Dichlorobenzene: an effective solvent for epoxy/graphene nanocomposites preparation

It is generally recognized that dimethylformamide (DMF) and ethanol are good media to uniformly disperse graphene, and therefore have been used widely in the preparation of epoxy/graphene nanocomposites. However, as a solvent to disperse graphene, dichlorobenzene (DCB) has not been fully realized by the polymer community. Owing to high values of the dispersion component (δd) of the Hildebrand solubility parameter, DCB is considered as a suitable solvent for homogeneous graphene dispersion. Therefore, epoxy/graphene nanocomposites have been prepared for the first time with DCB as a dispersant; DMF and ethanol have been chosen as the reference. The colloidal stability, mechanical properties, thermogravimetric analysis, dynamic mechanical analysis and scanning electron microscopic images of nanocomposites have been obtained. The results show that with the use of DCB, the tensile strength of graphene has been improved from 64.46 to 69.32 MPa, and its flexural strength has been increased from 97.17 to 104.77 MPa. DCB is found to be more effective than DMF and ethanol for making stable and homogeneous graphene dispersion and composites.


Introduction
Epoxy resins can be seen to be widely used in aerospace, automotive, marine, structures and construction, owing to their superlative mechanical properties and favourable thermal stability, which are important for various applications [1,2]. Epoxy resins are of particular interest to engineering because these resins provide a unique balance of chemical and mechanical properties combined with ease of processing [3].
For epoxy/graphene nanocomposites, graphene can significantly improve the physical and chemical properties of

Characterization
Tensile properties, flexural properties and fracture toughness were measured by a universal mechanical testing machine (Instron 3382); for all tests the crosshead speed used was 2 mm min −1 . Tensile tests were conducted according to ASTMD638 (Type V) with a specimen thickness of 4 mm. Flexural properties were measured according to ASTM D790 with a specimen dimension of 48 × 12.7 × 3 mm. A singleedge-notch three-point bend specimen was used to measure the fracture toughness (K 1C ) according to ASTM D5045; the specimen dimension was 36 × 6 × 3 mm with a crack of length of 3 mm. The K 1C was calculated using where W is the sample width (millimetres), a is the crack length and kept between 0.45 and 0.55 W, B is the sample thickness (millimetres), f (a/w) was calculated using equation (2.2), which is constant and related to the sample geometry and P max is the maximum load of the load-displacement curve. The critical strain energy release rate (G 1C ) was calculated using equation ( and Vickers microhardness was measured by Buehler Micromet II. A dwell time of 10 s with a load of 200 g was used for all the samples. Six samples were measured for each set of conditions and mean values were then reported. A Perkin Elmer-8000 dynamic mechanical analysis (DMA) instrument was used to conduct the analysis to determine the storage modulus (E ) and loss factor (tan δ). Sample dimensions of 30 × 8 × 2.5 mm were tested by the single cantilever method. All tests were carried out under the temperature sweeping mode (temperature ramp from 30°C to 150°C at 5°C min −1 ) at a constant frequency of 1 Hz. The glass transition temperature (T g ) was taken as the temperature value at the peak of tan δ curves. Thermogravimetric analysis (TGA) of the nanocomposites was carried out with a TA Instruments Q500 thermal analyser. The temperature range was from room temperature to 600°C at a ramp rate of 5°C min −1 under N 2 atmosphere. To evaluate the fracture modes of the nanocomposites, a FEI Quanta 200 electron microscope was used to obtain scanning electron microscopy (SEM) images. A layer of gold with 10 nm thickness was applied to the samples' fractured surfaces using an Emscope sputter coater model SC500A.

Visual stability
Successful fabrication of nanocomposites crucially depends on maintaining a stable dispersion of graphene before polymer curing. Figure 1 shows the colloidal suspensions of graphene in DCB, DMF and ethanol at different time intervals after sonication. The pictures show that graphene settled down in ethanol within two hours after sonication; the graphene-DMF suspension also reaggregated significantly and settled down within 12 h of sonication. However, stable dispersion was achieved only by DCB, which suggests that DCB is the best for preparing uniformly dispersed epoxy/graphene nanocomposites. Figure 2 shows the SEM image of as-received stacked multilayer graphene.

Mechanical properties
The uniformly dispersed graphene in the matrix will in return affect the macroscopic properties of the nanocomposites, while most of the work on epoxy/graphene nanocomposites aims at exploiting the mechanical enhancement of the nanocomposites. Therefore, the mechanical properties of nanocmposites have been tested and are summarized in   hardness of 0.247 GPa. Meanwhile, DMF-and ethanol-prepared samples show only medium increase in the properties, indicating that DCB produces better graphene dispersion than DMF and ethanol. To sum up, the improvements in the macroscopic properties of epoxy/graphene nanocomposites are due to the enhanced dispersion of graphene. The uniformly dispersed graphene forms a continuous network in the matrix, which could release the stress concentration, and thus enhance the mechanical strength and improve the energy-absorbing capacity. Moreover, the uniformly dispersed graphene has changed the microstructure of the polymeric network, which will be discussed in below.

