Influence of hematite nanorods on the mechanical properties of epoxy resin

The mechanical properties of nanocomposites obtained by incorporation of fairly uniform hematite nanorods (α-Fe2O3 NRs) into epoxy resin were studied as a function of the content of the inorganic phase. A thorough microstructural characterization of the α-Fe2O3 NRs and the nanocomposites was performed using transmission electron microscopy (TEM) and atomic force microscopy (AFM). The TEM measurements revealed rod-like morphology of the nanofiller with a uniform size distribution (8.5 nm×170 nm, diameter×length). High-magnification TEM and AFM measurements indicated agglomeration of α-Fe2O3 NRs embedded in the epoxy resin. Stress at break, strain at break, elastic modulus and tensile toughness of the nanocomposites were compared with the data obtained for pure epoxy resin. Significant influence of nanofiller on the mechanical properties of epoxy resin, as well as on the glass transition temperature, could be noticed for samples with low contents of the inorganic phase (up to 1 wt. %).


INTRODUCTION
Epoxy resins have found wide applications in adhesives, coatings, electronic encapsulants, medical devices, and optical and structural components due to their superior properties. 1 However, the brittle nature of most epoxy resins has driven extensive research into the influences of different fillers on their mechanical pro-perties.Hitherto, a variety of fillers, such as rubber particles, glass beads, thermoplastic particles, inorganic particles, as well as combinations of these materials have been used in order to improve mechanical properties of epoxy resins. 2 Recent developments in nanoscience, including synthetic procedures for size and shape control of various nanoparticles, have presented possibilities to prepare epoxy-based nanocomposites with significantly improved mechanical properties.3][4][5][6][7][8][9] Nanoparticles can significantly alter the mechanical properties of a polymer matrix, even at low concentrations, due to their high surface to volume ratio and consequential small inter-particle distance.Incorporation of nanorods into a polymer matrix can improve the mechanical properties of the matrix that cannot be achieved using the same amount of a spherical nanofiller.1][12] Interaction between nanofillers and polymer chains alters the mobility of the polymer chains, which is indicated by a change in the glass transition temperature. 13Most studies concerning inorganic nanoscale fillers and epoxy resins have been conducted using commercially available spherical nanoparticles, such as SiO 2 , 3,5,8,9,[13][14][15][16] Al 2 O 3 , [17][18][19][20] TiO 2 , 21 hybrid TiO 2 -SiO 2 , 22 ZrO 2 23 and SiC. 24On the other hand, there is a lack of information about the influences of zero-dimensional (0D) and onedimensional (1D) inorganic nanofillers 25 on the mechanical properties of epoxy resins, with the exception of graphene, 26 carbon nanofibers, [27][28][29] and carbon nanotubes. 30To date, ferric oxide particles of different sizes and morphologies, 0D (spherical and cubic) and 1D (rod-like), have been used to improve the thermo-mechanical properties of thermoplastic polymers (polystyrene [31][32][33] and poly(methyl methacrylate) [34][35][36] ).To the best of our knowledge, only the influence of commercially available spherical Fe 2 O 3 particles (diameter in the size range from 20 to 60 nm) on the mechanical properties of epoxy resins has been reported in literature. 20n this study, uniform hematite nanorods (α-Fe 2 O 3 NRs) with average diameter of 8.5 nm and aspect ratio of about 20 were introduced into epoxy resin with the view of improving the mechanical properties of the composite systems.Mechanical performance of nanocomposites (stress at break, strain at break, elastic modulus and tensile toughness) with low contents of inorganic phase (up to 1 wt.%) was compared with the mechanical properties of pure epoxy resin.The morphology of the obtained nanocomposites was studied using different microscopy techniques, while the influence of the content of α-Fe 2 O 3 NRs on the glass transition temperature (T g ) was studied using differential scanning calorimetry (DSC).

