- Email: [email protected]
Morphology and physical properties of poly(ethylene oxide) loaded graphene nanocomposites prepared by two different techniques
European Polymer Journal 47 (2011) 1534–1540
Contents lists available at ScienceDirect
European Polymer Journal journal homepage: www.elsevier.com/locate/europolj
Macromolecular Nanotechnology
Morphology and physical properties of poly(ethylene oxide) loaded graphene nanocomposites prepared by two different techniques Waleed E. Mahmoud ⇑ King Abdulaziz University, Faculty of Science, Physics Department, Jeddah, Saudi Arabia
MACROMOLECULAR NANOTECHNOLOGY
a r t i c l e
i n f o
Article history: Received 28 December 2010 Received in revised form 26 April 2011 Accepted 7 May 2011 Available online 14 May 2011 Keywords: Nanocomposites Graphene Optical microscopy Mechanical Electrical Casting
a b s t r a c t Organic–inorganic hybrids are artificially created structures presenting novel properties not exhibited by either of the component materials alone. In this contribution one addresses processing, morphology and properties of polymer nanocomposites reinforced graphene. First, synthesis routes to graphite oxide (GO) and foliated graphene sheets (FGS) are illustrated. Physical characterization of these graphene sheets were conducted using atomic force microscopy and X-ray diffraction techniques. Processing, structure and properties of graphene/poly(ethylene oxide) (PEO) nanocomposites are discussed. FGS was dispersed into PEO via two different composite manufacturing techniques: melt compounding and solvent mixing. Morphology of dispersed graphene and properties from different blending routes are compared. TEM showed that graphene distributed parallel to the composite surface using solvent method, while distributed randomly in melt blended method. Optical measurements indicated that the transparency of PEO/graphene prepared by solvent method is higher than that of melt blended method in the visible region. Electrical conductivity measurements are employed to evaluate threshold concentration for rigidity and connectivity percolation. The percolation concentration of the composites prepared by solvent method is less than those of melt blended method. The mechanical performance of the composites prepared by solvent method is higher than melt blended. Halpin–Tsai model has been used to confirm the distribution of the graphene into PEO by the two different processing techniques. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Plastics have many outstanding properties: light weight, toughness, good elongation, easy processing and low cost. However, comparing with ceramics and metals, low stiffness, strength, flammability and high permeability to gases and solvents can be their weaknesses. In some applications, higher thermal and electrical conductivity could be advantageous. Reinforcement with nanometer sized fillers can overcome many of these drawbacks if they are well dispersed in the matrix polymer. Most property enhancements can be achieved at significantly smaller ⇑ Permanent address: Suez Canal University, Faculty of Science, Physics Department, Ismailia, Egypt. Tel.: +20 0162 277 227. E-mail address: [email protected] 0014-3057/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.eurpolymj.2011.05.011
loadings than conventional micron sized glass or carbon fibers and resulting ‘‘nanocomposites’’ are much lighter in weight. Graphene, as an emerging two-dimensional (2D) structure of free-standing carbon atoms packed into a dense honeycomb crystal structure, is being predicted to have numerous potential applications [1,2] because of its unusual electron transport properties and other distinctive characteristics. Various methods have been developed to produce graphene, including mechanical [3], physical [4], and chemical methods [5]. Chemical production of graphene is facile and low-cost and is one of the favored methods for producing graphene. Currently, three major carbon sources are used as starting materials for chemical production of graphene: graphite oxide (GO) [6], expanded graphite (EG) [7], and sieved graphite powder (SGP) [8].
