatures and elastic moduli were measured and checked by TGA, DSC and DMTA.
Results showed that the fire retardant properties of nanocomposites improved significantly specially for COx, which presented a very high LOI (35%) and a homogeneous and uniform surface after burning. This effect was attributed to the very high aspect ratio of COx tubes.

3.2 Preparation of epoxy nanocomposites
Three samples containing 2 wt.% nanotubes (CNT, CNx or COx), were prepared using a three roll mill, mixing appropriate amounts of EP and nanotubes at 120 ºC to decrease epoxy viscosity (equivalent weight of DGEBA per epoxide group is 480 g.mol?1). After milling, the mixture was gently stirred at the same temperature to partially eliminate trapped air from the mixture. Stoichiometric amount of 4,4?-Diaminodiphenyl sulfone (DDS) was then added to the mixture and the temperature was then increased to 130 ºC to dissolve the curing agent for 20 minutes under vacuum to get a clear, homogeneous and degassed mixture. The mixture was then poured into a preheated Teflon mold and cured in an oven at 140 ºC for 5 h and then post cured for 3 h at 180 ºC. Specimens for the entire tests were cut from this block. The sample without nanotubes was the reference Paper I.
Fig. 3.1. Preparation of epoxy nanocomposites in the required size for LOI test (
3.3 Structure characterization of carbon nanocomposite
The existence of a CNT network was confirmed by a detailed microscopic examination of cryo-fractured specimens. Fig. 3.2 shows some representative examples of the different microstructural features of the nanocomposites.

Fig. 3.2. FESEM images of cryo-fractured nanocomposites. a) Representative image of the surface of CNT, CNx and COx nanocomposites. b) An example of a bad interfacial coupling in CNT. c) An example of a good interfacial contact in CNx. d) Presence of an aggregate in CNx. e) Nanotube network detected inside a defect in CNx nanocomposite.

A representative image of the dispersion degree is presented in Fig. 3.2. (a) where it can be observed an apparently homogeneous distribution of the nanofiller on the surface of CNT nanocomposites. The surfaces of CNx and COx appear to be similar. Some aggregates, as depicted in Fig. 3.2. (b) and (d), were observed for CNT and CNx nanocomposites while no aggregates could be observed for COx. Concerning the interfacial contact, we have observed a wide range of cases within the same samples. In some cases a complete decoupling of the nanotubes from the matrix was observed (Fig. 3.2. (b)) but inspection of other portions of the same sample revealed a good interaction of the tubes (Fig. 3.2. (c)). Therefore, we can conclude that there are no specific features of the microstructure of the nanocomposites that could be attributable to the effect of doping except for the case of oxygen doping. However, the most interesting finding appears in Fig. 3.2. (e), were inspection inside a defect free of polymer matrix revealed the presence of a network of interconnected nanotubes, confirming thus the existence of a percolative network. (Paper I)
3.4. Thermal stability and mechanical behavior of epoxy nanocomposites
3.4.1 Thermogravimetry analysis (TGA)
The influence of the pristine and doped CNTs on the thermal stability of epoxy nanocomposites was investigated by TGA, as shown in Fig. 3.3 and Table 3.1.
Fig. 3.3. TGA curve of epoxy and its nanocomposites in N2
Table 3.1. The TGA data of EP and its nanocomposites
Pure epoxy shows a sharp mass loss in the temperature range of 400 – 450 ºC, and its residual mass percentage is 13.3% under N2 atmosphere. The introduction of either pristine or doped CNT caused the earlier initial decomposition of the nanocomposites in about 16ºC as reflected by T5% and Tmax data. This effect may be attributed to the high thermal conductivity of the nanofillers. However, as degradation starts under N2 atmosphere, a char layer begins to be formed; this layer has a protective effect and it should be therefore expected a delay or reduction of gaseous products. This is just what was observed since the incorporation of the nanotubes led to an increase in the char yield at 700 ºC in about 6–10% (Table 2.3) following the sequence COx/EP>CNT/EP>CNx/EP. Notably, these experimental char residues were higher than the calculated ones assuming additivity of char residue (see Table 2.3 for the weight loss under N2 at 700 ºC), suggesting the existence of synergism among the modifiers. Furthermore, it is interesting to note that the thermal oxidative resistance of the nanocomposites close correlates with the length to diameter ratio of the nanotubes (Table 2.2). It seems that the higher length to diameter ratio, the higher char yield is produced, and this suggests to be an effect of the percolative network formed by the nanotubes as already suggested by other authors .

However, analysis of char residue is concerned with the end of the degradation process and some effect of the percolative network should be expected at any other stage of the process. We have carefully examined the TGA thermograms of both the pure nanotubes and the nanocomposites, and we have found a significant difference in the maximum rate at which weight is lost, (dm/dt)max which can be considered as a measure of the degradation rate. Among the three tubes examined in this work, COx presents the maximum degradation rate (Table 2.3) in air but when it is incorporated into the epoxy, the resulting nanocomposite presents the lowest degradation rate (Table 3.1). The three nanotubes lower the degradation rate of pure epoxy but COx almost halves the value. Since COx alone degrades quickly in air, this effect may be also associated to its high aspect ratio.

3.4.2 DSC and DMTA
DSC was used to determine the glass transition temperature (Tg) of the nanocomposites using a Mettler Toledo DSC 822 with a liquid nitrogen reservoir. Samples of about 5 mg were scanned from 25 ºC to 250 ºC at 10 ºC.min-1. In order to minimize the effects of previous thermal history, data from a second scan were used for analysis.

Dynamic mechanical thermal properties were measured using a DMTA Q800 Dynamic Mechanical Analyzer (TA Instruments), with amplitude of 30 µm at 1 Hz. Nanocomposite specimens with nominal dimensions of 60×10×2 mm3 were mechanically tested in single cantilever mode. The samples were heated from room temperature to 250 ºC at a linear rate of 3 ºC.min-1.

The influence of addition of nanofillers on the relaxational and mechanical behavior of nanocomposites has been evaluated also by DSC and DMTA. Glass transition temperatures of all the composites were measured by DSC, and DMTA (maximum in tan, Fig. 3.4) and the results are summarized in Table 3.2.

