Graphene Nanosheets as Novel Nanofillers in an Epoxy Matrix for Improved Mechanical Properties

Abstract

In this paper, we will be interested in bending tests on a polymer matrix reinforced with graphene nanofillers. The mechanical behaviour and the damage kinetics were determined. The samples were made using controlled dispersions of graphene nanosheets (GNP) in EPON 862 matrix. Various samples with different contents of GNP were made (0, 0.5, 1, 2.5, 5 and 10% by weight). Mechanical properties such as maximum stress, strain at break and Young’s modulus were determined. After each test, the fracture surfaces were characterised using optical microscopy (OP) and scanning electron microscopy (SEM). Experimental results show that the fracture toughness of the GNP/epoxy-based nanocomposites decreases with an increasing percentage of nanofillers. The flexural strength of the samples with 10 wt% of graphene significantly decreased compared to neat epoxy. Based on stress-strain data and the fracture surface analysis, graphene nanosheets seem to impact the mechanical behaviour and the kinetics of the damage. The influences of the weight percentage of GNP on the EPON matrix properties and the performance of the nanocomposites are discussed. In addition, the evolution of bending performance and damage kinetics with graphene content was obtained and analysed.

1. Introduction

Polymer-based composites were introduced in the 1960s as new structural materials. It disperses stiff, durable, and high-strength fibres in a polymer matrix. As a result, we fabricated lightweight composites [1]. Graphite nanosheets, carbon nanotubes (CNTs), and functionalised graphene have been used as new additives due to their excellent mechanical strength. In particular, graphene, a single layer of aromatic carbon, has one of the strongest among all materials and outstanding thermal and electrical properties [2, 3, 4, 5]. The latest research on the impact of graphene sheets as a filler in polymers suggests that the addition of graphene can significantly improve the resulting composite materials’ mechanical, thermal, and electrical properties. The graphene filler can enhance the polymer’s stiffness, strength, and toughness and increase its thermal and electrical conductivity [6]. Furthermore, integrating graphene into a polymer matrix can produce a new type of mechanical material by combining graphene’s exceptional mechanical properties with the polymer’s structural properties. Kuilla wrote an excellent state-of-the-art review article on “Recent advances in graphene-based polymer composites” [7].

Moreover, the dispersion of graphene in the polymer matrix is a critical factor in determining the properties of the resulting composite material. Researchers are exploring various methods for achieving a high degree of dispersion, such as graphene’s chemical functionalization and surfactants’ use.

The mechanical performance of composites with continuous fibre in a polymer matrix is well documented [8, 9, 10]. However, these composites have drawbacks that are essentially related to the intrinsic properties of the matrix and which subsequently limit their field of application [11, 12]. Recently, polymer-based nanocomposites have generated enormous interest because of their ability to improve the mechanical performance of epoxy resin [11, 13, 14]. Nanocomposites containing Nano additives like carbon black (CB), carbon nanotubes (CNTs) and graphene nanosheets (GNPs) offer excellent mechanical properties [15, 16, 17]. Many kinds of fillers, such as metal particles [18, 19], carbon nanofillers (CNF) [20, 21], carbon nanotubes [22, 23] and graphene [24, 25, 26], have been used. Graphene offers much better mechanical performance than other nanofillers, resulting in exceptional mechanical strength, significant elongation and large surface area [27, 28].

Mohan et al. [29] present a study of the impact of carbon allotropes on a self-powered triboelectric humidity sensor based on a flexible polydimethylsiloxane (PDMS) film. The authors used different carbon allotropes such as carbon nanotubes, graphene, and reduced graphene oxide as additives to the PDMS film. The study found that adding carbon allotropes to the PDMS film increased the humidity sensitivity of the sensor and improved its response time. The authors also observed an increase in the film’s dielectric constant with the addition of carbon allotropes, which is attributed to the adsorption of moisture and the formation of an electric double layer. The results suggest that adding carbon allotropes to PDMS films can create a self-powered triboelectric humidity sensor with improved performance.