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
1535
Although chemical reduction of GO can produce graphene [9], the product contains significant numbers of oxygen containing functional groups and irreversible lattice defects. Graphene sheets exfoliated from EG and SGP have fewer defects and are more conductive than chemically reduced graphene oxide [10]. The superior properties of graphene compared to polymers are also reflected in polymer/graphene nanocomposites. Polymer/graphene nanocomposites show superior mechanical, thermal, gas barrier, electrical and flame retardant properties compared to the neat polymer [11–13]. It was also reported that the improvement in mechanical and electrical properties of graphene based polymer nanocomposites are much better in comparison to that of clay or other carbon filler-based polymer nanocomposites [14–16]. Although CNTs show comparable mechanical properties compared to graphene, still graphene is better nanofiller than CNT in certain aspects, such as thermal and electrical conductivity [16–18]. However, the improvement in the physicochemical properties of the nanocomposites depends on the distribution of graphene layers in the polymer matrix as well as interfacial bonding between the graphene layers and polymer matrix. Interfacial bonding between graphene and the host polymer dictates the final properties of the graphene reinforced polymer nanocomposite. In this contribution, one presents a comparison study between two different processing techniques for investigating the dispersion of graphene into polymer matrix and their effect on the mechanical, electrical and optical properties of polymer.
for 15 min and washed four times with a solution of 10% HCl and five times with water. The obtained graphite oxide was rapidly heated to 1000 °C and maintained for 60 s under an atmosphere of forming gas (5% H2 and 95% Ar), to get exfoliated graphite oxide nanosheets. The resulting exfoliated graphite oxide nanosheets (1 mg) was added to 20 mL acetonitrile (reducing agent), and the resulting mixture was transferred to a Teflon-lined autoclave (25 mL) and maintained at 180 °C for 12 h, to get graphene layers. The reaction products were sonicated for 60 min and centrifuged at 600 rpm for 90 min, in order to remove un-exfoliated layers. The overall mechanism of reaction steps is shown in Fig. 1.
2. Experimental
2.2.2. Melt blended method Required amount of graphene added to PEO powder and mixed using a Brabender mixer with an initial screw speed of 100 rpm/min for 5 min; then the screw speed was raised to 200 rpm/min within 10 min. This compound was compression molding under a heating press (KARL KOLB, Germany) at temperature T = 200 ± 2 °C under a pressure of P = 4 MPa for 30 min.
Graphite oxide was synthesized from natural graphite using Hummer’s method [19,20]. Typically, 1 g of graphite, 0.5 g of NaNO3 and 23 ml of H2SO4 were mixed in an ice bath. Subsequently, 3 g of KMnO4 was gradually added. The mixture was allowed to react at 35 °C ± 5 for 30 min, and to result in a thick paste. Then, 46 ml of water was added and the mixture was refluxed for 15 min. The reaction was terminated by addition of 170 ml of aqueous solution of H2O2 (0.05 wt.%), resulting in a yellow brown mixture. Then the mixture was centrifuged at 1000 rpm
2.2.1. Solvent method Graphene (0.5 mg) was dispersed in distilled water (3 mL) in an ultrasonic bath for 30 min at room temperature to yield a homogeneous dispersion. Poly(ethylene oxide) (PEO) (1 g) was dissolved in distilled water (7 mL) at 80 °C until clear viscous solution has been formed. This solution was left to be cooled until reaches 40 °C, the graphene aqueous dispersion was gradually added to the PEO solution and sonicated for 30 min at room temperature. Finally, this homogeneous PEO/graphene solution was poured into a Teflon Petri dish and kept at 60 °C for 6 h for film formation. A series of PEO/graphene nanocomposite films with different graphene loadings (0.1–2 vol.%) were prepared by the same procedure.
2.3. Characterization and measurements 2.3.1. Characterization Typical tapping-mode atomic-force microscopy (AFM) measurements were performed using Multimode SPM
Fig. 1. The mechanism of preparation of graphene sheets from graphite stack.