Thermogravimetric analysis test
Thermal decomposition is one of the fundamental thermal properties and is critical for practical applications. Figure 4 shows the TGA curves of the nanocomposites in a nitrogen atmosphere. It can be seen that all samples have a similar two-stage weight loss, indicating that all the samples have a similar thermal degradation mechanism. The first weight loss from 100°C to 230°C was attributed to the decomposition of small molecules on the side chain. The second weight loss that occurred from 250°C to 500°C shows the decomposition of the main polymer chain [54,55]. As can be seen from the figure, DCB samples show the lowest decomposition rate, which means the highest thermal stability compared with that of the DMF and ethanol samples.
The reason of this phenomenon can be explained by the fact that graphene has affected the cross-linking structure of the matrix. Generally, the cross-linking density means the concentration of cross-linked bonds per volume. For typical polymeric materials, the higher the cross-linking density, the stronger the polymer chains bond with each other, thus resulting in a stronger capacity to withstand heat. Compared to the structures of DMF-and ethanol-prepared samples, DCB-prepared samples tend to produce more uniform graphene dispersion, and thus an increase in the cross-linking density [    On the other hand, the homogeneously dispersed graphene could form a continuous network in the matrix, and thus could reduce the volatilization rate of the decomposition products.
In general, the increased thermal stability of DCB samples resulted in a higher heat capacity of nanocomposites and a better barrier effect of the graphene network, which was caused by the uniform dispersion of graphene by using DCB.    The segmental motion of polymers is characterized by glass transition temperature (T g ), which was taken as the temperature value at the peak of tan δ curves and are shown in figure 5b. In this figure, it can be seen that the tan δ peak is observed at 69.28°C for G-0.3 samples. For nanocomposites processed by DCB, DMF and ethanol, the T g were observed at higher temperatures. This can be ascribed to the phenomenon that the uniformly dispersed graphene has restricted the chain mobility of the epoxy matrix, and thus increased the T g values [58]. Among all the samples, DCB-prepared samples show the highest T g of 76.57°C, which is a more than 7°C increment compared with the G-0.3 samples, while only slight increases (approx. 4°C) in T g are obtained for samples prepared with DMF and ethanol. As mentioned above, the uniformly dispersed graphene increased the cross-linking density of the nanocomposites, and then plays a positive role in improving the thermal stability.

Scanning electron microscopy test
The fracture surfaces of nanocomposites were examined by SEM and are shown in figure 6. For G-0.3 samples, as shown in figure 6a, graphene aggregates are sparsely located on the surface; the inset of figure 6a shows the typical morphology of a graphene aggregate. This poorly dispersed surface indicates a poor interfacial interaction between the epoxy matrix and graphene, and features the brittle nature of the material and its poor resistance to crack initiation and propagation. Compared to the G-0.3 samples, the fracture surfaces of DCB samples are smoother, as shown in figure 6b. The clear fracture patterns show the fracture mechanism of sheet/sheet delamination of this nanocomposite, and reveals that the usage of DCB leads to uniform graphene dispersion. The homogeneously dispersed graphene is able to bridge growing cracks, impede crack propagation and thus improve the properties of nanomaterials. However, for DMF-and ethanol-prepared samples, as shown by the insets in figure 6c,d, some poorly dispersed graphene can still be seen on the surface. The non-uniformly dispersed graphene forms defects in the nanocomposites, which act to concentrate the stresses locally, effectively causing a localized weakness, thus decreasing the properties of the nanocomposites.

Conclusion
A prerequisite for property enhancement of epoxy/graphene nanocomposites is the uniform dispersion of graphene in the matrix; however, the ultra-high specific surface area of graphene results in high van der Waals forces and thus induces a strong tendency to reaggregate. Solvents are able to provide a low-viscous medium, which is an ideal environment for graphene to disperse. While dispersing, lower viscosities are responsible for promoting the de-bundling of graphene sheets. This results in the triggering of the van der Waals force on graphene surfaces in lower proximity with each other, thus leading to dispersion. Therefore, the selection of dispersion media is very important for the final properties of nanocomposites.
In this work, DCB was selected to test its effectiveness for epoxy/graphene nanocomposites preparation. The colloidal stability and mechanical properties were determined, and TGA, DMA and SEM imaging of nanocomposites were conducted. The results show that DCB is able to produce stable graphene dispersion; however, DMF and ethanol are not sufficient to disperse graphene homogeneously in the nanocomposites. Nanocomposites prepared with DCB also show higher mechanical properties and better thermal stability compared to the samples prepared with DMF and ethanol.
In general, it is concluded that DCB is more efficient than DMF and ethanol for making uniform and stable graphene dispersions. The usage of DCB can endow the nanocomposites with outstanding mechanical properties and improved thermal stability. This finding can be significant in practical applications and gives an impetus for DCB usage not only in epoxy/graphene nanocomposites preparation, but also in other polymer nanocomposites where the usage of solvents is required in the processing.
Data accessibility. All the data are provided in the main text. Authors' contributions. J.W. conducted most of the laboratory work and wrote the manuscript. M.S.S. assisted in part of the experiments and testing. T.V. and F.I. supervised the project and gave research guidelines. All the authors contributed to the revision and proofreading of the manuscript.