SC-15 epoxy procured from Applied Polyremic, Inc., was selected as the matrix in preparation of nanocomposites. SC-15 Epoxy is a two-phase, low viscosity toughened resin system. This resin system is excellent for structural and ballistic applications. SC-15 has two parts: part A consists of diglycidylether of bisphenol-A and aliphatic diglycidilether epoxy toughener whereas part B consists of hardener which is a mixture of cycloaliphatic amine and polyoxylalkylamine. This is room temperature curing resin having viscosity 300 cps, density 1.09 g/cc and pot life of 6 hours. This resin system can be cured at high temperature also.

Amino functionalized multi-walled carbon nanotubes (MWCNT-NH₂) were procured from Nanocyl, Belgium. These nanotubes (nanocyl-3152) have average diameter of 10 nm, length of 1 micron, carbon purity >95% and functionalization <0.5%. Carboxyl functionalized multiwalled carbon nanotubes (MWCNT-COOH) was procured from Nanostructured & Amorphous Materials Inc., USA. These nanotubes have average diameter of 10 - 20 nm, length of 10 - 30 micron, carbon purity >95% and functionalization <2.1%. Figure 1 shows the probable synthesis reaction of carboxyl and amino functionalized CNTs [17]. This figure illustrates that carboxyl group could be formed firstand may have higher number of reactive sites that could lead to better interaction with polymers.

MWCNTs were mixed with Part A of the epoxy resin system as per required weight ratio. The mixture was then sonicated (Sonics Vibra Cell ultrasonic processor) for 1 hour at 35% amplitude on pulse mode (20 second on/20 second off). During the sonication process, there is a significant increase of pressure and temperature which is highly localized resulting in the increase in temperature of polymer-CNT mixture. In order to reduce this temperature increase, the container was kept in a cooling bath filled with refrigerant maintained at 5˚C. After the sonication process, mixture was added to a 3-roll machine (high shear mixer) for three passes. The gap setting among the rolls was 20 µm (1^st pass), 10 µm (2^nd pass) and 5 µm (3^rd pass) and the highest speed was 150 rpm in each pass. After sonication and calendaring process, Part B of epoxy resin system was added as per stoichometric ratio (Part A:Part B = 10:3) to the 800 rpm. During mechanical stirring, air bubbles are introduced in the resin mixture. To remove entrapped air and volatile materials from the mixture, a degasification process was introduced by placing the mixture in a vacuum oven for 35 minutes at room temperature. Two curing cycles were

used; in cycle A, the mixture was cured at room temperature for 24 hours and post-cured at 93.33˚C for 4 hours. In cycle B, mixture was cured at 65.56˚C for 5 hours in a convection oven. Schematic diagram of fabrication of nanocomposites with epoxy containing functionalised MWCNTs is shown in Figure 2.

Flexure tests were performed according to ASTM D 790-99 under a three point bend configuration using Zwick Reoll Z 2.5 machine. The machine was run under the displacement control mode at a crosshead speed of 1.2 mm/min. All the tests were performed at room temperature. Test samples were cut according to ASTM D 790-99 where length to thickness was 16:1. At least seven samples were tested from each type.

Dynamic mechanical analysis (DMA) was performed according to ASTM D4065-01 to study viscoelastic behaviour of composite samples. DMA test was performed using TA Instrument DMA Q 800 operating in a three-point bending mode at an oscillation frequency of 1 Hz and amplitude of 15 µm. The temperature was ramped from 30˚C to 200˚C at a heating rate 5˚C/min. The span length was 35 mm and the span length to thickness ratio was 10. Average sample size was 60 × 12.5 × 3 mm³. Data were collected from machine and

data analysis was performed using Universal Analysis 2000-TA Instruments Inc. data acquisition system. Glass transition temperature, loss modulus and storage modulus of samples were determined from the tests to evaluate the viscoelastic and damping properties of nanocomposites.

Thermo-mechanical analysis (TMA) was performed according to ASTM D696 using TA Instrument Q400 to determine coefficient of thermal expansion (CTE) of composites. Nitrogen was used during the tests. The samples were cut into small pieces of dimensions 6 × 6 × 4 mm³. The dimensional change was measured in the thickness direction. Tests were performed from 30˚C to 200˚C at a heating rate 5˚C/min. Four to five samples were tested for each type.

The analysis of fracture surfaces was carried out using JEOL JSM-6400 scanning electron microscope (SEM) at 5 kV accelerating voltage. Specimen surfaces were positioned on the aluminum holder using carbon tape and coated with a thin gold film to increase their conductance for SEM observation.