MACROMOLECULAR NANOTECHNOLOGY
2.1. Synthesis of graphene
2.2. Synthesis of PEO/graphene composite films
1536
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
MACROMOLECULAR NANOTECHNOLOGY
from Digital Instruments with a Nanoscope IIIa Controller. Samples for AFM images were prepared by depositing a dispersed graphene/H2O solution (0.1 mg mL1) onto a freshly cleaved mica surface and allowing them to dry in air. Transmission electron microscopy (TEM) was performed using a JEOL 2010 electron microscope at an acceleration voltage of 200 kV. The graphene/H2O solution was dropped onto carbon-coated copper grids (mesh size 300) and allowed to dry under ambient conditions. The X-ray measurements were performed using a Philips X’pert diffractometer supplied with copper X-ray tube (kka1 = 1.5406 Å), nickel filter, graphite crystal monochromator, proportional counter detector, divergence slit 1° and 0.1 mm receiving slit. The working conditions were 40 kV and 30 mA for the X ray tube, scan speed 0.05° and 2 s measuring time per step. For each measurement, a complete scan was made between 5° and 40° (2h). To calibrate the measured Bragg 2h-angles, a standard reference material (SRM 640a) of pure Si powder was used. 2.3.2. Measurements Transmission (T) spectra of nanocomposite films were measured at normal incidence in the spectral ranging from 200 to 800 nm using a double-beam spectrophotometer (JASCO model V-570 UV–vis–NIR). The stress–strain behavior was measured at room temperature by using a material tester (AMETEK, USA), which connected by a digital force gage (Hunter Spring ACCU Force II, 0.01 N resolutions, USA) to measure stress forces. The force gage interfaced with computer to record the obtained data. The stress–strain behavior was measured at strain rate 0.01 mm/s. The samples were in form of strips (length 2 cm, width 2 mm, and thickness 1.2 mm). For electrical measurements, digital electrometer (616 Keithly, USA) was used. The samples were in the form of disc with cross-sectional area 1 104 m2 and 1.2 mm height. 3. Results and discussion 3.1. Characterization of graphene and PEO/graphene composites Representative contact mode height profile of FGS lying on mica is shown in Fig. 1a. Typical lateral size of FGS determined from AFM is 650–700 nm. Considering the diameter of the starting graphite material (100 lm), significant size reduction during ultrasonic is evident. Fig. 1b represents height scan of graphene sheets along the straight line drawn in Fig. 1a. Height differences of FGS plates are in multiples of 1.43 nm. This means that the graphene consists of few-layers. HRTEM is also used to characterize the degree of exfoliation of graphite. Fig. 2c presents typical TEM images of FGS, from which the thickness as well as the number of layers can be estimated. From the typical image of a FGS, approximately four to five layers of FGS were identified. The interlayer spacing was measured to be 0.34 nm, which is consistent with the theoretical value of 0.335 nm. X-ray diffraction (XRD) is an important tool for determining whether graphene-based sheets are indeed present
Fig. 2. Contact mode AFM scan of (a) FGS on mica and height profiles of (b) FGS along the straight white lines and (c) HRTEM of FGS.
as individual graphene sheets in the nanocomposites. Fig. 3 shows the XRD patterns of GO, graphene (FGS) and pristine graphite. A sharp reflection at 2h = 26.4° in the X-ray scattering pattern of pristine graphite originates from the inter-layer (0 0 2) spacing (d = 0.34 nm). The typical diffraction peak of GO was observed at about 2h = 10.9° with spacing (d = 0.91 nm). However, after thermal reduction treatment of GO, the diffraction peak of GO disappeared. The XRD results clearly demonstrate that GO was fully exfoliated into individual graphene sheets (FGS) and that the regular and periodic structure of graphene disappeared [21–23]. Transmission electron microscopy (TEM) was also employed to investigate the influence of mixing techniques on the foliated graphene sheets in the polymer matrix. The TEM image for PEO loaded 0.3 vol.% of graphene prepared by solvent method is shown in Fig. 4a. The graphene sheets shown at a low magnification in Fig. 4a appeared
θ
Fig. 3. XRD intensity profiles of (a) Graphite, (b) GO and (c) FGS.