Ai et al. [30] explore the potential of using graphene/electrospun carbon nanofiber sponge composites induced by magnetic particles as a multi-functional pressure sensor. The composites were fabricated via electrospinning and then exposed to a magnetic field. The results showed that the resulting composite had a high sensitivity, good repeatability and durability, and a low hysteresis effect. Furthermore, the composite demonstrated excellent performance in terms of pressure sensing, temperature sensing, and humidity sensing. This demonstrates the potential of graphene/electrospun carbon nanofiber sponge composites for use in pressure-sensing applications.

Topal et al. [31] present a novel method to improve the performance of repaired secondary load-bearing aircraft composite structures. This technique employs scarf-bonded patches as part of a multiscale nano-integration strategy. This method preserves the structural integrity and durability of restored composite structures while minimising the loss of mechanical strength. This technique employs a nanomaterial-based adhesive that permanently attaches the scarf-bonded patches to the composite framework. The results of this study show that the performance of the repaired composite structures can be improved significantly by using this approach.

Wang et al. [32] have developed and studied different graphene nanofillers such as reduced graphene oxide, graphene nanosheets (GNP) and three-dimensional graphene. They demonstrated that epoxy reinforced with nanofillers has exceptional mechanical properties. In another work, an epoxy adhesive’s mechanical properties and toughness were improved with graphene nano additives [33].

The epoxy matrix material is a composite’s most commonly used primary phase because of lesser shrinkage, outstanding adhesion, and better solvent resistance. Some applications of epoxy-based nanocomposites include aerospace, automotive, marine, sporting goods, industrial tooling, and other general consumer products. To the best of the authors’ knowledge, the effects of adding GNPs on the flexural properties of the epoxy matrix have not been reported in the literature. Herein, the epoxy matrix has been reinforced by graphene nanosheets (GNP) with different weight percentages. The effects of adding (GNPs) with different concentrations of GNPs (0, 0.5, 1, 2.5, 5 and 10% by weight) on the flexural properties of the polymer matrix were investigated. Enhanced GNPs-epoxy interface bonding is achieved using the three-roll mill. Homogeneous dispersion is further facilitated through sonication. The evolution of the flexural strength with the graphene content was obtained and analysed. The effect of nanofillers on the kinetics of damage is also reported.

2. Materials and methods

These samples include EPON Resin 862 epoxy. The graphite used in these studies was a natural flake with a 500 m diameter. First, exfoliated graphite (ExG) was manufactured using a mixture of nitric, sulphuric, and natural graphite. Intercalation with graphene sheets appears to produce an intercalated graphite product after 24 h of reaction. The mixture was subsequently filtered, rinsed with water, and dried at low temperatures in an oven. The intercalated graphite compound was then subjected to a hurried heat treatment at 900°C and a rapid extension.

Figure 1 displays the morphological characterisation of ExG nanofiller by SEM at three magnifications: 1 mm, 100 μm, and 20 μm. ExG nano-filler has a cellular structure associated with considerable expansion, as seen in SEM images. Before exfoliation, graphite exists as flakes with the graphite c-axis perpendicular to the plane of the flake. Due to the substantial expansion along the c-axis, the exfoliated flake becomes longer in the direction corresponding to the flake’s c-axis prior to exfoliation. Figure 1a depicts the ExG generated from a graphite flake under low magnification (1 mm) and high magnification (100 μm). Figure 1b depicts the ExG under high magnification (100 μm). Figure 1c shows graphene sheets with thicknesses varying from 1 nm to around 16 nm and widths ranging from sub-micrometre to hundreds of m at 20 m magnification. Figure 1 depicts that the gap between graphitic stacks is roughly 5–10 nm.

Figure 1.

Size distribution and SEM image of the graphene nanosheets.

To fabricate samples, a traditional method of dry mixing and adding a different mass percentage of graphene (0.5, 1, 2.5, 5 and 10 wt%) to the epoxy resin was used. The reference polymer and the GNP/epoxy nanocomposites were treated under the same conditions to ensure the samples’ homogeneity. This study used the exfoliation and reduction method to produce graphene, Figure 1d. The synthesis of the product requires the use of Epon 862 resin and the hardener Epikure W. This epoxy matrix was selected for its characteristics at ambient temperatures, such as low viscosity and ease of handling. In addition, the chosen hardener offers good chemical resistance at high temperatures and excellent performance. Figure 2 shows the molecular formulae of both compounds. One of the big problems related to the synthesis of graphene and its dispersion in a polymer is the phenomenon of agglomeration. To overcome this problem and a dry mixing process was used.

