Preparation of graphene and inclusion in composites with poly(styrene-b-butadiene-b-styrene)

2.1 Materials

Graftech (GT) 220-50N expandable graphite (OH, USA) and Cheap Tubes (Vermont, USA)/HDPlas (CT) >700 m²/g nano-platelets (grade 4), 1–2 μm diameter, <3 nm, 1–3 layers (two different materials denoted #1 and #2). Poly(styrene-b-butadiene-b-styrene) (SBS) was from Aldrich (New South Wales, Australia), CAS 9003-55-8. p-xylene was from Merck (Victoria, Australia). 10 g graphite 5 g KClO₃ (2:1) [11–14]. Sodium hydroxide (NaOH) from Chem Supply Pty Ltd. (South Australia, Australia). Iron(II) chloride tetrahydrate (FeCl₂·4H₂O) and iron(III) chloride hexahydrate (FeCl₃·6H₂O) were obtained from BDH Chemicals (Queensland, Australia).

2.2 Preparation of nanocomposites

GT (in 1 g amounts) was expanded for 30 s in a furnace preheated to 1000°C to create expanded graphite oxide (EGO). The EGO was heated in a ceramic tube furnace at 1000°C for 8 h in an inert atmosphere of 100% nitrogen. A second EGO sample was heated at 1000°C for 8 h in a reducing atmosphere of 5% hydrogen and 95% argon. The production method, surface coating, and supplier of graphene are summarized in Table 1.

2.3 Preparation of magnetic-grafted graphene

FeCl₂·4H₂O and FeCl₃·6H₂O (molar ratio 1:2) were dissolved into 100 ml deoxygenated water (with constant stirring for 30 min). EGO was added to the Fe²⁺/Fe³⁺ solution and dispersed with ultrasonication for 30 min under N₂. The solution was stirred at room temperature for a further 30 min. While stirring, NaOH was added dropwise to precipitate the magnetite particles onto the graphene. The black precipitate was magnetically isolated and solution decanted. The magnetite-coated graphene was washed repeatedly with water and dried in a vacuum oven at 35°C for 24 h. SBS (1 g) was dissolved in xylene (10 ml) by standing overnight at 23°C. Graphenes 1% (w/w) were ultrasonicated in xylene (∼2 ml) for 10–20 min. Ultrasonication was used to further increase surface area of the graphenes by layer separation. SBS and graphene solutions were combined and ultrasonicated (to disperse graphene) as shown in Figure 1A and B. Ultrasonication was carried out with a Sonics Autotune series high-intensity Ultrasonic Processor Model GEX 500 (power 500 W) with a frequency of 20 kHz. Amplitude used was 25% for 20 min. Composites were precipitated with methanol, dried, and consolidated in a heated press at 155°C and 6 t pressure. A small round mold of ∼4.2 mm diameter and 0.5 mm thickness was used with polytetrafluoroethylene (PTFE) sheets on both sides. The sheets were sandwiched between metal plates. The mold was ∼30% overfilled to force out bubbles and holes.

Figure 1

(A) SBS dissolved in xylene with graphene in dispersion using π-π interactions and (B) SBS-graphene π-π interaction that kept graphene dispersed evenly.

