This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
This report is written for my final year project as a literature review. Based on the subject - Preparation of CNT/Epoxy nanocomposite, a brief introduction of Carbon Nanotubes (CNTs) and epoxy, a study of curing behavior of epoxy matrix composite and a detail review of preparation of CNT/Epoxy nanocomposite are given in this report. With its superior properties, CNT has a widely market to improve the performances of current materials in many areas.
The main problem to manufacture CNT/Epoxy composite is that how to achieve a well-condition of CNTs dispersion. Considering the nano-scale of CNTs and the property of aggregation, some methods tried by previous works were reviewed. Mainly, pre-dispersion in solvents and direct dispersion in matrix were commonly used. Different factors which influence on the dispersion of CNTs, like solvents and ambient temperature, were compared according to previous work. The methods to describe the dispersion condition like optical microscope and SEM, were shown to achieve a better understanding of processing route. Further work to manufacture a well-dispersion composite was pointed out at last.
1. INTRODUCTION OF CNTS
1.1 Structures of CNTs:
Carbon nanotubes (CNTs) are generally nano-scale hollow tubes, which are curl formed by a graphite-like plane, composed of carbon six-membered ring . They have a nominal diameter between 2 and 100 nano-meters and a length between tens and hundreds microns, depending on their manufacturing techniques and their surface functional coatings . In generalï¼ŒCNTs can be divided into two types, Single-Walled carbon nanotubes (SWNT) and multi-walled carbon nanotubes (MWNT), according to the numbers of the atomic plane which forms CNTs  .
Fig. 1 The structure of SWNT and MWNT 
SWNT is an ultimate state of CNTs, forming by one carbon atomic plane typically. MWNT has more than one carbon atomic planes. MWNT is more easily manufactured and more popular used than SWNT .
Fig. 2 The SEM micrograph of CNTs 
1.2 Properties of CNTs:
For physical property, CNTs have a structure which is a kind of entangled with each other and no ends condition. CNTs can achieve very high aspect ratios since their diameters are in the range of a few nanometers with lengths of several hundred nanometers .
For mechanical property, many researchers have reported mechanical properties of carbon nanotubes that exceed those of any previously existing materials . CNTs have Young's moduli ranging from 270 GPa to 2 TPa (or five times that of steel) and tensile strength from 11 to 200 GPa (or one hundred times that of steel) and 10-100 times higher than the strongest steel at a fraction of the weight. Their elastic and failure strains are greater than 5% and 12%, respectively . These mechanical property makes them ideal as reinforcing fillers for nanocomposites .
For thermal and electrical property, CNTs have a stable thermally property up to 2800oC in vacuum. The thermal conductivity of CNTs can be about twice as high as diamond  and CNTs have a extraordinary electric-current-carrying capacity, 1000 times higher than copper wires .
1.3 Fabrication of CNTs:
The main methods to fabricate CNTs are arc discharge, chemical vapour deposition (CVD) and catalytic decomposition of hydrocarbons. MWNT is much easier to be manufacture than SWNT as there are less factors needed to control .
The properties of CNTs are limited by the presence of impurity. Hence, reflux wet oxidation and dry oxidation are two methods which are often used for purification . However, the chemical methods partially destroy the structure of CNTs, especially the part of bending branch and tube ends and physical approaches are usually not good at cleaning up the amorphous carbon and carbon nano-particles .
1.4 Modification of CNTs
Functionalization of CNTs is a useful method to introduce the functional groups into the CNTs to expand the application areas of CNTs. As we know, the part of bending branch and tube ends of CNTs can be easily oxidized and transferred into carboxyl . The reactions of functionalization may roughly be divided into two categories: direct attachment of functional groups to the graphitic surface and use of the nanotubes-bound carboxylic acids. "The latter refers to the creation of terminal carbons in the shortening of nanotubes, in which oxidation is converted to carboxylic acids" .
Sun  stated that "The nanotube-bound carboxylic acids are the sites to attach a variety of functional groups for the solubilization of both shortened and full-length carbon nanotubes."
Abdalla  and co-workers made the MWCNTs treated with a mixture of sulfuric/nitric acid to obtain-COOH groups grafted MWCNTs.
Similarly, Zhu and co-workers made SWNTs immersed in a concentrated 3:1 H2SO4 / 70% HNO3 acid to obtain-COOH groups grafted SWNTs.
1.5 Applications of CNTs:
CNTs can be used in the reinforcement of multi-functional nanocomposites like coatings, adhesives, potting compounds, encapsulates, structural materials, liquid crystal display etc .