Figure 2.

Molecular formulation of EPON Resin 862. (a) Epon 862 Resin (b) Epikure W Hardener.

Six concentrations of GNP/epoxy (0, 0.5, 1, 2.5, 5 and 10 wt%) were used. The control sample consists of pure resin. It has been noted that for concentrations greater than 10% by weight, the mixing process becomes ineffective and difficult to obtain a well-dispersed filler.

GNPs were mixed with the resin/hardener solution for the desired concentration using a three-roll mixer. The same procedure was adopted for all percentages except for 10 wt% because the mixture’s viscosity increases as the number of nanofillers increases. Therefore, several steps are required to prepare the specimen at 10 wt%. First, a small percentage of graphene is mixed with the resin using the three-roll mill until the mixture’s viscosity is reduced, and then we add more GNP. This step was repeated several times until all the graphenes were mixed uniformly in the resin, Figure 3. Based on previous studies, the three-roll mill is the most appropriate apparatus to obtain a uniform GNP/epoxy mixture [28].

Figure 3.

Schematic diagram of GNP/epoxy nanocomposite production fabrication.

The speeds of the three cylinders 1, 2, and 3 are 21, 64, and 200 rpm, respectively. The space between the rolls was kept small to produce significant shear stress. Here, the distance between the rolls was fixed to 25 μm. It was decided to pass the mixture several times in the mill for all formulations. As graphene content increases, so does the speed of the feed roll, Table 1. Since there is an increase in the temperature in the system, which is generated by the shear stress, the temperature of the rolls was maintained at a temperature of 21°C using a cooling system (Figure 3).

Formula	GNP (%)	Velocity (rpm)
1	0	333
2	0.5–1	400
3	2.5	480
4	5	550
5	10	600

Table 1.

Velocity of the feed roller.

Table 2 shows the mechanical properties of graphene nanosheets and epoxy matrix.

	GNP (%)	Epon 862
Young Modulus (GPa)	1030	2.72
Poisson coefficient	0.19	0.3

Table 2.

Properties of epoxy and graphene [29, 30, 31, 32, 33].

3. Results and discussion

This study made composite specimens consisting of an Epoxy matrix reinforced with GNP. Six specimens with 0, 0.5, 1, 2.5, 5 and 10% of GNP were manufactured and experimentally tested under Bending (ASTM D2344). The mechanical bending performances of the epoxy matrix reinforced by graphene nanofillers were evaluated. Specimens with L = 50 mm, W = 15 mm, and T = 5 mm has been used. The schematic representation of the dimensions of the specimens studied, and the bending device is given in Figure 4. The specimens are placed on the two supports 10 mm in diameter and are separated by a fixed distance.

Figure 4.

Schematic presentation of test specimen dimensions and bending test. (a) Dimension of specimens (b) Bending test.

A bending test was used to study the mechanical behaviour of neat epoxy and GNP/epoxy nanocomposites. In addition, a bending test was used to assess the effect of nano-reinforcement on the mechanical characteristics of epoxy reinforced with graphene nanosheets. The three-point bending test was conducted using the universal testing machine (Instron type 5585H). The load frame was equipped with Instron’s ± 50 kN static cell. All measurements were taken at room temperature, and all tests were performed at a speed of 0.5 mm/min. The mechanical properties of manufactured specimens were evaluated for each GNPs fraction using the bending test’s force-displacement and stress-strain curves. For reproducibility of results, all specimens were subjected to a set of bending tests (3 tests).