1537
uniform and the graphene sheets looked more like a lump with an irregular shape than a flat sheet, and exhibited rough surface and rounded edges, which may be caused by the evaporation of water upon drying/heating. Further examination using a high resolution image (inset of Fig. 4b) gave a deeper insight into understanding of the exfoliation and morphology of graphene layers. The sheets are found to consist of four graphene layers and the thickness of graphene sheets in this case was found 1.4 nm, like the FGS seen in (Fig. 2c). According to TEM and HRTEM measurements, one can propose the following the mechanism. When the FGS incorporated into polymer under sonication in the solvent method, the polymer solution tends to diffuse between graphene sheets. This results in a further exfoliation of the graphene sheets to bilayers and in turn a reduction of graphene thickness. This made the graphene bilayers tends to align parallel to the polymer surface when the solution cast into petri dish, which results in transparent nanocomposites. This mechanism is shown in Fig. 5a. For melt blend method, the FGS added to polymer powder in the mixer and then this mixture compression molded under heating press. In this case, the polymer tends to intercalated with FGS instead of to diffuse between graphene layers. So, the thickness of FGS did not change and tends to wrinkle and randomly distributed into polymer matrix. This mechanism is shown in Fig. 5b. 3.2. UV–vis spectra of PEO/graphene nanocomposite films The UV–vis spectra of PEO and PEO loaded 0.3 vol.% graphene nanocomposites films prepared by solvent and melt blended techniques were recorded in transmittance mode as shown in Fig. 6. It is clear that the prepared PEO/graphene nanocomposite film by solvent method is highly transparent (87.6%) in the visible range. Comparing with the neat PEO (95%), the loss of transparency did not exceed 9%. For PEO/graphene nanocomposite film prepared by melt blended method, the transparency decreased to 66%, with loss around 30%. This drop in transparency is most likely to be due to the scattering of the light by the large aggregates formed in the nanocomposite film. This finding agrees with the TEM observation which indicated the random dispersion of graphene into polymer matrix by melt blended method. 3.3. Electrical conductivity
Fig. 4. TEM micrographs of (a) solvent blended and (b) melt blended PEO/ graphene composite and HRTEM inset of each image.
flat and transparent, with thickness 0.7 nm. Further examination using a high resolution image (inset of Fig. 4a) gave a deeper insight into understanding of the exfoliation and morphology of graphene layers. The sheets are found to consist of two graphene layers, unlike the FGS seen in (Fig. 2c). In contrary, TEM image for PEO loaded 0.3 vol.% of graphene prepared by melt method (Fig. 4b) showed that the distribution of graphene in the PEO matrix is not
Fig. 7 shows the variation of electrical conductivity of the prepared composites with filler volume fraction. A typical S-shaped curve is observed that separates three regions: insulating, transition, and conductive. It is clear that, the electrical conductivity of PEO is dramatically influenced by graphene addition. Both graphite and FGS effectively reduce the surface resistance, but they differ greatly in the onset concentration for electrical percolation. The model that is most often used to quantify the changes in the transition and conductive regions is the so called statistical percolation model [24]. Proposed by Kirkpatrick [25] and Zallen [26], this model predicts the electrical conductivity of an insulator–conductor binary
MACROMOLECULAR NANOTECHNOLOGY
Intensity (a.u.)
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
Fig. 5. The mechanism of FGS incorporation into polymer for (a) solvent technique and (b) melt blend technique.
mixture. The result is a power-law variation of the conductivity (r):
r ¼ ro
λ Fig. 6. Transmittance spectroscopy of (a) neat PEO, (b) solvent blended and (c) melt blended PEO/0.3 vol.% graphene composite.