Synthesis of uniform hematite nanorpds (α-Fe 2 O 3 NRs)
Powder consisting of α-Fe 2 O 3 NRs was obtained by courtesy of Dr. Milena Marinović--Cincović.The synthetic procedure and structural characterization of the α-Fe 2 O 3 NRs are described in a previous work. 35Briefly, a dispersion consisting of α-Fe 2 O 3 NRs was obtained by forced hydrolysis of FeCl 3 solution in a manner similar to the method developed by Matijević. 37Thus, 100 mL of 5.4 M NaOH was added to 100 mL of 2 M FeCl 3 , and the solution was stirred at room temperature for 15 min.Subsequently, the solution was heated to 100 °C and the temperature was maintained for 8 days.The α-Fe 2 O 3 NRs were recovered by centrifugation, washed several times with water and dried in a vacuum oven.As expected, the XRD analysis indicated that synthesized powder was α-Fe 2 O 3 with a corundum crystal structure, known for weakly ferromagnetic or superparamagnetic behavior depending on particle size. 31reparation of α-Fe 2 O 3 /epoxy nanocomposites A commercially available epoxy resin, CHS-Epoxy 210 X 75, was used as the polymer matrix, which was cured with Epikure 3115 X 70 hardener, using a weight ratio of 100:35.Nanocomposites with different contents of α-Fe 2 O 3 NRs were obtained according to the same procedure.The required amount of α-Fe 2 O 3 NRs was dispersed in 3.9 g of xylene (isomeric mixture) , added to 6 g of CHS-Epoxy 210 X 75, and mixed in an ultrasonic bath for 10 min.Then, the corresponding amount of curing agent, Epicure 3115 X 70, was added.The obtained mixture was additionally mixed in an ultrasonic bath for 10 min and then poured into a Teflon dumb-bell-shaped mould.The curing of the so-prepared samples was performed at room temperature for 21 days.The declared content of inorganic phase was calculated with respect to the total mass of epoxy resin and curing agent.A pure epoxy resin sample, without α-Fe 2 O 3 nanorods, was prepared under the same experimental conditions.

Characterization of the α-Fe 2 O 3 /epoxy nanocomposites
The morphology and particle size distribution of the α-Fe 2 O 3 powders as well as the morphology of the α-Fe 2 O 3 /epoxy nanocomposites were analyzed using a JEM-2100 LaB6 transmission electron microscope (TEM) operated at 200 kV.In addition, the topography and heterogeneity "bulk" relief from the nano to micrometer levels was realized by AFM.The nanocomposite sheets were measured after previous freeze-fracture at liquid nitrogen temperature.An atomic force microscope (Dimension Icon, Bruker), equipped with an SSS-NCL probe, Super Sharp SiliconTM-SPM-Sensor (NanoSensorsTM Switzerland; spring constant 35 N m -1 , resonant frequency ≈170 kHz) was used for the analysis.Measurements were performed under ambient conditions using the tapping mode AFM technique.AFM scans covered areas from 1×1 μm 2 to 50×50 μm 2 .
Mechanical measurements of the pure epoxy resin and the α-Fe 2 O 3 /epoxy nanocomposites were performed on a Galdabini 1890 (Quasar 5) tensile testing machine with nominal capacity of 5 kN.The dumbbell-type specimens were 2.50±0.05mm thick and 4 mm wide at the neck.For stress-strain measurements, the crosshead speed was 20 mm min -1 .The presented values for the stress at break, the strain at break and the Young's elastic modulus are the average of at least five runs.The tensile toughness of examined nanocomposites was determined from the area underneath the stress-strain curves.
Differential scanning calorimetry (DSC) was performed at a heating rate of 20 °C min -1 under a nitrogen atmosphere using a Perkin Elmer DSC-2 instrument.