Reported values of the Tg for epoxy resins cured with DDS are in the range 184 ºC to 230 ºC . Differences with our case are attributed to the high molecular weight of the prepolymer used in this work (480 g/mol instead of the usual 340 g/mol used by other authors).
Storage modulus (Fig. 5) shows an apparent drop at about 60ºC that is attributed to the tail of the relaxation that typically appears at -59ºC, as reported by other authors on a similar system 87. R.t. values are presented in Table 3. No significant variations were found in the Tg on adding 2% nanotubes. Neither in the storage modulus nor in the width of the loss tangent peak meaningful alterations were detected as well. Bad CNT-Epoxy interfacial interactions usually manifest as a decrease in both, the storage modulus and Tg and it is usually justified as an increase in free volume. Our results indicate that addition of nanofillers do not negatively affect the relaxational and mechanical behavior of the nanocomposites.
Fig. 3.4. Storage modulus and tan ? versus temperature plots of epoxy and its composites.

Table 3.2. DSC and DMTA results for nanocomposites.

* Full width at half maximum
3.5 Flammability
The flammability behavior of nanocomposites was investigated by LOI and MCC.

3.5.1 Limiting Oxygen Index (LOI)
The limiting oxygen index (LOI) is defined as the minimum oxygen concentration (in vol%) that is necessary to sustain a stable combustion of the specimen after ignition. “The higher the LOI of a polymer material the lower the flammability”. Air consists of approximately 21% oxygen. Therefore, any material with an LOI ?21 will probably support burning in an open-air situation.

The LOI value is a basic property of the plastic but tells us nothing about how the plastic will react to burning in an open atmosphere. Therefore, this test is simply used to compare the relative flammability and rank polymer and composite materials due to the simple and repeatable condition .

LOI tests are performed under standard conditions (as specified by Fire Testing Technology, UK) according to ASTM D 2863-97 with specimens of dimension 130 × 6.5 × 3.2 mm3 .
A rod sample with special size is placed vertically at the center of a glass chimney and a slow stream of O2/N2 mix is fed from the bottom of the column. The O2/N2 ratio is changeable (Fig. 3.5 (a) and (b)).

The test stars by ignition the top edge of the test sample with a candle-like flame. The O2/N2 ratio of the flow is decreased until the flame is no longer supported. 
LOI percent is calculated from the minimum concentration of O2 that will just support combustion. The experimental error in LOI estimation was ± 0.2%.

Fig. 3.5. (a) LOI test equipment; (b) Checking the flammability of a sample
The three different types of CNTs (CNT, COx and CNx) have been incorporated into the epoxy matrix and their flammability was examined by measuring their limiting oxygen index (LOI); results are summarized in Fig. 3.6.
Pure epoxy exhibited a LOI value of 21.5%, indicating an easily ignited material. With the addition of 2 wt.% CNT, the LOI of the resultant epoxy composite increases to 33.5%. LOI values above 27 are normally indicative of materials that are self-extinguishing . The incorporation of COx into epoxy results in a highest LOI value of 35%. In contrast, the addition of CNx does not lead to a significant increment in LOI as COx, but still exhibiting a high LOI value of 31.5%. In terms of LOI increment: 12% (CNT), 13.5% (COx) and 10% (CNx) Paper I
Fig. 3.6. LOI data of pure EP and its nanocomposites
Obviously, LOI values of the composites strongly depend on the LOI value of the polymer matrix and on the presence of other additives. Therefore, a more indicative parameter of the effect of carbon nanofillers may be the LOI increment with respect to the same material without carbon nanofillers. A literature search on the maximum LOI increments found for epoxy thermosets modified with carbon nanotubes yields the following data: 0.2% for epoxy loaded with double walled CNT , 3% for epoxy with CNT synthesized by the sol-gel method , 5% for epoxy with amino functionalized CNT . Therefore, it seems that the LOI increments found in this paper represent the maximum values ever reported for epoxy thermosets.
Complimentary data to TGA and LOI experiments are supplied by micro-scale combustion calorimetry (MCC) experiments, a testing technique that provides a variety of information on the combustion behavior of polymeric materials .

3.5.2 Microscale combustion calorimetry (MCC)
Microscale combustion calorimetry (MCC) (Fire Testing Technology, UK) was used to investigate the combustion of nanocomposites according to ASTM D7309. About 5 mg of sample were heated up to 700 ºC at 1 ºC.s-1 under nitrogen at 80 cm3.min-1. The volatile and anaerobic thermal degradation products in the nitrogen stream were mixed with a 20 cm3.min-1 N2/O2 stream (20% O2) prior entering the combustion furnace at 800 ºC. Each sample was run in three replicates.

The relevant parameters obtained from the MCC include peak heat release rate (pHRR), total heat released (THR), temperature at pHRR (Tmax) and char yield. Results are illustrated in Fig. 3.7 and Table 3.3.
Fig. 3.7. The HRR curves of samples from MCC.

Table 3.3. Peak heat release rate (pHRR), total heat released (THR), temperature at pHRR and char residue as obtained from MCC measurements.

Table 3.3. shows that pHRR and THR of neat epoxy (EP) are reduced by 22.7% and 25.3% respectively, by the addition of 2 wt.% of CNx. A further minor reduction in pHRR and THR was achieved by the incorporation of CNTs in comparison with the sample containing CNx. However, the highest reduction in pHRR (38.7%) and THR (32.8%) was achieved by incorporation of 2 wt.% of COx into the epoxy matrix. These results are in accordance with previously reported experiments on polypropylene/MWCNT at low loadings 85. It can be also noted from Table 4 that the char residue increased on addition of nanotubes in accordance with TGA results presented previously (Table 3.1), except for the CNx case. Interestingly, char residue for COx composites is higher than the char residue of pure epoxy plus amount of nanotubes, indicating that the tubes may induce some additional charring (5 – 7%) of the epoxy matrix.

3.6 Morphology of the char residue
The surface morphology of the samples after LOI test was checked by SEM and presented in Fig. 3.8. EP, CNT/EP and CNx/EP (Fig. 3.8a, b and c) nanocomposites showed a fluffy and cracked char layer full of open holes all along the surface; however COx/EP displayed a thick and homogeneous layer which may explain the significant reduction of heat and mass transfer associated to the enhanced flame retardancy presented by the nanocomposites made with this filler.

Fig. 3.8. SEM images of the surface of the residual chars of EP (a), CNTs/EP (b), CNx/EP (c), and COx/EP (d) after the LOI test
3.7 Discussion
Based on the LOI and MCC results, the COx/EP system is illustrative of a condensed phase burning mechanism with a significantly lower mass loss and higher char yield compared to the other samples.
The higher char residue quenches the fire. Therefore, less epoxy converted into combustible fuel. The early appearance of a network structure on the external surface of the sample (Scheme 3.1b) right after the ignition, followed by formation of a thick char layer, served as a thermal barrier to separate burning materials from oxygen also preventing feeding the flame zone from combustible gases, as shown in (Scheme 3.1c).
Scheme. 3.1. A proposed flame retardant mechanism
3.8 Conclusion
Epoxy nanocomposites containing the three types of carbon nanotubes were prepared by three roll mill at a constant loading of 2% in an epoxy/DDS thermoset and curing them in two stages to ensure the maximum conversion degree. Morphological observations revealed that the CNTs were homogeneously dispersed in the epoxy matrix. Some aggregates and some regions in which nanotubes decoupled from the matrix were observed for CNT and CNx systems. However, no such defects were observed for COx nanocomposites (Fig. 3.2).