Glass fiber reinforced epoxy composites,the introduction of MWNT into the composite increased the ILSS by up to 33% .
CNTs also can be used in field-emission displays, scanning probe microscopy tips , and microelectronic devices [9,10].
CURE CHARACTERISTICS OF EPOXY RESINS
2.1 Cure study of epoxy resins
Fig. 3 The structure of Epikote resin 862 and Epikote curing agent W 
From the structure of Epikote Resin 862, it can be obviously seen that a C-O-C ring is linked at the end of a monomer. Usually during polymerization, "the C-O-C rings are opened and the bonds are rearranged to join the molecules" . In similar projects, Epikote Resin 862 is often used as a thermoset polymer matrix with CNTs embedded in. Epikote curing agent W is often used as curing agents. Usually, the curing reaction is divided into two steps. According to the curing mechanism of epoxy, Epikote Resin 862 opens its C-O-C rings first and the bonds are rearranged to connected with each others to produce chains. The chains are co-reacted with curing agents that provide cross-linking. In the last step, because there is usually a C-O-C ring at the end of each chain, the amino groups, which is attached at Epikote curing agent W, is the active groups to react with these C-O-C rings. After that, the chains are cross-linked to achieve its final structure.
Fig. 4 The reaction of -NH2 and C-O-C 
Shanku  and co-workers measured epoxy 862 isothermally at different temperature as a function of time to investigate the curing viscosity. The experiment time required ranged from 450 min at 23oC to 20 min at 100oC. The resin samples were taken from the viscometer chamber at specific times while at these high temperatures and tested using the DSC to obtain the resin degree of cure. Based on the experimental data, a viscosity predicted equation was given:
 â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ Eqs. 1
where K1 and K2 are two parameters assumed to be temperature dependent, and U is a constant assumed to be independent of Î±.
2.2 Factors which affect the cure degree of epoxy resins:
Usually, epoxy value is a standard value to each specific product like Epikote Resin 862. With a same weight, the number of C-O-C rings usually is a constant. During the curing reaction, there are three situations. First one is that the number of active groups on the curing agents is less than the reactive C-O-C rings, which are used for crosslinking. In this situation, the C-O-C rings can not be used up. So the cure degree can not match at a high value. Some chains could still be separable. Second situation is that the curing agents could be enough to make epoxy resin crosslinked and reach the full cured position. It is a variable calculation fields that a number of curing agent can influence the cross-linking level. The cuing degree can be perfect. The last situation is that the curing agents are much more than the consumption of cross-linking. The curing agents would stay their monomers, and be trapped in the thermoset. The curing degree is often too high to be useless.
Usually, the curing reaction needs time. Often, the cure time is determined by kinds of epoxy and curing agents and the ratio of them under an invariable cure temperature and pressure. In general, more cure time a sample has, more cure degree it would reach.
In general, cure temperature has a great influence on cure degree. With the same time and pressure, a sample with higher temperature could reach full cure degree faster than one with lower temperature. Also, heating rate is an important factor, which should be taken into consideration. A faster heating rate often brings higher cure degree.
Pressure is one of factors which could affect the cure degree. However, it does not play an important role, comparing with cure temperature. In general, higher pressure means active groups could have a smaller distance with others. This would be helpful for reaction. So higher pressure could take a higher cure degree.
2.3 Testing equipment
DSC is one of the techniques which are used for the characterization cure degree of epoxy resins. Often, Thermal characterization was performed using a DSC. The curing processes of the epoxy were studied using isothermal scans. Quasi-isothermal scans were conducted in the modulated DSC (MDSC) mode. Total area under the exotherm curve was used to calculate the heat of reaction at each temperature. Cured samples were also tested with DSC to determine the glass transition temperatures using a dynamic procedure .
3. CNTS DISPERSION IN EPOXY RESINS
3.1 CNTs dispersion methods in epoxy resins
Nowadays, many researches have been conducted to develop methods to separate MWNT and disperse them in various polymer matrices.
In order to disperse CNTs in the epoxy resin, Gojnya and co-worker manually added the nano-particles into the neat resin and added the resulting suspension batch-wise to a mini-calender for final high shear mixing. A first primary dispersion of the agglomerates is performed in the knead-vortexes, whereas the final distribution of the nano-particles occurs in the thin gap between the rolls. The suspension was collected, mixed with the hardener for 10 min by intense stirring, cured for 24 h at room temperature (RT) and finally post-cured at 60oC for 24 h.