RESULTS AND DISCUSSION
The shape and size distribution of α-Fe 2 O 3 particles were estimated using TEM.A typical TEM image at low magnification of the α-Fe 2 O 3 NRs is shown in Fig. 1A.It could be noticed that the α-Fe 2 O 3 NRs had a tendency to array themselves parallel to each other and to form bundle-like aggregates.The corresponding size distribution histograms for the diameter and length (Figs.1B and C, respectively) were determined on a population of over 250 particles.The average diameter and length of α-Fe 2 O 3 NRs were found to be 8.5 and 170 nm, respectively, and consequently the aspect ratio of rod-like particles was 20.The standard deviations of the diameter and length were found to be 2 and 75 nm, respectively.Based on the particle size distribution data, it is clear that it is more difficult to control length than the diameter of α-Fe 2 O 3 NRs.It is well known that the ability to control length of nanorods decreases with increasing aspect ratio, while, on the other hand, their tendency to array parallel to each other and to form bundle-like aggregates increases. 34The morphology of the α-Fe 2 O 3 /epoxy composites was studied using AFM and TEM measurements.The AFM phase image of the epoxy filled with 0.5 wt.% of inorganic phase is shown in Fig. 2A.It is clear that an epoxy-rich phase and inorganic-rich phases existed after the addition of α-Fe 2 O 3 NRs, indicating a high degree of agglomeration of the nanofiller.Additional TEM measurements at high magnification (Fig. 2B) confirmed that the α-Fe 2 O 3 NRs formed bundle-like aggregates in the epoxy matrix, although some individual nanorods were also visible.It should be kept in mind that no attempt was made to improve compatibility of epoxy resin and α-Fe 2 O 3 NRs by surface modification.The glass transition temperatures (T g ) of the pure epoxy resin and the α-Fe 2 O 3 /epoxy nanocomposites were determined by DSC.The DSC thermograms of the examined samples are shown in Fig. 3.The values of the glass transition temperature were taken as the midpoint of the glass transition event and are listed in Table I.Except for the sample with the highest content of α-Fe 2 O 3 NRs (1.0 wt.%), all the other composite samples had a lower glass transition temperature than that of the pure epoxy resin.The composite sample with 1.0 wt % of α-Fe 2 O 3 NRs has the same T g as the pure epoxy resin.The ratio of epoxy resin to hardener was constant for all the prepared samples, which were treated in the same way.Moreover, no post-curing reactions were observed by DSC.This indicates that the maximal degree curing was reached in the examined samples and that all samples have the same crosslinking density.The obtained results indicate that the observed reduction of the T g of nanocomposites was caused by diffusion of free volume holes towards the filler-polymer interfaces. 38,39This reduction of T g is directly related to the ratio of the surface the α-Fe 2 O 3 filler/volume of epoxy matrix.In the sample with 1.0 wt.% of α-Fe 2 O 3 NRs, agglomeration of the nanoparticles and formation of aggregates occurred.This drastically reduced the interfacial area between the α-Fe 2 O 3 NRs and the epoxy matrix, making the influence of the filler on the glass transition temperature negligible.These results are contrary to the results obtained by Dudić and co-workers. 40They examined the electrical properties of nanocomposites prepared from hematite nanorods and epoxy resin.The calculation of the T g of the examined samples was based on the results obtained by conductivity measurements.Thus, the estimated values of the T g indicated that incorporation of hematite nanorods shifted the T g of the epoxy matrix towards higher temperatures compared to the pure epoxy resin.The increase in the T g was explained by the reduced mobility of the epoxy segments in the presence of high specific surface fillers.The curing agent was a low molar mass diamine while, in the present case, the curing agent was polyaminoamide.According to Dudić and co-workers, a layer with reduced mobility was formed at the interface of the polymer/nanoparticles during curing, because the diamine could easily reach the surface of the nanoparticles.In the present study, this layer was not formed because the polyaminoamide hardener could not reach the filler surface so easily during the curing reaction and this is a possible reason for the contradictory results.I.The stress and strain at break values, and the Young's modulus of pure epoxy resin were 26.3 MPa, 7.0 % and 410 MPa, respectively.The values for the stress at break, strain at break and elastic modulus for a nanocomposite depends on the content of inorganic phase.For example, a continuous increase in the strain at break with increasing content of inorganic phase could be observed with the exception of the sample with the highest content of inorganic phase (1.0 wt.% of α-Fe 2 O 3 NRs).The strain at break for sample with 0.5 wt.