Flexure tests were performed to evaluate flexural strength and stiffness of neat and nanophased epoxy SC-15 samples and carbon/epoxy laminates. Typical stress-strain curves from flexure tests are shown in Figure 3. Most of samples failed immediately after reaching maximum stress. Summary of flexure test results are shown in Table 1. Typical stress-strain behavior from the flexural tests is shown in Figure 3(a). The stress-strain curves showed considerable non-linearity before reaching the maximum stress, but no obvious yield point was found in the curves Flexure strength and modulus of the nanophased epoxy cured using two cycles increased continuously with MWCNTs content up to 0.2 wt% loading as shown in Figure 3(a) and Table 1. Here, neat epoxy SC-15 cured atroom temperature is considered as control sample. Flexural strength and modulus of room temperature cured (cycle A) samples with 0.2 wt% MWCNTsNH₂ improved by 11.3% and 28%, respectively. Samples prepared at high temperature cure (cycle B) with 0.2 wt% MWCNTs-NH₂ loading showed better properties with 31.36% and 37% improvement in flexural strength and modulus, respectively in comparison to control samples. Flexural properties were found to lower with increase in cure time under isothermal condition.

Table 1. Flexure test results of epoxy SC-15 samples with different wt% of MWCNT-COOH and MWCNT-NH_2.

Similar results have been reported in the literature [18, 19].

Polymerization process is accelerated at elevated temperatures. Effect of temperature on reaction rates is calculated using the Arrhenius equation.

where k is the rate coefficient, A is a constant, E_a is the activation energy, R is the universal gas constant, and T is the temperature (in kelvin). R has the value of 8.314 × 10⁻³ kJ·mol⁻¹·K⁻¹.

Samples with 0.2 wt% MWCNTs-COOH loading prepared at high temperature cure (cycle B) showed best properties with 41.45% and 48% improvement in flexural strength and modulus, respectively in comparison to control samples. Good dispersion of CNTs and crosslinking of functional groups present on the surface of CNTs with polymer that restricts the mobility of polymerchains enhanced strength and modulus. In addition to good interfacial adhesion between the CNTs and matrix high aspect ratio, high modulus and strength of CNTs, also contribute to enhancement in the properties. However, decrease of strength at higher CNT content can be attributed to following two effects: creation of voids and agglomeration of CNTs in the sample that may have decreased the strength and improper dispersion of CNTs at higher loading. Acoustic cavitation is one of the most efficient methods to disperse nanoparticles at small loading [20]. Choi et al. claimed that few voids were created during the fabrication process and that voids increased with the higher nanoparticle contents [21]. Agglomerations of CNTs act as areas of weakness and cause stress concentration leading to lower flexure properties of epoxy samples modified with CNTs at higher loading.

Figures 4(a)-(f) shows the scanning electron micrograph (SEM) of fractured neat samples and samples with 0.2 wt% and 0.3 wt% CNTs processed at room and high temperature (65.56˚C). Rougher fracture surfaces in were noticed in all samples with 0.2 wt% MWCNTs irrespective of the curing cycle in comparison to neat epoxy samples. A dense rough surface was observed in samples with 0.2 wt% MWCNT-NH₂ cured at high temperature (see Figure 4(c)) whereas much uniform and rougher surface was observed in samples with 0.2 wt% MWCNT-COOH cured at high temperature (see Figure 4(e)). Figures 4(b) and (c) demonstrated that MWCNT-NH₂ sample cured at elevated temperature had more closely space rougher surface when compared with that cured at

room temperature. When the CNTs loading increased to 0.3 wt%, the failure mode of nanocomposite changed. Agglomeration of CNTs was noticeable for both types of CNTs. A crack initiated at an agglomerated location forsample with 0.3 wt% MWCNT-NH₂ (Figure 4(d)) whereas a localized failure without crack was noticed for sample with 0.3 wt% MWCNT-COOH (Figure 4(f)). This type of stress concentration could be due to agglomeration of several MWCNTs and formation of bubbles leading to voids due to viscous mix of epoxy and MWCNTs. The fissures or cracks in Figure 4(d) indicates that samples with 0.3 wt% MWCNT-NH₂  cured at high temperature cured samples had an earlier stress failure compare to sample with 0.3 wt% MWCNT-COOH cured at high temperature.

Generally after mixing MWCNT-NH₂ with epoxy part A, the interfacial reaction takes place between amino functional groups of CNTs and epoxy part A which consists a ring opening reaction followed by a cross-linking reaction creating interlocking as shown in Figure 5. MWCNT-NH₂ contains -NH₂ group which is a strong nucleophile that opens the epoxide. The resultant product