2.4 Characterization

Electron microscopies were used to characterize surface morphology. For the Philips XL30 scanning electron microscope (SEM), graphene was mounted on conductive carbon tape and high vacuum mode was used (∼1.2×10^-5 mbar). The Quanta 200 SEM was used in low vacuum mode with the SBS mounted upright on right angled aluminum mounts on conductive carbon tape. The SBS was fractured while held in liquid nitrogen. For the JEOL 1010 transmission electron microscope (TEM), 4 mg graphene was suspended in 3 ml N-methyl-2-pyrolidone (NMP) and further dispersed with 10 min ultrasonication. A drop of suspended graphene was placed on GYCU200 Holey support film (200 mesh copper grids). The solvent was left to evaporate for 1 h before imaging. Raman spectroscopy was used to measure inelastic scattering in the graphene. A Perkin-Elmer Raman Station 400F 785 nm 250 mW and spot size 100 μm was used. One hundred twenty scans of 1 s were carried out. The graphenes were compressed in a press (9 t for 5 min) to obtain a stronger response. Wide-angle X-ray scattering (WAXS) was employed to determine any change in crystalline structure of the graphene. The instrument used was a Bruker D8 Advance diffractometer X-ray diffraction (XRD; Cu Kα radiation with λ=0.154 nm). The films were placed on a sample holder and analyzed using a 1 μm slit. The diffractograms were scanned in the 2θ range from 10° to 90° at a rate of 2°/min. Thermogravimetry was employed to determine thermal stability of SBS using a Perkin-Elmer TGA-7. A ∼2 mg mass of the SBS composite was analyzed in an open platinum pan. The sample was heated from 30°C to 850°C at 20°C/min in nitrogen at 20 ml/min. At 700°C, the gas was switched back to air at 20 ml/min. Mechanical force thermomechanometry (mf-TM) was used to measure viscoelastic properties with frequency and temperature. Measurement was carried out with a Perkin-Elmer Diamond Dynamic Mechanical Analysis (DMA). A standard target position of 10 mm, temperature range of -120°C to 120°C, and frequencies of 0.5, 1, 2, 5, and 10 Hz (only 1 Hz is reported) were used. Electrical conductivity was measured with a HP 4192A impedance analyzer. A clamp with two 1.2 cm circular electrodes was used to measure electrical capacitance.

3.1 Microscopy

Graphene was produced by rapid thermal expansion and compared with a high-quality commercial graphene (CT). The graphene was dispersed in SBS using π-π interactions (Figure 1) to prevent agglomeration. Figure 2A shows the SEM image taken at low resolution (80×) to demonstrate the worm/accordion-like expansion that is typical of thermally expanded graphene. A TEM image that demonstrates the effect of ultrasonication on the sample (increased exfoliation and decreased width) is shown in Figure 2B. Two SEM images were taken at 20,000× magnification allowing easy comparison. Figure 2C shows long folds of carbon joined at the edges in an accordion-like fashion: expanded graphite looks like this prior to ultrasonication or compression. Figure 2D shows that a commercial graphene (CT) of one to three layers is comparable in thickness but has much smaller particle size. SBS was compared with and without graphene at 1000× magnification; neat SBS (Figure 2E) shows regular straight ridges with small fracture marks. SBS with GT 1000°C 30 s graphene (Figure 2F) shows irregular fracture marks and straight breaks suggesting reinforcement. SBS with H₂ reduced graphene (Figure 2G) shows irregular fracture marks and straight breaks, which are less pronounced suggesting finer reinforcing.

Figure 2

SEM image of (A) GT N₂ 500 μm 80× showing worm/accordion-like expansion 20,000× (2 μm), (B) TEM image of GT 1000°C 30 s with ultrasonication, and SEM images of (C) GT N₂ (inert gas to remove oxides), (D) CT#1 >700 m²/g and 1000×, (E) SBS neat, (f) SBS-1000°C 30 s, and (g) SBS-H₂ reduced graphene.

3.2 Vibrational spectroscopy

Raman spectroscopy is the most convenient way of characterizing graphenes. Raman spectroscopy records vibrations of covalent bonds in molecules. It is sensitive to changes in polarizability (stretching and deformation) rather than dipole moment as with infrared spectroscopy. It is most sensitive to symmetrical bonds and slight changes in bond angle or strength. Raman spectroscopy is particularly sensitive to the phenomenon that breaks the symmetry of sp² carbons [15]. Raman spectra (Figure 3) show significant peaks at ∼2654, ∼1582, and ∼1316 cm^-1 (2D, G, and D bands).

Figure 3

(A) Raman spectra showing highest to lowest at the D peak (smoothed using 15-point moving average), (B) comparing D peak intensity of graphene produced using four different methods (intensities compared using unsmoothed data), and (C) graphene edges zigzag and armchair.