σ
MACROMOLECULAR NANOTECHNOLOGY
1538
u uc 1 uc
t ð1Þ
where ro is the conductivity of unfilled polymer, u the volume fraction of filler, uc percolation threshold and t is a universal exponent (t P 2 for a randomly 3D dispersion of filler and t < 2 for 2D dispersion) [27]. According to this model the percolation concentrations were 1.8 vol.% for pristine graphite versus 0.5 vol.% for FGS melt blended and 0.3 vol.% for solvent blended. The exponent t equals 2.1 for pristine graphite, 3.6 for FGS melt blended, and 1.2 for solvent blended. Conductivity measurements indicate that the percolation threshold is inversely proportional to aspect ratio Af. Among FGS composites, conductivity of solvent blended samples is higher than melt blended ones at the same filler volume fraction. Resistance decrease takes place at even less than 0.3 vol.% of FGS in case of solvent mixed samples while it required more than 0.5 vol.% for melt intercalation. The values of exponent t indicate that both of pristine graphite and graphene melt blended are randomly distributed into PEO matrix, while graphene solvent blended is distributed parallel to the surface of PEO film. This trend agrees with the TEM observation which indicated the dispersion of graphene into polymer matrix. 3.4. Mechanical properties
Fig. 7. Conductivity vs. volume fraction for (a) melt blended PEO/ graphene, (b) solvent blended PEO/graphene and (c) pristine graphite/ PEO.
Typical stress–strain curves of PEO/graphene composites are shown in Fig. 8. The mechanical performance of the graphene/PEO nanocomposites (solvent blended) was significantly increased as compared to that of the melt blended method. However, in case of solvent blended (Fig. 8a), composite films all broke at higher elongation
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
1539
(a)
Fig. 9. Young modulus as a function of graphene volume fraction.
(b) Ejj ¼ Em
1 þ ML f uc 1 ML uc
ð3Þ
0 E g
Em ML ¼ @ Eg Em
0 E g
Em M T ¼ @ Eg Em
F¼ Fig. 8. Stress–strain behavior for (a) solvent blended and (b) melt blended PEO/graphene composite.
than melt blended (Fig. 8b). For example, with only 0.3 vol.% FGS, the tensile strength increased by 189% against 104% for melt blended, the elongation at break increased by 111% against a decrease by 70% for melt blended, the Young’s modulus increased by 200% against 120% for melt blended. This may be due to the aspect ratio and dispersion state of graphene by solvent blended method were similar in nanocomposites with different FGS loading, the enhanced mechanical properties are believed to arise from the strong interaction between graphene and the PEO matrix due to high aspect ratio of graphene [24], while less well dispersed morphology of graphene into PEO by melt blended may lead to smaller interfacial area and may reduce the adhesion. To confirm such assumption, Halpin–Tsai model [28–30] is selected for predicting the modulus of unidirectional (E||) as well as the modulus of randomly distributed (Er) filler-reinforced nanocomposites. The Young’s modulus of the nanocomposite for FGS randomly dispersed as 3D network throughout the polymer matrix is defined as
Er ¼ Em
3 1 þ ML f uc 5 1 þ 2M T uc þ 8 1 ML f uc 8 1 M T uc
1 1 A þf
ð4Þ
1 1 A þ2
ð5Þ
2Af 2lg ¼ 3 3t g
ð6Þ
Eg and Em are Young’s modulus of the FGS and the polymer matrix. Af, lg, and tg represent the aspect ratio, length and thickness of the graphene oxide sheet, and uc is the volume fraction of FGS in the nanocomposites. Fig. 9 depicts the young modulus as a function of graphene volume fraction. It is clear that, there is a good agreement between the experimental data obtained for graphene/PEO nanocomposites, solvent blended, and the calculated results under the hypothesis that FGS is aligned parallel to the surface of the nanocomposite film. This consistency indicates that external tensile loads were successfully transmitted to the FGS filler across the graphene–PEO interface via strong interfacial interactions, [15,18] and presumably the graphene sheets prefer to align parallel to the surface of the sample film within the nanocomposites [18]. In contrary, there is a good agreement between the experimental data obtained for graphene/PEO nanocomposites, melt blended, and the calculated results under the hypothesis that FGS is randomly dispersed as 3D network throughout the polymer matrix. This trend agrees with the TEM observation which indicated the dispersion of graphene into polymer matrix. 4. Conclusion