% of α-Fe 2 O 3 NRs was higher by 49 % compared to that for the pure epoxy resin.The strain at break of all nanocomposite samples was larger than the strain at break of the pure epoxy sample.This results in a larger area under stress-strain curves, indicating increasing toughness of the epoxy matrix.6][47][48][49][50][51] Tensile toughness data of the prepared α-Fe 2 O 3 / /epoxy nanocomposites and pure epoxy resin, calculated as the area under the stress-strain curve, are given in Table I.The increase of tensile toughness from 1.06 MJ m -3 for pure epoxy resin to 2.79 MJ m -3 for the nanocomposite with 0.5 wt.% of α-Fe 2 O 3 NRs could be noticed, while a further increase of filler content caused a decrease of the tensile toughness (2.02 MJ m -3 for the nanocomposite filled with 1.0 wt.% of α-Fe 2 O 3 NRs) due to filler aggregation.Dudić and co--workers 40 also showed that the fracture surface of the pure epoxy resin was relatively smooth and featureless while a pronounced roughness was observed in the case of the α-Fe 2 O 3 nanorods/epoxy resin nanocomposite, indicating that the α-Fe 2 O 3 nanorods promoted ductile deformation of the epoxy matrix.The 2.63-fold increase in the tensile toughness upon incorporation of 0.5 wt.% of unmodified α-Fe 2 O 3 NRs into the epoxy resin is at least comparable to or better than that attained by other nanofillers.For example, incorporation of 1.0 wt.% of TiO 2 increased the tensile toughness 1.70-fold, 21 1.0 wt.% of Gr/ /Fe@Fe 2 O 3 increased the tensile toughness more than 2-fold, 27 15.5 wt.% of clay increased the tensile toughness by about 1.75 times, 48 1.0-wt.% of Fe@FeO increases the tensile toughness more than 1.5-fold, 49 while incorporation of 13.4 wt.% of nanosilica increased the fracture toughness of epoxy resin by 2.40 times. 50n the other hand, the stress at break values for the nanocomposites with low contents of inorganic phase (≤0.5 wt.%) were slightly lower compared to that for the pure epoxy resin, while the value for the nanocomposite with the highest concentration of α-Fe 2 O 3 NRs (1.0 wt.%) was 13 % larger than that for the pure epoxy resin (29.7 and 26.3 MPa, respectively).The values of the Young's modulus of the nanocomposites were lower compared to that of the pure epoxy resin, except the sample containing 0.2 wt.% of α-Fe 2 O 3 NRs.
According to the obtain results, the most likely toughening mechanism is plastic void growth.The repulsive interactions at the interface between the α-Fe 2 O 3 NRs nanofiller and the epoxy matrix facilitate debonding of the nanofiller and formation of voids around the α-Fe 2 O 3 NRs, which allows plastic deformation of the matrix.
The yield point could be observed from the stress-strain curves of the nanocomposite samples with 0.2 and 0.5 wt.% of α-Fe 2 O 3 NRs.The yield stress was about 32 MPa for these samples, but at the different elongations (a higher value for the sample with the higher content of inorganic phase).Most likely, this effect is a consequence of alignment of the α-Fe 2 O 3 NRs together with the polymer chains when yielding occurred.
The obtained results indicate a significant influence of the 1D nanofiller on the mechanical properties of the epoxy matrix even when the content of inorganic phase was smaller than 1 wt.%.However, there is a lack of information in literature concerning the performance of epoxy-based nanocomposites with embedded ferric-oxide nanoparticles.Recently, Reis et al. 20 showed that reinforcement of polymer mortars with spherical Fe 2 O 3 nanoparticles in the size range from 20 to 60 nm led to an increase of their elastic and fracture properties.Although the shape and concentration range of the nanofiller differed in the case of published data and this study, there is a reasonable good agreement in observed effects.CONCLUSIONS Epoxy-resin-based nanocomposites were prepared by embedding fairly uniform ferric oxide nanorods in order to understand better the influence of 1D nanofillers on the mechanical properties of polymer matrices.The tensile properties (stress at break, strain at break, Young's elastic modulus and tensile toughness) and the glass transition temperature of the epoxy resins were significantly altered in the presence of α-Fe 2 O 3 NRs at the low concentration levels (≤1.0 wt.%).The tendency of inorganic nanoparticles with rod-like morphology to form bundle-like aggregates was enhanced in the surrounding media of different polarity.This is the main obstacle that limits reinforcement of polymer matrices with 1D nanofillers.For this reason, future work will be focused on the improvement of the compatibility of epoxy resin and α-Fe 2 O 3 NRs.
Acknowledgements.Financial support for this study was granted by the Ministry of Education, Science and Technological Development of the Republic of Serbia (Project No. 45020).The AFM analysis was realized within the project 13-06700S (Czech Science Foundation).

Fig. 1 .
Fig. 1.A typical TEM image of the α-Fe 2 O 3 NRs (A) and the corresponding size distributions of the diameter and length (B and C, respectively).

Fig. 3 .
Fig. 3. DSC curves of pure epoxy resin and of the α-Fe 2 O 3 /epoxy nanocomposites.Typical stress-strain curves of pure epoxy resin and the nanocomposites with different α-Fe 2 O 3 NRs loadings are shown in Fig. 4. Data for the stress and