3.1 GNPs distribution in an epoxy matrix

In this work, the distribution of GNP in the epoxy matrix was analysed using SEM and optical analysis. Figure 5a shows optical microscope images of graphene-filled epoxy. Arrows indicate the graphene sheets. In all cases, one can see that the graphene nanosheets are well dispersed all over the surface. Figure 5 also shows that the dispersion of GNP in epoxy is reasonably uniform. Figure 5 gives an overview of the GNP/epoxy nanocomposites identifying the graphene sheets, while Figure 6 shows a zoom with a detailed view. Figure 5b illustrates the surfaces of GNP/epoxy nanocomposite, showing the graphene morphology and their dispersions into a polymer matrix. The graphene platelets were exfoliated, exhibiting a wide range of interatomic interlayer spacing. They are well dispersed with no evidence of aggregates. This morphology has an essential role in the mechanical performance and load transfer from the epoxy matrix to graphene sheets, as reported by*** Shen et al. [34].

Figure 5.

Overview of GNP/epoxy nanocomposite surfaces. (a) Distribution of GNP in epoxy matrix and (b) zoom of GNP/epoxy nanocomposite and graphene morphology.

Figure 6.

Stress-strain curves for a different formulation of GNP/epoxy composites. (a) GNP, 0 wt%, (b) GNP, 0.5 wt%, (c) GNP, 1 wt%, (d) GNP, 2.5 wt%, (e) GNP, 5 wt% and (f) GNP, 10 wt%

3.2 Bending properties of GNP/epoxy nanocomposites: effect of GNPs

GNPs were randomly distributed in the Epoxy matrix. The mechanical properties of the GNP/epoxy samples were determined for each mass fraction using a flexural test. For each GNP/epoxy configuration, five tests per configuration were carried out to ensure the reproducibility of the tests. Figure 6 illustrates the reproducibility of obtained results. The elastic moduli, peak stress, strain at break, and stiffness were evaluated for each test.

Figure 7 summarises the results of the tests carried out on the specimens with different percentages of graphene: 0.5, 1, 2.5, 5, and 10 wt%. Again, the effect of nanofillers on the overall behaviour of the material is apparent and noticeable.

Figure 7.

Bending behaviour of different formulations of GNP/epoxy composites.

As shown in Figure 8, the resin with added nanofillers has a lower resistance than the neat epoxy. The variation of the bending properties (Peak stress, Strain at break, Stiffness) of GNP/epoxy specimens with graphene content is shown in Figures 8 and 9. As expected, the stiffness of the specimens increased with the increase of the mass fraction of the nanofillers. However, the peak stress and the strain at break (or strain at maximum stress) dropped when adding graphene nanofillers. The improvement can be attributed to GNP’s high flexural modulus. The nanocomposite reinforced with 1% GNP demonstrated a more significant increase in elastic modulus, as seen in Table 3. The GNP/epoxy mixture with 5% GNP has a slightly lower flexural modulus than pure epoxy. For 10% GNPs, the addition of graphene nanosheets has almost no effect on the stiffness of the base material.

Figure 8.

Bending properties of GNP/epoxy composites. (a) Peak stress vs. wt.% GNP and (b) strain at break vs. wt% GNP.

Figure 9.

Stiffness of GNP/epoxy composites.

GNP (wt%)	0	0.5	1	2.5	5	10
Average (MPa)	3203.12	3425.65	3569.57	3410.22	2998.21	3250.36
Performance (%)	—	+6.95	+11.44	+6.47	−6.40	+1.47

Table 3.

Flexural modulus variation vs. wt% GNPs.

The bending strength of the GNP/epoxy nanocomposites decreased peak stress and strain at break, Figures 6 and 7. This degradation has already been noted in some research [35, 36, 37], which has explained the poor dispersion of graphene. The formation of aggregates can cause a stress concentration. The interfacial strength controls the ultimate mechanical properties. Thus the weak interfacial adhesion between GNPs and the epoxy matrix and the aggregation of nanosheets explain the strength reduction. As can be seen in Tables 4 and 5, the resistance of nanocomposites decreased with the increase in the GNP contents. As expected, the presence of GNPs led to more fragile materials because the peak load and the stress at break are conditioned by the interactions between the phases and the chemical bonds between the graphene and the epoxy matrix, which have been confirmed by different studies [38].