ð2Þ
The Young’s modulus of the nanocomposite for FGS aligned parallel to the surface of the sample film is defined as
Graphene sheets were prepared by chemical oxidation of pristine graphite flakes followed by thermal exfoliation and reduction of graphite oxide. Typical lateral size of FGS determined from AFM is 650–700 nm and thickness
MACROMOLECULAR NANOTECHNOLOGY
where
MACROMOLECULAR NANOTECHNOLOGY
1540
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
1.43 nm. The incorporation of FGS into the polymer by solvent technique results in a reduction of FGS thickness to 0.7 nm, while the FGS thickness did not change when incorporated into polymer by melt blend technique. The XRD results clearly demonstrate that GO was fully exfoliated into individual graphene sheets. The UV–vis spectra of PEO loaded 0.3 vol.% graphene nanocomposites films prepared by solvent and melt blended techniques were recorded in transmittance mode. The results showed that the prepared PEO/graphene nanocomposite film by solvent method is highly transparent (87.6%) comparing with melt blended method (66%). The electrical measurements depicted that the percolation concentration is 0.5 vol.% for by melt blended method versus 0.3 vol.% for solvent blended. The mechanical performance of the graphene/ PEO nanocomposites (solvent blended) was significantly increased as compared to that of the melt blended method. Halpin–Tsai model indicated that the graphene sheets prefer to align parallel to the surface of the nanocomposites using solvent method. In contrary, there is a good agreement between the experimental data obtained for graphene/PEO nanocomposites, melt blended, and the calculated results under the hypothesis that FGS is randomly dispersed as 3D network throughout the polymer matrix. References [1] Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y, Dubonos SV, et al. Science 2004;306:666. [2] Lee C, Wei X, Kysar JW, Hone J. Science 2008;321:385. [3] Di CA, Wei DC, Yu G, Liu YQ, Guo YL, Zhu DB. Adv Mater 2008;20:3289. [4] Liu ZF, Liu Q, Huang Y, Ma YF, Yin SG, Zhang XY, et al. Adv Mater 2008;20:3924.
[5] Ang PK, Chen W, Wee ATS, Loh KP. J Am Chem Soc 2008;130:14392–3. [6] Liu Z, Robinson JT, Sun XM, Dai HJ. J Am Chem Soc 2008;130:10876. [7] Stankovich S, Dikin DA, Dommett GHB, Kohlhaas KM, Zimney EJ, Stach EA, et al. Nature 2006;442:282. [8] Berger C, Song ZM, Li XB, Wu XS, Brown N, Naud C, et al. Science 2006;312:1191. [9] Kim KS, Zhao Y, Jang H, Lee SY, Kim JM, Ahn JH, et al. Nature 2009;457:706. [10] Si Y, Samulski ET. Nano Lett 2008;8:1679. [11] Xu YX, Bai H, Lu GW, Li C, Shi GQ. J Am Chem Soc 2008;130:5856. [12] Li XL, Wang XR, Zhang L, Lee SW, Dai HJ. Science 2008;319:1229. [13] Li XL, Zhang GY, Bai XD, Sun XM, Wang XR, Wang E, et al. Nat Nanotechnol 2008;3:538. [14] Valles C, Drummond C, Saadaoui H, Furtado CA, He M, Roubeau O, et al. J Am Chem Soc 2008;130:15802. [15] Liu N, Luo F, Wu HX, Liu YH, Zhang C, Chen J. Adv Funct Mater 2008;18:1518. [16] Viculis LM, Mack JJ, Kaner RB. Science 2003;299:1361. [17] Fan XB, Peng WC, Li Y, Li XY, Wang SL, Zhang GL, et al. Adv Mater 2008;20:4490. [18] Tung VC, Allen MJ, Yang Y, Kaner RB. Nat Nanotechnol 2009;4:25. [19] Hummers WS, Offeman RE. J Am Chem Soc 1958;80:1339. [20] Park S, An JH, Piner RD, Jung I, Yang DX, Velamakanni A, et al. Chem Mat 2008;20:6592. [21] Zhu J. Nat Nanotechnol 2008;3:528. [22] Liu W, Hoa SV, Pugh M. Compos Sci Technol 2005;65:2364. [23] Zhu J, Kim J, Peng H, Margrave JL, Khabashesku VN, Barrera EV. Nano Lett 2003;3:1107. [24] Mahmoud WE, El-Mossalamy EH, Arafa HM. J Appl Poly Sci 2011;121:502. [25] Kirkpatrick S. Rev Mod Phys 1973;45:574. [26] Zallen R. The physics of amorphous solids. New York: Wiley; 1985 [Chapter 4]. [27] Derrida B, Stauffer D, Herrmann HJ, Vannemius J. J Phys Lett 1983;44:701. [28] Kalaitzidou K, Fukushima H, Miyagawa H, Drzal LT. Polym Eng Sci 2007;47:1796. [29] Schaefer DW, Justice RS. Macromolecules 2007;40:8501. [30] Halpin JC, Kardos JL. Polym Eng Sci 1976;16:344.