In Fan  and co-work study, they used a technique combining high-speed mechanical stirring, ultrasonic agitation and acid. First, MWNTs was oxidized by refluxing in a 3:1 mixture of sulfuric and nitric acid for 30 min at 140oC and then filtering with distilled water to wash away any acidic residue from the nanotubes surfaces. Then the oxidized MWNTs were placed in acetone and agitated in an ultrasonic water bath for 1 h. A predetermined amount of epoxy was added and the mixture was agitated for another 1 h. Finally, the mixture of MWNT, epoxy and acetone was placed into a water bath at 80oC and continuously stirred by a high-speed magnetic stirrer for 4 h to fully evaporate the acetone. This method generated a nearly uniformly separated MWNT/ Epoxy suspension as shown in Although there were still a number of 0.5-1 um MWNT aggregates in the suspension, the fraction of separated, single MWNT fibers was significant.
Epoxy composites containing well-dispersed MWCNTs in Guo's work  were prepared by the following procedures: One gram of the crude MWNTs was purified by dilute nitric acid in a water bath for 24 h at 40oC in order to remove the unwanted amorphous carbon and metal catalyst. Then the purified MWNTs were chemically modified by mixed acids. The purified MWCNTs were suspended in a 3:1 (vol/vol) mixture of concentrated sulfur acid (98%) and nitric acid (67%) and sonicated for 5 h at room temperature. The resultant acid-treated MWCNTs were then washed with distilled water until pH 7, and dried at 80oC in a vacuum oven.
Gojnya  and his fellow used a quite bigger dimension about 5mm. The sample they had made could be bigger particle of highly agglomerates. It may not suit for this project. Fan  and co-work had found out a better way to disperse CNTs. First treatment is acid. Then they used acetone as a carrier to transfer the CNTs to epoxy. This step means the main procedure of dispersion of CNTs was finished in a more mobility and polar solution than neat epoxy. This step brought much better dispersion. Guo had done almost the same steps with Fan. A main difference between them is the Guo's method did disperse CNTs in acetone instead of acetone. So in this project, I think Fan's and Guo's works can be referred.
3.2 CNTs/epoxy fabrication
After said above, Guo's method contains that modified MWNTs were manually suspended in epoxy resin and then sonicated for 4 h using a high-power (600 W) ultrasonic machine. During the last 10 min of sonication, a few drops of curing agent EMMZ were added into the system. Thereafter, the blend was degassed for 3 h in a vacuum oven, cast into a stainless steel mold and was cured at 120oC in an oven under atmospheric pressure for 4 h.
Zhu  and his fellows goes their work in this way: the functionalized nanotubes were dispersed in DMFby sonication for 5 min using a high-power CUP-HORN ultrasonic processor and then for 1 h in an ultrasonic bath (40 kHz). Thereafter, the epoxy resin was added, and the solution was stirred for 30 min. DMF was evaporated at 100 Â°C in a vacuum chamber. The SWNT/epoxy blend was prepared by stirring for 5 min with a high-shear mixing homogenizer to ensure good homogeneity. A 100:26 ratio of EPI-CURE W curing agent was added, and further stirring was performed with a high-shear mixer. The blend was degassed for 5 h in a vacuum oven and then cast into an aluminum mould. The curing cycle was 2 h at 100 Â°C under a pressure of 0.3 MPa followed by another 2 h at 160 Â°C. During mixing, an airrelease agent, BYK-A 555, was added to help reduce Porosity. All nanotube/epoxy composites were prepared using a 1 wt % load of both pristine Bucky Pearl SWNTs and functionalized SWNTs.
Abdalla  goes his work in this way: an appropriate amount of MWCNT to prepare 1% CNT samples was dispersed into the epoxy resin by using an extrusion process. Using two syringes, the mixture was extruded from osne syringe into the other syringe with the help of pneumatic device. This process was repeated up to 50 times to ensure a good dispersion. Then the equivalent amount of curing agent was added and mixed using the same process .
In Guo's method, the dispersion of CNTs were manually suspended in epoxy resin directly. The way may cause the lack of CNTs and the aggregates of CNTs. I think this way may be not as good as Zhu's method. In the latter method, DMF was used as solution. Then epoxy was added into DMF. The mixture was treated together. The dispersion of CNTs could be better than Guo's method. However, Zhu's method has a problem about the evaporation of DMF. While, Abdalla's method could be used in factory at a large amount. For this project, I think Zhu's method is more suitable.