GNP (wt%)	0	0.5	1	2.5	5	10
Average (MPa)	161.28	87.54	85.93	67.12	44.86	41.43
Performance (%)	0	−45.72	−46.72	−58.38	−72.19	−74.31

Table 4.

Peak stress variation vs. wt% GNPs.

GNP (wt%)	0	0.5	1	2.5	5	10
Average (MPa)	8.802	3.0955	2.7445	2.169	1.839	1.4965
Performance (%)	0	−64.83	−68.82	−75.36	−79.11	−83

Table 5.

Strain at break variation vs. wt% GNPs.

Tables 4 and 5 show the average peak stress and strain at break versus GNP mass fraction. Adding GNPs significantly affects the peak stress and strain at failure, which decreases with the addition of GNPs. The peak stress decreases by 45.72, 46.72, 58.38, 72.19 and 74.31% for 0.5, 1, 2.5, 5 and 10 wt%, respectively. As presented in Table 1, the elastic part of the curve was slightly higher for samples containing 0.5, 1, and 2.5 wt% GNP than the neat epoxy. Considering the standard deviation, the stiffness of all specimens is nearly consistent, Figure 8. However, the material’s stiffness decreased when the contents of GNP exceeded 2.5%. It can be concluded that, in the case of bending tests, the addition of GNPs causes a decrease in mechanical performance.

3.3 Bending resistance of GNP/epoxy nanocomposites

Figure 8 and Tables 2 and 3 show the evolution of the mechanical properties for the different specimens with different graphene contents. The five nanocomposites (GNP/epoxy) have maximum stress and a maximum strain lower than the neat epoxy. Indeed, under three-point bending tests, the higher the graphene content, the more it weakens the material (the bending strength continues to decrease).

After each test, the fracture surface of specimens was examined by an optical microscope (Kayence) and scanning electron microscope (MEB). Figure 10 visualises the different damage mechanisms in the Kayence and MEB micrographs of the fractured specimen (central region), which are activated during the bending test of the manufactured composite. From these results, it is clear that, under bending stress, these materials (matrix doped with GNP) have an unfortunate tendency to be damaged rapidly compared to neat epoxy (their performance decreases); this is the result of changes in their structure. By examining these various evolutions, we will be able to describe and classify precisely the various types of damage to materials. The embrittlement is the outcome of internal material changes that are not accompanied by the formation of new surfaces but lower the material’s ductility. The first explanation for the decrease in flexural strength as the graphene content increases is the aggregation of GNP in the epoxy matrix, Figure 11. The second reason is that nanofillers tend to form a barrier for the deformation of the specimens, which leads to a faster failure with too low levels of load and deformation.

Figure 10.

Fracture facies, GNP/epoxy with different percentage of GNP at different scales.

Figure 11.

Simulation of GNP distribution in the epoxy matrix. (a) 1 wt% GNP and (b) 5 wt% GNP—with agglomeration.

Microscopic analyses of the GNP/epoxy samples were carried out to understand the relationship between their morphology and mechanical behaviour. The analysis of the stress-strain curves, as well as the fracture surfaces, allows us to state the following points:

• Dispersion of nanofillers in a reference material is a recurrent problem of nanocomposites, but it is more visible for a high percentage. This dispersion can cause the formation of agglomeration (Figure 11) and poor adhesion, which will cause the fall of the performances of the nanocomposites. So, we can say that the more the nanofiller percentage increases, the more this problem is present.

• The adhesion of the GNPs to the matrix is not perfect, Figure 10. As previously stated, graphene nanofillers act as a barrier to crack propagation and limit the deflection of the specimens. The kinetics of damage, unfortunately, does not favour the hardening of these nanocomposites because of the weakness of the GNP/epoxy interface.

• The bending test of the GNP/epoxy results in various damage mechanisms responsible for the suppression of crack initiation and propagation in the composite. As the graphene content in the epoxy matrix increases, the propagating crack begins to interact with the rigid GNP; it must tilt or twist at greater angles to pass the GNP due to the high aspect ratio of the particles. As a result, the fracture surface becomes rougher with more peaks and valleys.