The influence of pristine and doped CNTs on the thermal stability of epoxy nanocomposites was investigated by TGA. The results showed that introduction of the nanofillers at 2 wt.% loading caused the earlier initial decomposition of the nanocomposites, an increased char yield, and a reduction of the degradation rate of in comparison with pure epoxy.

The influence of addition of nanofillers on the relaxational and mechanical behavior of nanocomposites was evaluated by DSC and DMTA. No significant variations in Tg, storage modulus or loss tangent width were detected, indicating that the addition of nanofillers does not negatively affect the relaxational and mechanical response of the nanocomposites.

The flammability behavior of nanocomposites was investigated by LOI and MCC. The LOI increments for CNx, CNT and COx nanocomposites were observed 31.5, 33.5 and 35 respectively, being the latter the highest increment observed for epoxy doped with carbon nanotubes.

The MCC results showed a remarkable reduction in pHRR and THR of 38.7% and 32.8% respectively for COx system.
From the SEM images taken from the residual char layer formed by COx nanocomposite after burning, a uniform and continuous char layer was observed. However, the other studied systems presented a fluffy surface with plenty of open holes.
In the absence of specific effects that could be associated to the surface chemical nature of the carbon nanotubes, our results suggest that the effective improvement in the flame retardant properties of the studied nanocomposites is due to a homogeneous dispersion of CNTs in the epoxy matrix. In the case of oxygen doped carbon nanotubes, which presents notably enhanced flame retardancy, a more effective barrier was formed and this finding must be associated to the formation of a denser percolative network due to its very large aspect ratio. Hence, the key role of percolative networks suggests the need of exploring the flame retardancy properties of systems with the varying aspect ratio which is the aim of our ongoing research. Chapter 4
Microwave-assisted Decoration of Graphene Oxide (GO) by Copper hydroxy nitrate Cu2(OH)3NO3, Cuprous oxide (Cu2O) and Copper oxide (CuO)
4.1 Synthesis of graphene
After discovery of graphene in 2004, different methods were developed to generate thin graphitic films and few layer graphene, depending on the desired size, purity and crystallinity of the final product . A summary of common synthesis methods is shown in Scheme 4.1. There are two main approaches; “Top- down” and “Bottom- up”. The first one involves the production of graphene or modified graphene sheets by separation/ exfoliation of bulk graphite or graphite derivatives (such as graphite oxide (GO) and graphite fluoride), whereas in “bottom- up” methods graphene is grown on different substrates, which are subsequently removed latterly . This means starting at atomic scale and building up atom by atom to the desired final size of the material.

Common graphene synthesis approaches includes mechanical exfoliation from graphite, chemical vapor deposition, and reduction of graphene oxide.

One graphitic layer is well known as or single-layer or monoatomic graphene. A two or three graphitic layers are known as bilayer and tri-layer graphene, respectively. More than 5 layer up to 10 layer graphene is generally called few layer graphene, and 20–30 layer graphene is referred to as multilayer graphene, thick graphene, or nanocrystalline thin graphite .

Scheme 4.1 Graphene synthesis methods
In the following, some more details of common graphene synthesis techniques will be brought.
4.1.1 Mechanical cleavage/exfoliation
Graphite is formed of mono-atomic graphene layers bundling through overlapping of partially filled p orbital perpendicular to the plane of the sheet together by weak van der Waals forces. The inter layer band energy and distance between layers is 2 eV/nm2 and 3.34 A°, respectively.

Exfoliation with the large lattice spacing via weak bonding in the perpendicular direction is the reverse of stacking with the small lattice spacing and stronger bonding in the hexagonal plane . An external force about 300 nN/µm2 is required to separate one mono-atomic layer from graphite via mechanical cleavage . The first reported production of graphene was by Novoselov and Gaim by mechanical exfoliation using a simple “scotch tape” method in 2004 .

This method is a top-down technique in nanotechnology, by which a perpendicular stress is created on the surface of the layered structure materials, single-layer graphene (SLG) or few-layer graphene (FLG) are peeled off Highly Oriented Pyrolytic Graphite (HOPG) via using a variety of agents like adhesive tape , electric field , ultrasonication or by transfer printing technique tape .

Although, the quality of the prepared graphene by this method is high with almost no defects, this process has major disadvantage to obtain larger amounts of graphene .
4.1.2 Chemical exfoliation
Directly graphite exfoliation in the liquid phase as another method to produce graphene have been already reported by some research groups . In these methods, choosing the most suitable solvent or surfactants to exfoliate and disperse graphene is very important. The low yield efficiency (typically ?0.01 mg mL-1) and longtime sonication (400h) are the most disadvantages of these methods .
Indirectly graphite exfoliation with preliminary oxidation is another chemical method. This strategy is a two-step procedure. First step includes reduction the interlayer van der Waals forces by phenyl hydrazine, hydroxylamine, glucose, ascorbic acid, hydroquinone, pyrrole, oxidation of layers, or inserting large alkaline ions between the graphite layers . In this step, graphene intercalated compounds is formed (GICs). The second step is the exfoliation of GICs to obtain few layers by reduction, rapid heating, sonication or ultracentrifugation , .

The most common ways for oxidation of graphite are Brodie and Hummers’ methods. Hummers’ oxidation method is still the favored method because of the following reasons :
i) Replacement of KClO3 by KMnO4 as the oxidation agent, which resulted in elimination of toxic gas byproducts and improving the reaction safety
ii) The shorter oxidation time is needed in this method
iii) It is easy to disperse the resulted GO in water
iv) A larger restoration of the pristine graphite 2D structure is achieved .