3.2.2 D peak

The D peak occurs when there are defects in the graphene matrix. Defects can include edges, grain boundaries, vacancies, implanted atoms, and a change from sp² to sp³ carbon bonding [17]. Defects that do not generate a D peak include perfect zigzag edges, charged impurities, intercalants, and uniaxial/biaxial strain [16]. Armchair edges are a major source of defects [15]. Defects are often introduced as a way of functionalizing a graphene. Using π-π interactions for dispersion avoids the introduction of new defects. Defects will adversely affect performance and will limit the uses of graphene. Defects are of particular relevance to applications that require pristine graphene. Any method of decreasing defects in graphenes is noteworthy. The CT >700 m²/g graphene showed a peak in the D band at 1316 cm^-1, whereas the 220-50N-based graphenes showed smaller D peaks (Figure 3). The order of these smaller D peaks corresponds to the amount of processing carried out on the 220-50N: (1) air (highest), (2) N₂ inert gas, and (3) H₂ reducing gas (lowest). However, these data cannot be compared quantitatively. The D/G ratio is used instead (Figure 4).

Figure 4

Raman spectra comparing the D/G ratio of graphene produced using four different methods (ratios compared using unsmoothed data).

3.2.3 G peak

The G band has been shown to correlate with the number of layers in graphene [3]. Single layers are indicated by an increase in Raman shift. Double or multiple layers (graphite) are indicated by a decrease in Raman shift. The CT graphene showed a G peak at 1580 cm^-1 (Figure 5). The GT-based graphenes showed peaks between 1580 and 1582 cm^-1. If the CT graphene is single layer (as determined by the 2D peak), then the GT graphenes also contain single layers (as indicated by the G peak).

Figure 5

Raman spectra comparing the G peak position of graphene produced using four different methods.

3.3 Mechanical properties

Graphene at 1% (w/w) in SBS was found to increase the tensile storage modulus by 68% (Figure 6A) and reduce the damping of SBS elastomer. It is seen in Figure 6B that energy absorption (loss modulus) at the lower temperatures is increased by 147%. Pure SBS has the lowest absorption and GT in 5% H₂ (a reducing atmosphere) had the highest energy absorption. The improvement plateaued at approximately -100°C at just under 2.5 GPa (compared with ∼1.9 GPa for SBS). CT#2 graphene enhanced the SBS damping (tan(δ)) at 25°C 74% (SBS v’s CT#2 Figure 6C). Styrene had the greatest effect at higher temperatures (strength and rigidity). The styrene π-π bonding with graphene enhanced the damping dramatically when using any graphene. The GT-Fe₃O₄ performed poorly because it had a graphene mass fraction of only 15% (established using TGA). However, the π-π bonding between the CT graphenes and styrene appeared to be stronger than that of the 220-50N based graphenes, suggesting that smaller size (CT) might be an important performance feature. Graphene enhanced styrene phase physical crosslinks increasing the modulus at low concentration. Graphene dispersion gave strong interactions with the polymer to stabilize the dispersion against agglomeration.

FIgure 6

DMA of graphene-SBS composites: (A) storage modulus, (B) loss modulus, and (C) tan(δ).

3.4 WAXS

SBS-graphene composites display a series of well-resolved diffraction peaks superimposed on an amorphous background (Figure 7A). The peak of maximum intensity at 2θ∼26.56° is attributed to graphene [20] that corresponds to the 002 reflection (plane in crystal) of graphene [21] and is the characteristic peak of pure graphite [20]. Pure SBS has an amorphous halo at 2θ∼19.60° [22]. The diffraction patterns were similar for all SBS-graphene composites differing only in intensity. It was reported that an Fe₃O₄ composite can be differentiated from graphene by differing diffraction angles [21]. The tallest peak within the area of maximum intensity was the CT graphene, whereas the pure SBS had no peak.

Figure 7

(A) Diffractogram: relative intensity versus scattering angle of SBS-G/graphene composites (listed in order of peak intensity) and (B) graphenes (in SBS) versus peak (002) intensity.

Each reflection is from a particular diffraction plane in the crystal. Therefore, the sample with the greatest intensity for the 002 plane displays a preference for this crystal orientation (Figure 7B). The order of this preference is from highest to lowest: (1) CT graphene, (2) GT 1000°C 30 s (least processed), (3) GT N₂ (an inert gas), (4) GT H₂ (a reducing gas), and (5) GT Fe₃O₄. As single-layer graphene will have no layer stacking, the reducing intensity of the (002) peak implies a move toward single-layer graphene.