• From the bending test, it is clear that the dispersion and doping of the epoxy matrix with different mass ratios of graphene nanoparticles has an effect not only on the load/stress at break and the bending modulus but also on the mode of failure. Indeed, the behaviour of the GNP-doped epoxy makes the material more brittle with a plastic zone that reduces with increasing GNP mass percentage. For 10% of nanofillers, the behaviour is purely brittle; failure occurs at the end of the elastic domain.

• The results show that the increase in the GNP mass fraction degrades the properties of the epoxy; they did not improve the epoxy performance. In Figure 10, the crack guidance mechanism can also be visualised. In crack guidance, cracks propagate along the GNPs, and lower fracture energy is consumed for the reinforced epoxy; this is because the GNPs hinder crack propagation in the load plane due to their unique wrinkling characteristic and strongly limit the plastic zone. In GNP-epoxy nanocomposites, the higher the percentage of GNP, the more localised the crack propagation path is (parallel to the direction of the applied load). This phenomenon is attributed to the robust bonding interfaces between the GNPs and the matrix grains. In contrast, extensive plastic deformation is observed for pure epoxy, leading to a bifurcation or branching of the crack, as shown in Figure 10.

• Two situations are to be considered (a) for low plastic deformations of the matrix, the sliding is still very heterogeneous, and the Zener mechanism of dislocation stacking can be considered on the obstacle constituted by the GNP inclusion (b) for more significant plastic deformations, the field of these deformations becomes much more homogeneous, in which case the mechanics developed by Eshelby provides answers.

• For this type of test and material, we can consider that the stress concentration in the inclusion is an increasing function of the strain-hardening slope and depends on the shape factor of the inclusion. This factor takes exceptionally high values for needle-shaped inclusions parallel to the maximum principal stress and disc-shaped inclusions stressed in their plane. We understand that such inclusions are those that cleave or whose interface breaks first. Hence, they are particularly harmful from the point of view of the damage; this is the case, for example, of GNP inclusions in the epoxy matrix.

• The distribution of GNP (volume fraction of inclusions) is far from constant within a material. Instead, we often find clusters of high volume fraction (agglomeration effect). The damage develops primarily in these zones. While this more or less random distribution of inclusions has a negligible influence on the elastic behaviour of the material, it plays an essential role in the damage. Once again, disorder favours it.

4. Conclusions

In this work, we studied the effect of graphene nanofillers (GNP) on the bending performance of an epoxy matrix. Six weight percentages of graphene nanosheets (0, 0.5, 1, 2.5, 5 and 10 wt%) were prepared and tested. At this level, the first conclusion is that flexural modulus increased with increasing GNP, reaching a maximum value of 1 wt% (an increase of 12%). However, the flexural strength of the nanocomposites decreased as the percentage of graphene increased, which can be explained by the formation of graphene aggregates and could also be the result of the interfacial bonding between the epoxy matrix and the GNP nanosheets. Toughness (peak stress and strain at break) drops with increasing graphene content for all nanocomposite formulations. Nanocomposites with 10 wt% of GNP are the least resistant to flexural stress. However, GNP/epoxy resulted in a higher increase in the flexural modulus with a maximum of 1 wt% of nanofillers. GNP/epoxy nanocomposites have a higher modulus than neat epoxy due to fewer crosslinked networks. The shear strength for nanocomposites at 0.5% GNP is greater than for 10% GNP, which is a consequence of agglomerates as the GNP content increases. The chemical bonds formed between the epoxy matrix and GNP had a major effect on the mechanical behaviour and the strength of specimens.

To our knowledge, and after analysis of the bibliographic references, there is no work dedicated to quantifying the performances of a matrix doped by nanofillers under 3-point bending type solicitations. Nevertheless, this work constitutes an exciting database for scientists and industrialists and complements our work on nanocomposites and understanding their mechanical responses and damage kinetics under static and dynamic loads [39, 40, 41, 42, 43, 44, 45, 46, 47, 48].

Conflict of interest

The authors declare no conflict of interest.