Different reduction methods such as thermal, photo catalyst, solvothermal, hydrothermal, microwave , sonochemical , and photo reduction have been reported in order to provide reduced graphene oxide (rGO) .
Elemental analysis (atomic C/O ratio) of the rGO revealed the existence of a significant amount of oxygen, indicating that reduced graphene oxide is not the same as pristine graphene.
Theoretical calculations of the rGO (the model used for graphene oxide had the graphene decorated with hydroxyl and epoxide groups) suggest that reduction below 6.25% of the area of the graphene oxide (C/O = 16 in atomic ratio) may be difficult in terms of removing the remaining hydroxyl groups . GO contains about 40% oxygen, whereas the oxygen content in rGO is typically lower than 10%. Oxygenated functionalities include carboxylic acid groups, ketones, ethers, and hydroxyls
Actually, the carbon to oxygen ratio 4:1 considered above is a bit larger than experimental values for strongly rGO and almost twice larger than the maximal
Normally maximum reduction can not be achieved and some parts of oxidized group
Hummers’ method used to produce GO
4.1.3 Pyrolysis methods
Graphene sheets could be smoothly detached by pyrolization of different hydrocarbon sources. This produced graphene sheets with dimensions of up to 10 µm. Although the quality of produced graphene in this method because of a large number of defects is not perfect, the most important benefit of this method is fabricating graphene in a low temperature with low-cost method . The yield and morphology of the final graphene are affected by pyrolytic conditions such as temperature of the reaction, composition of the hydrocarbon source and the size of catalyst .

4.1.4 Epitaxial growth
The graphitization of SiC was discovered in 1955, but it was not considered as a method of preparing graphene in the beginning . Nowadays, epitaxial growth directly from a carbide surface is precisely controlled locations approach to synthesize of graphene. Graphene can be prepared by simply heating and cooling down on silicon carbide (SiC) crystal . The results are highly dependent on temperature, pressure and heating rate. For instance, the growth of CNTs instead of graphene in a very high temperatures and pressures possibly occurs .
Epitaxial growth of graphene on another attractive alternative surface like Ru , Cu have been also reported.
4.1.5 Chemical vapor deposition (CVD)
In chemical vapor deposition method, graphene is grown directly on a metal or thin film substrate via saturation of carbon upon exposure to a hydrocarbon gas. CVD growth of graphene has been mainly practiced on copper and nickel substrates. Deposition of graphene on the other transition metals like Pd and Ir have been also reported. This method contains chemical reaction of heating and changing molecules to a gaseous state, which is called precursor. Methane is a common hydrocarbon precursor gas at a high temperature. By cooling the substrate and decreasing the solubility of carbon on the substrate, the carbon precipitates and form mono- to multilayer graphene sheets on the substrate.
There are different types of CVD processes that are Depending on the desired structure and quality of final graphene, different forms of CVD processes are available: thermal, plasma enhanced (PECVD) cold wall, hot wall, reactive, and so on .

4.1.6 Other methods
Some other techniques are also reported such as longitudinal unzipping of CNTs , microwave assisted exfoliation of CNTs , arc- discharge of graphite , electron beam irradiation of PMMA nanofibers , thermal conversion of nano diamond .

4.2 Graphene-based nanocomposites
Graphene with significant conductivity, excellent electron mobility and fascinating high specific surface area is a good substrate to produce graphene-based composites via functionalization. Graphene with remarkable properties is a “magic bullet” for the composite world .

In recent years, decoration of graphene sheets with metal oxides nanoparticles (MONPs) to prepare graphene metal oxide composites has attracted considerable attention for their potential applications in photovoltaic devices, energy harvesting, catalysis and photocatalysis areas. The synergetic effect between graphene and MONPs resulted in outstanding properties of graphene composites .
Anchoring transitional metal oxides such as TiO2, ZnO, SnO2, MnO2, Co3O4, Fe3O4, Fe2O3, NiO, Cu2O on graphene are good examples to obtain enhanced efficiency in various applications of these composites. To date, different strategies to synthesize and support MONPs have been proposed. The most effective methods to MONPs decoration on graphene includes solution mixing method, sol–gel method, hydrothermal/solvothermal method, self-assembly and microwave irradiation .
4.3 Graphene-metal oxide composites in photocatalytic degradation
Actually, activated carbons with high surface area have been widely used for water purification . Graphene which theoretically shows nearly twice the surface area of activated carbon can supply as a very good alternative for water purification .

Metal oxides have also been employed as a photocatalyst for the degradation of various water pollutants and have shown promising results . Metal oxides are not still competitive with the traditional water purification techniques such as reverse osmosis and filtration due to some drawbacks. Firstly, metal oxides are only active in UV region because of their wide bandgap. Secondly, due to the possibility of charge carrier recombination, the lifetime of the active species responsible for the degradation is short.

Combination of MO oxides with electron scavenging agents has been proposed as one the most common strategies to improve the photocatalytic efficiency of the metal oxides.

The most advantages of utilizing GO as support material with electron scavenging properties to improve the photocatalytic performance are summarized as below:
(i) The functional groups of GO (carbonyl, epoxy and hydroxyl groups) act as the nucleation and increase growth sites for nanoparticles
(ii) Improving the surface area of the composite because of the decrease aggregation of GO layers
(iii) Decrease agglomeration of MONPs by positive effect of graphene sheets on enhancing nanoparticle dispersion
(iv) Inhibition of MONPs’ leaching into the treated solution and extend the lifetime of the adsorbent material
(v) Reduce the bandgap energy of metal oxide by GO via energy-favored hybridization of 2p atomic orbitals of oxygen and carbon
(vi) Limitation of the electron hole recombination by efficiently separation of the photo-generated charge carriers due to the GO role (see section 4. )
4.4 Graphene copper based nanocomposites
To date, among the aforementioned transitional metal oxides, low cost copper based metal oxides are materials with potential and diversified heterogeneous catalytic applications including electrochemical catalysis , photocatalysis , NOTEREF _Ref519780446 h * MERGEFORMAT 139 , chemical catalysis in organic synthesis . Therefore, due to their excellent properties, preparation of graphene-copper oxide nanohybrides have attracted our interest.

In our work, Hummers’ method used to produce GO because of the mentioned reasons (section 3.1.2)
Our focus in the present thesis was to design a convenient and cost effective method using domestic microwave oven to prepare and decorate copper salts on GO.

A unique characteristic of microwave processing is rapid, sectional and localized heating, because materials are heated directly through the interaction with microwave energy which is opposite to the slow heating in a conventional furnace .
In the next step of our work, we have checked the performance of the prepared copper salts/GO nanocomposites in different catalytic areas and compare to the previous reports in literature. As we expected, incorporation of copper oxides (copper with different oxidation number) onto GO via our method resulted in enhancing the properties of initial metal oxides and it was comparable with the copper salts/GO hybrids produced of conventional methods.
4.5 Experimental procedures
4.5.1 Materials
Graphite powder (with a purity ;99.999%) was purchased from Alfa Aesar., H2O2 30% W/V (Panreac), KMnO4 (Panreac), NaNO3 (Sigma-Aldrich) and H2SO4 98% V/V (Panreac) were employed for graphite oxidation and used without any further purification. Copper (II) nitrate hemi (pentahydrate) Cu(NO3)2. 2.5 H2O was purchased from Sigma-Aldrich was utilized to decoration on GO.