3.3 CNTs/Epoxy Cure Characteristics
In Tao's research , Epon 862 was used as epoxy matrix and three types of CNTs, S_SWNT, HiPco-SWNT and XD-CNT, were used. Fig. 5 is one of his results of Dynamic DSC thermograms of the curing of the neat epoxy and epoxy nanocomposites. From the results, CNTs do influence on the curing behaviour of epoxy. A heat flow exothermic peak appears at 190oC. Compare with pure epoxy, composites with CNTs fillers exhibited general low curing energy consumption. More energy had been absorbed in composites with CNTs fillers. Because the amount of CNTs is 1%wt, it is clear to say that CNTs have more complex and influential reaction mechanisms with epoxy matrix.
Fig. 5 Dynamic DSC thermograms of the curing of the neat epoxy and epoxy nanocomposites 
Fig. 6 showed the four types of samples' thermograms at 177oC 2.5h curing. We can see that all curves showed typical one-step changes corresponding to glass transitions. It could be predicted that "all CNT-filled composites exhibited lower glass transition temperatures (Tg) regardless of the rigidity of the nanotubes" . Tao also pointed point that "nanotubes might be a major factor impairing the mobility of the active groups in epoxy and the curing agent, thus leading to significantly lower curing degrees".
Fig. 6 DSC thermograms of the epoxy samples cured at 177oC for 2.5h 
Based on Adballa's work (Fig. 7), it can be obviously found that neat epoxy needs more energy during curing period. This also proves CNTs can have an influence on curing behavior of epoxy, which reduce the energy absorption.
Fig. 7 Isothermal DSC thermograms at 120oC for neat EPON828, 1% F-MWCNT/EPON828 AND COOH-MWCNT/EPON828 nanocomposite 
3. 4 Factors that influence the dispersion of CNTs
3.4.1 the selection of dispersion solvents:
Boiling point (oC)
Table 1 The property of four solvents
Due to the functional groups attached to CNTs, CNTs often has a low polarity, as we know, solute prefers to solute into solvent which has a similar polarity with it. Low viscosity means better working quality. Low boiling point means good evaporation ability. So acetone is a more suitable solvent among these four.
3.4.2 Sonication time
There are three main steps which needs sonication. First one is the time that CNTs are dispersing in solvent. Second one is the time that CNTs are dispersing in solvent and epoxy mixture. The third one is the time that CNTs are dispersing in epoxy and curing agents mixture. Each step could cause a big influence on dispersion.
3.4.3 Ambient temperature
Usually, the ambient temperature is related to the viscosity property of solvent or epoxy. In general, high temperature brings low viscosity. This may be helpful for dispersion. Like in Fan's  work, a water bath temperature for sonication is 80oC.
3.4.4 Functional groups on CNTs
Different functional groups display different polarity, viscosity and density. So it cause a difference between different functional groups on CNTs. Zhu  and his fellow had done some work about this point, "The average aggregate size of pristine BuckyPearl nanotubes in DMF , measured to be 3 um by the scattering method, was significantly reduced to an average size of 300 nm for functionalized nanotubes."
3.5 Techniques which are used for observing the dispersion of CNTs
3.5.1 Optical microscopy
Optical microscopy is an easy way to observe the dispersion roughly. Optical microscopy has its limitations because of its light-wave. Basically, 0.2Î¼m is the smallest resolution. While, the length of single CNT is about 100nm and the width is about 10nm. Because of that, optical microscopy can not be used to get a accurate figure of dispersion. However, as a reason of CNTs aggregation, the feature of dispersion can be distinguished. As shown in Fig. 8, black area is covered with CNTs.
Fig. 8 Optical microscopy pictures taken of the nanotube dispersions (2 mg/mL) in DMF, Pristine BuckyPearl SWNTs 
Lots of work had been done on SEM. From the figure below, a real appearance of CNTs/Epoxy interface had been observed. The distribution and dispersion can also be seen. The best way to investigate CNTs/Epoxy is using SEM. However, using SEM is really a hard work and costs much money. Generally, SEM has a quite large range of resolution. As shown in Fig. 9, lower(10,000x) and higher (50,000x) magnifications can be both observed. Single CNT can be seen using SEM. An accurate figure could be got and a better analysis for its dispersion.
Fig. 9 SEM micrographs of the epoxy composites filled with S_SWNT (a), HiPco-SWNT (b), and XD-CNT (c), at lower(10,000x) and higher (50,000x) magnifications 
After this literature review report, the preparation of CNT/Epoxy nanocomposite was mainly described. A well understanding of this subject was developed. Lots of helpful information were collected. Further work in this subject is that a proper processing route is going to be designed and practiced to achieve a well-condition of CNTs dispersion. Mechanical measurements and observations of CNTs dispersion are going to be taken to provide the advantage of CNT/Epoxy nanocomposite.