contains an epoxide ring. In this way, MWCNT-NH₂ show cross-linking with epoxy resin. This crosslink reaction creates interlocking structure in the resin blend through the covalent bond which facilitates impediment of the mobility of polymer chains in the system. This improved interfacial interaction may have facilitated better load transfer and thus resulting in increased flexural properties. Similar reasons for improvement in epoxy incorporated with functionalized MWCNTs were also reported previously [1,22] . On the other hand, Figure 6 shows the interfacial reaction of MWCNT-COOH and epoxy resin which creates interlocking. Epoxide opening is governed by strong nucleophile. Since -COOH groups are not so strong nucleophile to open the epoxidegroup, they react with epoxy part B which has a pair of electron. The resultant product has -NH₂ which is a strong nucleophile and susceptible to attach the epoxide group and therefore open the epoxide ring. The resultant product contain -OH group which is strong nucleophile and susceptible to react further with epoxy part A. Thus MWCNT-COOH shows cross linking with epoxy system. In our research, MWCNT-COOH has higher functional group (2.1%) than MWCNT-NH₂ (0.5%). For this higher functionalization in MWCNT-COOH, MWCNT-COOH showed higher cross linking reaction with epoxy system than MWCNT-NH₂ which  improved interfacial interaction leading to better stress transfer during loading and thus better flexural properties.

Dynamic mechanical Analysis (DMA) was used for characterization of visco-elastic material properties as function of temperature. Storage modulus, loss modulus and glass transition temperature of fabricated epoxy samples were determined by DMA. Figures 7-9 show the DMA results of epoxy samples. Table 2 shows the average storage modulus, loss modulus and glass transition temperature of epoxy samples with different wt% loading of MWCNT-NH₂ and MWCNT-COOH. Epoxy samples modified with 0.2 wt% MWCNT-COOH showed best storage modulus, loss modulus and glass transition temperature. Table 2 shows that storage modulus of samples with 0.2 wt% loading MWCNT-COOH improved by 47% compared to control. Glass transition temperature improved by 13˚C compare to control system.

In Figure 7, positive effect of adding CNTs was ob-

Table 2. DMA results of epoxy polymer.

served on storage modulus, loss modulus and glass transition temperature. Glass transition temperatures (T_g) were determined from onsets of the sharp drop in tan δ curve (Figure 9). T_g increased by adding CNTs as the mobility is reduced in epoxy matrix around the CNTs by the interfacial bonding and thermal stability became higher [16]. The improved interfacial bonding between CNTs and epoxy system can lead to a higher shift of T_g. The carboxyl-functionalized MWCNTs are supposed to react with epoxy SC-15 and create covalent bonding which is the cause of reduction in matrix mobility and a stronger shift of T_g. Table 2 shows that visco-elastic properties are increasing with addition of CNTs up to 0.2 wt% and then decreases for 0.3 wt%. Formation of micro agglomeration could be a reason for decreasing properties in 0.3 wt% MWNCT samples [23].

Figure 8 shows the storage modulus of epoxy samples in rubbery plateau region. Cross-linked density can be calculated by the storage modulus in this rubbery plateau region (150˚C - 200˚C) by using following equation:

In this equation, λ is cross-linked density, G is storage modulus in rubbery plateau region, R is universal constant (8.314 J·mol⁻¹·K⁻¹) and T is absolute temperature in Kelvin. Using above equations the calculated crosslinked density values are listed in Table 3. From Table 3, it is observed that the cross-linked density is highest for 0.2 wt% MWCNT-COOH modified epoxy samples. Because of highest cross-linked density, properties of 0.2 wt% MWCNT-COOH modified epoxy samples have the best values compared to other types of epoxy samples [24].

Thermal expansion of samples was measured by using thermo-mechanical analyzer (TMA). Coefficient of thermal expansion (CTE) before and after glass transition temperature (T_g) were measured. Five samples were tested in each type. Table 4 shows the average CTE values of epoxy samples before and after T_g. It was observed that dimensional change at higher temperature was higher than the dimensional change at lower temperature. The CTE was measured from the slope of curves. The thermal expansion was investigated before (glassy region) and after (rubbery region) glass transition temperature [25]. In both states, 0.2 wt% MWCNT-COOH modified epoxy samples showed best improvement compare to neat epoxy samples. When 0.2 wt% MWCNT-COOH were added to epoxy, CTE value improved by 15% and 25% glassy region and rubbery region respectively as compared to neat samples.

Figure 10 shows the positive effect of CNTs in CTE values due to good dispersion in epoxy that has reduced the segmental motion of molecules within the resin system. Improvements in properties of MWCNT/epoxy samples prepared using cycle A were comparatively

Table 3. Calculated cross link density of epoxy polymer samples.

Table 4. TMA results of epoxy polymer samples.

lower than samples prepared with cycle B. MWCNT-COOH modified epoxy samples showed better mechanical and thermal properties than MWCNT-NH₂ modified samples. 0.2 wt% MWCNTs modified epoxy samples fabricated with cycle B showed optimum enhancement in strength and modulus. Similar improvements were observed in thermo-mechanical properties. Flexural and thermo-mechanical results demonstrated maximum improvement in 0.2 wt% MWCNT-COOH modified epoxy samples prepared using cycle B.