4.5.2 Preparation of graphene oxide (GO)
Graphene Oxide was prepared from natural flake graphite powders using a modified Hummers’ method NOTEREF _Ref519074339 h 11. Briefly,
1. Concentrated H2SO4 (180 mL) was added to a mixture of graphite (4 g) and NaNO3 (2 g)
2. The mixture was cooled down to 0 °C in an ice bath
3. KMnO4 (11 g) was added slowly in small doses to keep the reaction temperature below 20 °C
3. The solution was heated to 35 °C and stirred for 2h
4. 300 mL water was added to complete the oxidation reaction
5. 30 mL H2O2 30% was slowly added for elimination of exceed KMnO4
6. The reaction mixture was stirred for 30 minutes
7. The mixture was centrifuged (3700 rpm for 30 min), after which the supernatant was decanted away
8. The remaining solid material was washed with water and centrifuged again, this process being repeated until the pH was between 6 and 7
9. The obtained brown-yellow graphene oxide was dispersed in water (quantified at 5 mg mL-1) and poured into the plastic tube
10. The tubes were frozen by liquid nitrogen and transferred into an ultra-low freeze dryer (-50°C, 15 Pa) for at least 72 h until flocculated powder was obtained
11. Using furnace at 950°C for ?5min leads us to partially reduced graphene oxides
Prepared GO was kept in closed for next experimental tests. ?? ???? ?????? ????
4.5.3 Decoration of copper oxides on graphene oxide
The schematic illustration of the Copper salts/GO composites shows in Scheme 1.
A mixture of GO 100 mg in Ethanol (dispersed under 5 min sonication) was added to 100 mL of Cu(NO3)2 0.01 M, then the solution was microwaved for 2 minutes (on-off system). The mixture was filtered after cooling, and the obtained precipitate was washed with deionized water and hot ethanol, at least for five times and dried in a vacuum oven at 90 ºC for 12h.
Decorated GO collected for characterization tests.
Scheme 4.2 Microwave assisted procedure
4.5.3.1 Solvent effect
We found that solvent has a key role in our method. According to the XRD pattern of decorated GO (see section 4.5.4) by changing the solvent to ethylene glycol in the same condition (2 min microwave), different copper salt was decorated.
Furthermore,
In this method, microwave can prepare copper salts nanoparticle and decorate them on GO simultaneously that can be mentioned as one pot synthesis. Therefore, all experiments have been done in the same condition without GO to compare pristine and decorated copper salts nanoparticles.
All collected products were well characterized with XRD, SEM, TEM, TGA, .

Characterization methods
4.5.4 XRD
Powder X-ray diffraction (XRD) is one of the primary tools of chemical analysis used in this work. It is a rapid analytical technique used for phase identification of crystalline or amorphous materials. XRD was collected in an automatic X’Pert Philips diffractometer using a Cu K? radiation (? = 1.54187A°). Data were collected in the 2? range from 5° to 80° in step-scanning mode with a step size of 0.01° and a counting time of 0.5 s per step.

4.5.4.1 XRD of GO
GO exhibited the strong diffraction peak at 2?= 10.7º (Fig. 1), attributed to the (001) plane due to oxygen containing functional groups on carbon sheets (Fig 4.2).
Fig. 4.2 X-ray diffraction patterns of graphene oxide
The layer distance or interlayer spacing between graphene oxides (d) can be calculated by Bragg’s law that:
n ? =2d sin (?)
where ? is the wavelength of the X-ray beam (0.154 nm), n is 0.9, ? is the diffraction angle. The distance between graphene oxide layers was about 0.7 nm.
4.5.4.1 XRD of Cu/GO composites
The X-ray diffraction patterns of decorated GO with different solvent shows well-defined and sharp peaks.

Microwave assisted copper decoration of GO in ethanol resulted in a special type of copper salts. According to the XRD data-base, copper hydroxy nitrate (Cu2(OH)3NO3) has been produced and decorated on GO (Fig. 4.3) .
Copper hydroxy nitrate (here is called as DS) is a double salts belonging to layered materials family with special application as an ion exchanger, will be explained more later (see section 5.1).

Changing the solvent to ethylene glycol in the same condition can decorate cuprous oxide nanoparticles (Cu2O) on GO with XRD pattern which is in a full agreement with the literature . (Fig. 4.4)
The characteristic peak for GO in XRD is not observed clearly due to the high concentration of Cu2O and DS nanoparticles on the substrate.
Fig. 4.3 X-ray diffraction patterns of DS and DS/GO nanocomposite
Fig. 4.4 X-ray diffraction patterns of Cu2O and Cu2O/GO nanocomposite
The average size of Cu particles can be estimated by the Scherrer equation:
d=0.9?/?1/2cos?
d is the average particle size (nm), ? is the wavelength of the X-ray used (0.15406 nm), ?1/2 is the width of the diffraction peak at half height in radians and ? is the angle at the position of the peak maximum. The calculated average particle size of DS and Cu2O are 47 and 9 nm respectively and for DS and Cu2O supported on graphene are 27 and 9 nm, respectively.
In addition, we found that the salts decorated and non-decorated could be stored in the air for at least six months without any changing in their XRD pattern.
4.5.6 Thermogravimetry analysis (TGA)
Approximately 5 mg of each sample used for TGA in a Perkin-Elmer 6000 STA system heating from room temperature to 800 ºC.
The thermal stability and the composition of GO and the Copper salts/GO nanocomposites were further investigated by using TGA (Fig. 4.6). Graphene oxide showed several degradation steps. Initially, ~10% weight loss at about 100 ?, may be assigned to the loss of physisorbed water. A major loss of weight of GO consists of two steps, first sudden weight loss at ~140 ? followed by another steep step in the temperature range from 160?–290?, which were assigned to the pyrolysis of oxygen-containing functional groups (carboxylate, anhydride, or lactone groups). Upon deposition of Copper nanoparticles onto GO nanosheets, the weight loss in this region decreased to less than 36% due to the reduction of these oxygen-containing functional groups. Between 300°C–500°C was associated with the removal of more stable oxygen groups such as phenol, carbonyl, and quinine . The weight loss observed for the composite above 500 °C is due to the pyrolysis of the carbon skeleton of GO nanosheets. The total residual weight of GO obtained at 800 ? was around 24%.
TGA is an effective analytical technique to evaluate the ratio of copper salts/GO. The content of DS and Cu2O in the nanocomposite calculated by amount of residue at 800 ºC are about 53.5 and 79.5 wt.%, respectively calculated by TGA, which is fairly close to the theoretical values, 46 and 69 wt.%, based on the experimental conditions.
The TGA of Cu2O/GO shows three weight loss steps.
Fig. 4.6. TGA traces of GO, DS/GO and Cu2O/GO nanocomposites
DS/GO in TGA exhibited no weight loss up to 200 ?; however, in the temperature range of ~200?–270 ?, showed a weight loss of ~30% due to the decomposition of DS in one step to produce CuO according to the chemical equation :
4 Cu2(OH)3NO3 ? 8 CuO(s) + 4 NO2(g)?+ O2(g)? + 6 H2O(g)
According to the above equation we could obtain black pristine copper oxide (CuO) and decorated on GO by heating DS and DS/GO respectively at about 250 for 10 min in an oven.
No significant mass loss was detected when the copper oxide and CuO/GO nanocomposite were heated up to 800?.