Contents lists available at ScienceDirect
European Polymer Journal journal homepage: www.elsevier.com/locate/europolj
Macromolecular Nanotechnology
Morphology and physical properties of poly(ethylene oxide) loaded graphene nanocomposites prepared by two different techniques Waleed E. Mahmoud ⇑ King Abdulaziz University, Faculty of Science, Physics Department, Jeddah, Saudi Arabia
MACROMOLECULAR NANOTECHNOLOGY
a r t i c l e
i n f o
Article history: Received 28 December 2010 Received in revised form 26 April 2011 Accepted 7 May 2011 Available online 14 May 2011 Keywords: Nanocomposites Graphene Optical microscopy Mechanical Electrical Casting
a b s t r a c t Organic–inorganic hybrids are artificially created structures presenting novel properties not exhibited by either of the component materials alone. In this contribution one addresses processing, morphology and properties of polymer nanocomposites reinforced graphene. First, synthesis routes to graphite oxide (GO) and foliated graphene sheets (FGS) are illustrated. Physical characterization of these graphene sheets were conducted using atomic force microscopy and X-ray diffraction techniques. Processing, structure and properties of graphene/poly(ethylene oxide) (PEO) nanocomposites are discussed. FGS was dispersed into PEO via two different composite manufacturing techniques: melt compounding and solvent mixing. Morphology of dispersed graphene and properties from different blending routes are compared. TEM showed that graphene distributed parallel to the composite surface using solvent method, while distributed randomly in melt blended method. Optical measurements indicated that the transparency of PEO/graphene prepared by solvent method is higher than that of melt blended method in the visible region. Electrical conductivity measurements are employed to evaluate threshold concentration for rigidity and connectivity percolation. The percolation concentration of the composites prepared by solvent method is less than those of melt blended method. The mechanical performance of the composites prepared by solvent method is higher than melt blended. Halpin–Tsai model has been used to confirm the distribution of the graphene into PEO by the two different processing techniques. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Plastics have many outstanding properties: light weight, toughness, good elongation, easy processing and low cost. However, comparing with ceramics and metals, low stiffness, strength, flammability and high permeability to gases and solvents can be their weaknesses. In some applications, higher thermal and electrical conductivity could be advantageous. Reinforcement with nanometer sized fillers can overcome many of these drawbacks if they are well dispersed in the matrix polymer. Most property enhancements can be achieved at significantly smaller ⇑ Permanent address: Suez Canal University, Faculty of Science, Physics Department, Ismailia, Egypt. Tel.: +20 0162 277 227. E-mail address: [email protected] 0014-3057/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.eurpolymj.2011.05.011
loadings than conventional micron sized glass or carbon fibers and resulting ‘‘nanocomposites’’ are much lighter in weight. Graphene, as an emerging two-dimensional (2D) structure of free-standing carbon atoms packed into a dense honeycomb crystal structure, is being predicted to have numerous potential applications [1,2] because of its unusual electron transport properties and other distinctive characteristics. Various methods have been developed to produce graphene, including mechanical [3], physical [4], and chemical methods [5]. Chemical production of graphene is facile and low-cost and is one of the favored methods for producing graphene. Currently, three major carbon sources are used as starting materials for chemical production of graphene: graphite oxide (GO) [6], expanded graphite (EG) [7], and sieved graphite powder (SGP) [8].
