Abstract: Epoxy resins with widely different mechanical properties were used as matrices in which functionalized TiO₂ particles were randomly dispersed to produce nanocomposites. These nanoparticles are 21 nm in diameter. They are dispersed in epoxy resin , using ultrasonication or magnetic stirring process for 5, 30 and 60 min. The composites with nanometer scale, TiO₂ particles at 0.5, 1, 5 and 20% weight were characterized by hardness tests and tensile tests. The results are then compared with the neat epoxy resin.

INTRODUCTION

Now-a-days, Epoxy (EP) resins are intensively used in various technical applications such as coatings, composite matrices, potting compounds or structural adhesives. Resin Transfer Moulding (RTM) is rapidly gaining acceptance as one of the most promising manufacturing routes for composite structures in applications such as aerospace and automotive industries (Merad et al., 2009).

Nanocomposites are a novel class of composite materials where one of the constituents has dimensions in the range between 1 and 100 nm. Nanocomposite materials garner most of their material improvements from interactions at the molecular scale, influencing physical and material parameters at scales inaccessible by traditional filler materials (Xu et al., 2004).

However, the effective reinforcement by TiO₂, nanoparticles of thermosetting polymers such as the epoxy resins favored in aerospace and other industries, still presents great challenges (Liu and Wagner, 2005). The mechanical properties of the final nanocomposite materials can be easily measured, using various kinds of standard tests for engineering materials. To achieve maximum performance from the nanoparticles, uniform dispersion and good wetting of nanoparticles within the matrix must be ensured (Xu et al., 2004). The aim of this study is to understand the influence of particles size of TiO₂ on the mechanical properties of epoxy reinforced with nanometer sized particles.

Choice of matrix material and reinforcement agent: The matrix used in this research was a commercially available grade of the diglycidyl ether of bisphenol A ((4-(2, 3 epoxypropoxy) phenyl) propane) abbreviated as DGEBA (Merad et al., 2009). It was used under its commercial designation (Dow Chemical Company; DER 332). Properties of resin cured with the ARADUR 3298 (HUNTSMAN) curing agent are shown in Table 1 and 2.

Table 1:	Matrix material properties

Table 2:	Characteristics of the TiO2 particles

Table 3:	Different times of dispersion

Fabrication of nanocomposites
Casting of TiO₂ epoxy
Sample 1: About 1% weight nanometer TiO₂ particles were added to the epoxy resin, mixed with a glass rod and then dispersed by placing the mixture in a beaker in an ultrasonic bath for about 5 min.

Sample 2: About 1% weight nanometer TiO₂ particles were added to the epoxy resin, mixed with a glass rod and then dispersed by placing the mixture in a beaker in an ultrasonic bath for about 5 min. At the end of the sonication hardener was added.

The dispersion was better in sample 2 because the curing agent was added during the reaction for sample 1 which caused the fluidity of the mixture to decrease resulting in poor dispersion of nanometer TiO₂ particles. The curing agent was simply added at the end of the sonication for sample 2.

Curing cycle: The curing cycle of epoxy resin modified the final properties. The curing cycle selected from those proposed was to cure the samples for 24 h at room temperature and then heated at 100°C for 4 h.

Dispersion of nanometer TiO₂ particles: New developments in the synthesis of nanometer TiO₂ particles have enabled the processing of exciting new nanoparticle/epoxy composites (Ng et al., 1999). Magnetic stirring and ultrasonic methods were used to disperse the nanoparticles in epoxy. Samples were successively removed after different periods of dispersion (typically 5, 30 and 60 min) (Table 3).

Fabrication of specimens: The specimens had nominal length L of 24 mm height, width of 4 mm and a thickness B of 2 mm as per NF T 51-034 standard; all samples were polished at room temperature for 30 min in order to have a plane area.

MATERIALS AND METHODS

Testing and characterization
Hardness test: It was performed with BUHLER MACROVICKERS 5114, equipped with a microscope that allows you to position the imprint on the desired phase. The force applied is selected from the 10 values between 100 mN to 20 N. We measured the average of the two diagonals, 0.002 mm meadows through the microscope micrometer connected to the machine.

Tensile test: Tensile testing of the TiO₂ epoxy composites was performed using an INSTRON 8801 with a 5 KN load cell at a crosshead speed of 0.01 mm sec^-1. The tensile test was carried out at room temperature. Typical sample dimensions were 4 mm (width) 2 mm (thickness) 24 mm (length). To ensure data accuracy and repeatability, five samples of each type were tested. In this case at least five samples were measured for each specimen type.

RESULTS AND DISCUSSION

Figure1 shows the effect of different modes of dispersion of nanometer TiO₂ particles ultrasonication or magnetic stirring process vs. the hardness.

The usefulness of nanometer sized particles, however is marred by processing difficulties such as the formation of particle agglomerates and non-uniform dispersion. This has motivated research for finding better ways of processing, the particles to achieve de-agglomeration and ensure better dispersion. Techniques such as ultrasonication and magnetic stirring have been used by Zunjarrao and Singh (2006).

Figure 1 shows the results obtained for the addition of TiO₂ nanoparticles which increases the viscosity of the resin. However, prolonged agitation significantly reduces the hardness of the nanocomposite. Aggregates start to form and eventually, we get the opposite of what is expected (Guan et al., 2006). Similarly, magnetic stirring is not a good solution as the viscosity of the TiO₂/epoxy is very important.