Fig 4.7. TGA traces of DS, CuO and CuO/GO composite
The content of CuO in the nanocomposite calculated from amount of residue at 800 ºC is 90.7 wt.%.
We have also checked the XRD of CuO and CuO/GO in the same method as the other salts. Fig. 4.8 shows the XRD pattern of free copper oxide and decorated on graphene which is in accordance with the literature .

Fig. 4.8 X-ray diffraction patterns of CuO and CuO/GO nanocomposite
The size of CuO and CuO/GO have been estimated by using Debye-Scherer equation 21 and 18 nm respectively.
4.5.5 Transition Electron Microscopy (TEM)
The morphology of synthesized samples was characterized using TEM measurements. As it can be seen in Fig. 4.8, nearly three kinds of morphology in copper products are obtained with copper nitrate as precursor: rods DS, spherical Cu2O and cubic CuO.
The characteristics of copper nanoparticles deposited on the surface of graphene oxide could be clearly seen in Fig. 4.5. The particles were well anchored on GO nanosheets. The GO in nanocomposites exhibited plentiful wrinkles, which resulted from the graphitization of GO sheets in exfoliation process. Apparently, GO nanosheets could effectively prevent the aggregation of copper particles.

Fig. 4.8. TEM images of DS, Cu2O and CuO alone and decorated on graphene oxide
Raman spectroscopy measurements were carried out in via Renishaw confocal microscope based Raman spectrometer using the 514.5 nm laser excitation. For each sample, various spectra were recorded in different places in order to verify the homogeneity of the sample.Chapter 5
Comparison Study of the synthesized Copper/GO nanohybrides in catalytic applications
5.1 Introduction
One of the most important 3d transitional metal with high natural abundancy and low cost is copper. Copper nanoparticles (Cu Np) are especially attractive in catalytic fields because of their interesting physical and chemical properties. The application of Cu NP is limited by instability under atmospheric conditions. Hence, Cu NP located on a generally inert support could be a suitable alternative to increase their stability ,. Graphite and its derivatives like graphene, graphene oxide, and reduced graphene oxide have been recognized as attractive catalyst supports because of their extremely high surface area (?2600 m2/g), high thermal/electrical conductivity, and chemical stability .
Up to now, a wide range of accessible oxidation states of copper (Cu0, CuI in Cu2O and CuII in CuO) decorated on graphene oxide via different methods, have been explored as a suitable heterogeneous catalyst in various chemical transformations NOTEREF _Ref521413842 h * MERGEFORMAT 155. Copper oxide (CuO), cuprous oxide (Cu2O) and copper hydroxy nitrate (DS) have been synthesized and decorated on GO as we explained in chapter 4.
Furthermore, since synthesize and decoration of DS on GO via the new method is significant achievement in our work, we will explain more about the properties, conventional synthesize methods and application of DS in details.
5.1.1 Copper hydroxy nitrate (DS)
In recent years, several articles involving transition metal double salts with general composition M2(OH)3X (M=Cu, Co; X= inorganic anion, organic anion, …) consists of a two-dimensional nanostructure of positively charged metal hydroxide and interlayered anions with tunable physical and chemical properties have been published .

Copper hydroxide (or hydroxy/ hydroxyl) nitrate, Cu2(OH)3NO3 (we show it by DS), one of the most important member of this group exists in two structurally dimorph varieties: a conventional synthetic metastable monoclinic phase and a natural orthorhombic phase occurring in the mineral Gerhardtite. There has been considerable research on the exact structure of these two dimorphs and the relationship between the two phases .
The structure of DS can be imagined as being obtained by substitution of 25% of the hydroxide anions (OH-) in the octahedral copper hydroxide sheets by NO3- anions directly coordinated to copper(II) cations through one of its oxygen atoms, as shown in Fig. 5.1. The nitrate ions can be displaced by others negative charged molecules.

Fig 5. 1 Schematic representation of copper hydroxy nitrate (DS)
Over the past few decades, different methods have been proposed to synthesize this double salt. The most commonly used method is the use of directly mixing aqueous solutions of copper nitrate and sodium hydroxide . Ammonium bicarbonate (NH4HCO3) and sodium carbonate (Na2CO3) were used by Wolf et al. and Bushong et al. , respectively, as a base reservoirs to synthesized Cu2(OH)3NO3. These methods demand strict control of experimental variables, especially pH, temperature and concentration of the reactants to avoid the formation of sub-products like Cu(OH)2, CuO and other stoichiometric ratios with available ions (Cu2(OH)3(NO3), Cu3(OH)(NO3)5, …).

Urea hydrolyzes as an OH? generator instead of NaOH for the formation of copper hydroxy nitrate was utilized by Henrist et al. and Cho et al. . Anandan et al. improved this method by using high-intensity ultrasonic irradiation. The inevitable contamination of Cu2(OH)3NO3 by CO3? ions in urea hydrolysis is the main drawback of this method .

Wang et al. has reported another method to obtain Cu2(OH)3NO3 by adding copper powder into the copper nitrate solution and high-intensity ultra-sonication for several hours. Copper hydroxy nitrate can be prepared through the reaction of magnesium hydroxide (Mg(OH)2) and copper nitrate aqueous solution .
Cu2(OH)3NO3 was also obtained by dispersion of copper oxide (CuO) in copper nitrate solutions in a very long reaction time (e.g. 7 weeks) .
Heating aqueous copper nitrate up to 200°C at about 3 days , and interaction of solution of copper nitrate and propanol in the solvothermal condition were alternative method to synthesize copper hydroxy nitrate which are time and energy consuming and require severe control of the process.