The Bragg law equation d=nλ/2sinθ was used to calculate the distance between graphene layers (Figure 8). The spacing of the graphenes at the 002 crystal orientation suggests that the CT graphene has the smallest distance between layers. The CT graphene shows a greater variation than any other graphene (although the variation is small ∼0.5%). The greatest distance between layers is for the composite with the minimally treated 220-50N (1000°C 30 s) and the graphene treated in the inert gas (N₂). Between the two ranges were the graphene treated in the reducing gas (H₂) and the graphene with Fe₃O₄ on its surface.

Figure 8

Graphenes (in SBS) versus distance between graphene layers.

3.5 Thermal stability

TGA shows that degradation for all the SBS-graphene composites starts at ∼350°C and is completed by ∼475°C (Figure 9). SBS with GT Fe₃O₄ and 1000°C 30 s decompose at the lowest temperature (466°C) and SBS with N₂ graphene is furthest right showing slightly more heat resistance (478°C). Pure SBS and SBS with H₂ reduced graphene and with CT graphene had an intermediate temperature of degradation (474°C). Often when multiple components are present (polystyrene, polybutadiene, and graphene), a TGA curve will show three deflections (this one does not). Thus, adding graphene to SBS can slightly decrease the heat resistance of SBS (Fe₃O₄ or 1000°C 30 s) or slightly enhance it (N₂). However, minor changes in degradation temperature are desirable, as SBS already has a fairly high temperature of degradation, that is, low heat resistance makes molding, disposal, and reprocessing easier.

Figure 9

TGA mass v’s temperature with first derivative of SBS and graphene (1%) composites created with GT and CT (listed in order of temperature of decomposition).

3.6 Conductivity

Composite capacitance varied between 215 and 267 pF/cm with pure SBS 217 pF/cm (virtually unchanged). Composite resistivity (30.5 MΩ cm) did not improve significantly with the addition of graphene at 1% (27.0–29.8 MΩ cm) (Figure 10). Graphene and SBS were compatible due to π-π interactions between their aromatic groups resulting in uniform dispersion. Styrene in SBS is a dispersed phase in a continuous butadiene phase, so graphene surrounded by styrene would not form a percolation network to enhance conductivity. High aspect ratio fillers such as graphene are known to decrease percolation thresholds, and in a polyethylene-graphite composites, it was estimated that the interparticle spacing had to be ∼1.2 nm before conductivity was achieved [23].

Figure 10

Capacitance versus SBS-graphene composites (measured at 11 kHz AC).

3.7 Surface modification

Graphene sheets readily aggregate. So to retain the high surface area and other unique properties of graphenes, they were functionalized with Fe₃O₄. The randomly distributed magnetic nanoparticles on the surface and edges of graphene make both surfaces of graphene readily accessible and facilitate a homogenous dispersion in a polymer. Fe₃O₄ nanoparticles also aid in the exfoliation of graphene.

DMA showed that SBS with GT-Fe₃O₄ (1%) storage modulus was higher than SBS but lower than other types of graphenes in SBS. Loss modulus showed that SBS with GT-Fe₃O₄ was only superior to SBS with CT#1 and plain SBS. tan(δ) showed that GT-Fe₃O₄ was superior to GT treated in an inert environment and SBS. WAXS showed that SBS with GT-Fe₃O₄ had a peak intensity (653) at 26.56°, which was the lowest of all graphene samples (suggesting more single-layer graphene). Interlayer spacing at 26.56° was not affected (0.3354 nm) and was the same as other GT graphenes. TGA showed that SBS with GT-Fe₃O₄ had the highest temperature of decomposition (a slight increase in temperature resistance). Conductivity testing showed that capacitance of SBS with GT-Fe₃O₄ was 219 pF/cm, which was higher than SBS but lower than all other GT and CT graphenes (except GT-N₂, which was lower than SBS). Thus, overall measured SBS properties were less enhanced by GT-Fe₃O₄ graphene than other graphenes. Only 15% of the mass fraction of GT-Fe₃O₄ was graphene (established by TGA), which explains why its performance was lower than expected. Other properties of interest include magnetism, which will be subsequently reported.