The curing agent was simply added during the dispersion of TiO₂ nanoparticles by ultrasonication for 30 min. The particles, thus treated were then dispersed in epoxy and composite was fabricated in the same way as before (Fig. 1).

Hardness test
Effect of load and times: The hardness tests are done on both sides of the samples. The hardness values are the same. Dispersion of nanofillers is therefore consistent throughout the resin and there is no effect of gravity that would have formed some agglomerates in the lower part of the epoxy resin during the reaction. The value of the load will then be used; we tested the influence of the load on the hardness of a specimen of neat resin. Initially, the hardness of the unit continues to increase. But from about 3000 mN, hardness stabilizes. We therefore, chose to work with a force of 2942 mN because the machine works by predetermined increments. Figure 2 shows the influence of the duration of application of this force on the same sample.

Fig. 1:	Effect of different mode dispersion TiO2 particles vs. hardness in 1% weight nanometer TiO2 particles in epoxy resin

Fig. 2:	Influence of the duration

We note that while the hardness decreases with time the reduction diminishes and tends to stabilize (Ratna et al., 2003). This variation is interesting and reveals the viscous behavior of material found for example in the creep. This part has led to the preparation of samples that will be the following:

Study on TiO₂-epoxy nanocomposite: The weight percent nanometer TiO₂ particles in the following samples:

Fig. 3:	Influence of the weight percent nanometer TiO2 particles vs. hardness

Figure 4 shows the hardness of steel XC 100 and nanocomposite. The steel hardness is about 215 HV. For comparison of the two materials, the hardness has been normalized at 1 (Hardness/hardness_max). After 500 sec, the hardness of nanocomposite decreased by >15% while that of the steel XC 100 dropped by only 5%. Hardness of steel has leveled but that of the nanocomposite stabilization has not yet stabilized.

Tensile test: Tensile tests measurements were carried out for neat epoxy and epoxy reinforced with TiO₂ nanoparticles. For each case, five samples were tested. The results were obtained for the Young’s modulus experiments.

Fig. 4:	Hardness of steel XC 100 and Nanocomposite

Table 4:	Young’s modulus for the neat epoxy

Table 5:	Young’s modulus for 1% TiO2

Table 6:	Young’s modulus for 5% TiO2

Table 7:	Young’s modulus for 20% TiO2

The statistical spread of experimental data is denoted by the standard deviation (Table 4-7). The presence of individual TiO₂ particles embedded in an epoxy matrix. There were some agglomerates but overall, the dispersion was good. During the fabrication of test specimens, they were degassed of any trapped air bubbles. These air bubbles result in fragile specimens and are initiation sites of breaking. The specimen will fracture at a stress lower than it would without bubbles. We thus obtain a fairly large dispersion for the breach of the same series of specimens. Between different Nanocomposites, comparison cannot be done directly for the stress and strain to failure because there is significant uncertainty of the results. However, we can compare the Young’s modulus. It was tensile tests with specimens from samples of neat epoxy 1, 5 and 20% TiO₂ (Table 4-7).

The Young’s modulus increases with the weight percent TiO₂ particles. Conversely if 5% weight of TiO₂ nanoparticles is exceeded, the Young’s modulus will decrease (Huang et al., 2006). The material at 20% weight of TiO₂ particle is very fragile. The break occurs for a weak constraint either because the high weight percent of TiO₂ prevent the complete polymerization reaction and due to the higher occurrence of agglomeration under the gravitational interaction between TiO₂ nanoparticles which hinder the extrusion of epoxy resin’s molecular chains or because the presence of very large aggregates acting as sites of boot failure like a bubble.

Fig. 5:	Normalised values vs. weight percent of TiO2 nanoparticles

Collects the results obtained during different tests (Fig. 5). The evaluation of various mechanical properties compared against the weight percent of TiO₂. These differences shows that low levels of particles almost do not affect the characteristics of the material.

Too much of these particles will also not be useful. It is assumed that the particles results in difficulty of obtaining a complete epoxy resin cure which makes the material less durable. If a significant change in characteristics of the epoxy resin is desired, adding 5% TiO₂ appears to be very interesting because all the features of nanocomposites (E = Young’s modulus, stress at break = sigma R, strain at break A% and Hardness (HV) are 35% better compared with neat epoxy. The biggest improvement is in the properties at break. The Young’s modulus defining rigidity has a maximum increase at 10% weight of TiO₂ while the stress and strain at break increase at 20 and 35% weight of TiO₂ nanoparticles.

CONCLUSION

Systematic mechanical property characterizations including hardness and tensile properties for a new functionalized TiO₂/epoxy composite were conducted. Investigations were carried out on the influence of some process parameters on the dispersion of nanofillers and the mechanical properties of the nanocomposite were obtained but we have to go further in testing. In fact, we tried to get the best possible protocol to disperse the nanofillers but be aware that this dispersion was not perfect and there is still room for improving the protocol to improve the dispersion of nanofillers in epoxy resin. Mechanical tests on nanocomposites with different weight percent of TiO₂ nanoparticles showed that they have certain influence on the mechanical properties of nanocomposite. The nanometer size TiO₂ was most interesting at 5% wt. of TiO₂.

ACKNOWLEDGEMENTS

The researchers are also grateful to the University of Mascara and University of Tlemcen for their support, encouragement and interest for this study.