Due to the drawbacks of previous methods, we proposed microwave oven as an alternative convenient and cost effective way to synthesize and decoration of DS on GO (see section 4.5.3). As we checked the XRD of the produced DS, it was found a good accordance with DS obtained from the conventional method.

DS has shown applications in vehicle airbags , ion exchangers , radiochemical applications , oxidation degradation for purification of wastewater , fire retardant of polymers , and recently as an effective protective material against organophosphate nerve agents . The copper hydroxy nitrate was used as a raw material for other products. For example, it can be a good precursor for preparing Cu(OH)2 and CuO , Some other different copper hydroxy double salts can be synthesized easily by anion exchange of nitrate anions . According to the literature, it has not found any reports about applying DS as a catalyst in chemical reactions.
In this chapter, we will compare the performance of the three copper salts/GO nanocomposites (DS, CuO and Cu2O/GO) prepared by the microwave assisted method as a catalysts in photocatalysis, electrocatalysis, and catalytic organic transformations.
5.2 Photocatlytic application
Dyes in wastewater as one of the most important groups of pollutants are usually recalcitrant nature and non-bio-degradable. Therefore, a large amount of different catalysts such as photocatalysts to accelerate the dye pollutant degradation has been recommended so far today .

In our work, the catalytic application of the prepared samples was estimated from the decomposition of Rhodamine B (RhB) with chemical structure (Fig. 5.2) in an aqueous solution under ambient dark conditions.
RhB was chosen as one of the famous dyes which is commonly used as a colorant in textiles industries due to its high stability. According to its toxic properties for human beings and animals, different removal methods in recent studies in the literature have been introduced. Using metal oxide nanoparticles (MONPs) especially semiconductor ones (e.g. TiO2, FeO, ZnO, Cu2O, etc.) decorated on GO as heterogeneous photocatalysis is an effective and rapid strategy to remove dye pollutants .

Furthermore, we have selected RhB as model compound because of its high stability in native light and its relatively nondegradability in the absence of any photocatalyst .

Fig. 5.2 Rhodamine B’s chemical structure
It is well known that Cu2O can be used as photocatalyst to remove organic pollutants via photodegradation .
5.2.1 Calibration curve of Rhodamine B concentration
Photocatalytic behavior of samples was recorded on UV–vis spectrophotometer (Jasco V-650) over the range of 450-600 nm because the strongest photon absorption of RhB can be detected in this range perfectly (?max (RhB)= 554 nm)
An initial calibration of the UV-vis spectrophotometer was performed by measuring known concentrations of RhB. A calibration line (Fig. 5. 3 (a)) was obtained from the absorbance analysis of the aqueous standard solutions of RhB in different concentration (2, 4, 6, 8, and 10 ppm) (Fig. 5. 3 (b)). The straight line shows that the absorbance (A) of a solution is directly proportional to the concentration (C) of the absorbing species in the solution and the path length (l) according to the Beer-Lambert’s law:
A= ?.C.l
? is molar extinction coefficient and characteristic for each substance at a particular wavelength (?).
5.2.2 Evaluation of photocatalytic activity of samples
The experimental covered equipment, which has been utilized to check the photodegradation of RhB in the presence of catalysts schematically depicted in Fig. 5.4. The heat effect in the illumination was minimized by condensed water.
10 mg amount of catalyst (as-prepared products (GO and decorated ones)) was dispersed in a 100 mL water by 5 min ultrasonic bath. The solution was added into 400 mL of RhB solution. Initial concentration of our solution should be around 10 ppm of RhB because in the range of 1 to 10 ppm the ratio between absorbance and concentration was linear (Fig. 5.2 (b)). Before starting irradiation, the suspension was stirred moderately in darkness for 30 minutes to ensure of the adsorption/desorption equilibrium of the organic dye. Then illumination under UV-vis light from a TQ 150 W lamp began.
During the illumination, approximately 2 mL suspension was withdrawn at fixed intervals (30 minutes) and analyzed by an UV–vis spectrophotometer over the specific mentioned range.
Fig. 5.4 Experimental set-up for checking the photodegradation of RhB
First test has been carried out in the absence of any catalyst to check the light effect. Then we have done the experiment in the presence of GO, DS/GO, Cu2O/GO, CuO/GO and also pristine DS, Cu2O and CuO to compare the effect of GO as substrate. Furthermore, the effect of GO without illumination has been checked.
We have tested the performance of TiO2 Degussa P-25 as a common photocatalyst reference .

Fig. 5.5 shows the absorbance light vs time of remained RhB in the solutions in presence of different degrading photocatalysts. The main absorption peak of RhB at 554 nm, decreases within the time as the degradation proceeds. The rate of degradation in copper salts/GO nanohybrides is higher than bare salts.

Ultimately, after collecting the data from the experiments and checking the maximum absorbance at 554 nm for each sample in the specific time, a plot of degradation percentage versus irradiation time for the aforementioned samples was obtained.

%Degradation=C-C0C0Fig. 5.6, the plot of degradation vs time shows that the decomposition of RhB under light without any catalyst (?5 % during 240 min exposure) is negligible. Degradation in the presence of GO with and without illumination was 20% and 17% respectively. However, in the presence of Cu2O/GO, DS/GO and CuO/GO nanocomposites, 97.7%, 97.5% and 93.7% of the dye is removed from the solution after 240 min, respectively.

RhB was decomposed in the presence of 10 mg of DS, Cu2O and CuO salts around 34.4%, 51.4% and 41.1% respectively under the same condition. The reference, TiO2 Degussa could decompose 62.8% of RhB in 4 hours. Fig. 5.6. %Degradation vs illumination time of RhB in presence of different catalysts
It is fully demonstrated that copper salts/GO nanocomposites have much higher photo-catalytic activity than the bare ones. This trend might be attributed to the improving of the collection of sunlight, excellent electron mobility and charge transfer in GO, and the higher specific surface area in nanocomposites to absorb a broad spectrum of radiation (see section 4.3). Whereas the decrease in bandgap and consequently promoting the charge efficiency is attributed to the enhanced photocatalytic activity , we decided to check bandgap energy of our catalysts.
5.2.3 Bandgap energy
The bandgap energy or forbidden energy zone (Eg) stands the minimum energy that is required to excite an electron up from a state in the valence band to the conduction band where it can participate in conduction .
One of the best method of determining Eg is based on Kubelka–Munk function evaluated from the diffuse reflectance data (DRS).
The reflectance data can be converted to absorption according to the K-M theory;
where, R? is reflectance, F(R?) is the K-M function.
The optical bandgaps of the catalysts can be estimated by using Tauc plot (Fig 4.14)
where ?, h?, A, and Eg are the absorption coefficient, incident light frequency, proportionality constant and band gap, respectively. The absorption coefficient (?= F(R?)) is obtained from kubleka munk function. The value of exponent ‘n’ determines the nature of electronic transition; for direct transition, n=1/2 and for indirect transition, n=2. The linear extrapolation of (?h?)1/n to zero gives the band gap energy of the sample (Fig 5.7).

Diffuse reflectance spectra (DRS) were measured with Lambda 14P, Perkin-Elmer spectrophotometer, using a 60 mm integrator sphere, the diffuse reflectance mode F(R?) (absolute reflectance) and BaSO4 as pattern. Fig. 5.7. Determination of the bandgap from the Kubelka-Munk transformed reflectance spectra for DS, Cu2O and CuO
The large bandgap energy and rapid recombination of photogenerated charge carriers in the case of bare copper salts limited their performance. Incorporation of the copper salts in GO leads to a decrease in bandgap energy in about 25-35% (Table 5.1).
Table 5.1. Bandgap energy extracted of Fig 5.7
Narrower bandgap reveals that the optical properties of composites has been improved, thereby improving the utilization of light and photodegradation .

The bandgap red shifts can be attributed to the crystalline size/morphology and type of electronic transition due to the covalent link between copper salts and graphene oxide .

5.2.4 Study of photocatalytic degradation kinetic
The narrower bandgap of the copper salts/GO composites in compared to the bare salts, high specific surface area of composites and excellent electron mobility in GO facilitate collection of sunlight, charge separation and stabilization of photo-excited electron- hole pairs which are responsible for photodegradation due to the suppress the recombination .
The heterogeneous catalysis mechanism can be divided in three steps, dye adsorption over catalyst, dye degradation and product desorption from catalyst, Fig 5.8.
Fig. 5.8
Langmuir-Hinshelwood (LH) kinetic model assuming absorption and desorption rate are much faster than RhB degradation.
Langmuir–Hinshelwood rate (LH) is the most commonly used kinetic expression to determine the relationship between the initial degradation rate and the initial concentration of the dye in heterogeneous photocatalytic degradation.
Olga
According to Wachs et.al. one of the interesting parameter to compare heterogeneous photocatalysis activity was T.O.R. (moles converted or produced per gram of photocatalyst per unit of time:
The heterogeneous catalysis mechanism can be divided in three stages, dye adsorption over catalyst, dye degradation and products desorption from catalyst, Fig 11. Langmuir-Hinshelwood kinetic model assuming absorption and desorption rate are much higher than dye degradation and Rhodamine B discoloration, Figures 11 and 12 can describe as a pseudo first order reaction. Figures 11 and 12.
K(kads? kdsp )>>>>k(khet)
In the case of bare cuprous oxide nanoparticles only one aparent constant appears, but nanoparicles /graphene oxide have two different aparent constant, possible due to the change in relative value between absorption desorption and catalytic constant.

Fig 5.9 Langmuir-Hinshelwood kinetic model versus irradiation time for RhB photodegradation in the presence of Copper salts.

Fig.5.10.Langmuir-Hinshelwood kinetic model versus irradiation time for RhB photodegradation in presence of graphene Copper salts hybrids.

The experimental kinetic parameters are detailed in Table 5.2 hybrid catalyst and in Table 5.3 for bare copper oxide nanoparticles and TiO2 like reference due his well know photocatalytic properties.
Table 5.2. Experimental kinetic parameters for GO, CuO, Cu2O, and DS nanoparticles
Table 5.3. Experimental kinetic parameters for GO and its nanocomposites
In all cases, apparent photocatalytic constant and TOR parameter TiO2 reference are higher than copper nanoparticles but copper graphene oxide hybrids improve all result
5.3 Electrocatalytic application
The efficiency of hydrogen fuel generation from photo-electrochemical water-splitting and electrolysis is severely limited by the sluggish kinetics of the oxygen evolution reaction (OER)
2H2O?O2+4H++4e? in acid
4OH??O2+2H2O+4e? in alkaline
Finding an effective electrocatalyst can lower the overpotential needed to sustain an appreciable current, and is therefore an avenue to improve the efficiency of fuel generation technologies .

The electrocatalytic properties of four samples (GO, DS/GO, Cu2O/GO and CuO/GO) were tested and evaluated via performing Oxygen Evolution Reaction (OER) of splitting water.
5.3.1 Experimental
The electrocatalytic properties were measured using a standard three-electrode system on an electrochemical work station (CHI660E, ChenHua, Shanghai, China). A graphite rod and an Ag/AgCl electrode (in saturated KCl) were used as counter and reference electrodes, respectively. The working electrode was a glassy carbon disk electrode with a diameter of 3.0 mm, polished with an Al2O3 paste, and washed thoroughly with ethanol and deionized water.
Typically, 4 mg of catalyst and 40 ?L of a Nafion solution (Sigma-Aldrich, 5 wt. %) were dispersed in 1.0 mL of a water/ethanol solution with a volume ratio of 4:1 by sonicating for 1 h to form a homogeneous ink. Then 5 ?L of the catalyst ink (containing 20 ?g of catalyst) was loaded onto a glassy carbon electrode (loading of ?0.285 mg.cm-2). Finally, the as-prepared catalyst film was dried at room temperature. The potentials reported in our work were referenced to the reversible hydrogen electrode (RHE):
ERHE = EAg/AgCl + 0.059 × pH + 0.1988
The polarization curves were obtained by linear sweep voltammetry (LSV) from 0 to 1 V versus the RHE at room temperature with a sweep rate of 5 mV s-1 was conducted in 0.1 M KOH for each sample. Fig 5.10 shows the electrochemical apparatus and the prepared electrodes.

Fig 5.11 Electrochemical apparatus
5.3.2 Result and discussion
The results were shown in Fig. 5.12. As it follows from these data, decorated GO with CuO exhibits premier activity with nearly 0.65V vs. Ag/AgCl of the onset overpotential, whereas GO shows poor activity with a negligible current response even at large overpotentials.
Fig. 5.12. LSV polarization curves for Oxygen evolution reaction performance of the various catalysts in 0.1 M KOH solution; scan rate: 5 mV-1 s at 25 ° C.

3.1. Introduction
4-1 Introduction