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
1535
Although chemical reduction of GO can produce graphene [9], the product contains significant numbers of oxygen containing functional groups and irreversible lattice defects. Graphene sheets exfoliated from EG and SGP have fewer defects and are more conductive than chemically reduced graphene oxide [10]. The superior properties of graphene compared to polymers are also reflected in polymer/graphene nanocomposites. Polymer/graphene nanocomposites show superior mechanical, thermal, gas barrier, electrical and flame retardant properties compared to the neat polymer [11–13]. It was also reported that the improvement in mechanical and electrical properties of graphene based polymer nanocomposites are much better in comparison to that of clay or other carbon filler-based polymer nanocomposites [14–16]. Although CNTs show comparable mechanical properties compared to graphene, still graphene is better nanofiller than CNT in certain aspects, such as thermal and electrical conductivity [16–18]. However, the improvement in the physicochemical properties of the nanocomposites depends on the distribution of graphene layers in the polymer matrix as well as interfacial bonding between the graphene layers and polymer matrix. Interfacial bonding between graphene and the host polymer dictates the final properties of the graphene reinforced polymer nanocomposite. In this contribution, one presents a comparison study between two different processing techniques for investigating the dispersion of graphene into polymer matrix and their effect on the mechanical, electrical and optical properties of polymer.
for 15 min and washed four times with a solution of 10% HCl and five times with water. The obtained graphite oxide was rapidly heated to 1000 °C and maintained for 60 s under an atmosphere of forming gas (5% H2 and 95% Ar), to get exfoliated graphite oxide nanosheets. The resulting exfoliated graphite oxide nanosheets (1 mg) was added to 20 mL acetonitrile (reducing agent), and the resulting mixture was transferred to a Teflon-lined autoclave (25 mL) and maintained at 180 °C for 12 h, to get graphene layers. The reaction products were sonicated for 60 min and centrifuged at 600 rpm for 90 min, in order to remove un-exfoliated layers. The overall mechanism of reaction steps is shown in Fig. 1.
2. Experimental
2.2.2. Melt blended method Required amount of graphene added to PEO powder and mixed using a Brabender mixer with an initial screw speed of 100 rpm/min for 5 min; then the screw speed was raised to 200 rpm/min within 10 min. This compound was compression molding under a heating press (KARL KOLB, Germany) at temperature T = 200 ± 2 °C under a pressure of P = 4 MPa for 30 min.
Graphite oxide was synthesized from natural graphite using Hummer’s method [19,20]. Typically, 1 g of graphite, 0.5 g of NaNO3 and 23 ml of H2SO4 were mixed in an ice bath. Subsequently, 3 g of KMnO4 was gradually added. The mixture was allowed to react at 35 °C ± 5 for 30 min, and to result in a thick paste. Then, 46 ml of water was added and the mixture was refluxed for 15 min. The reaction was terminated by addition of 170 ml of aqueous solution of H2O2 (0.05 wt.%), resulting in a yellow brown mixture. Then the mixture was centrifuged at 1000 rpm
2.2.1. Solvent method Graphene (0.5 mg) was dispersed in distilled water (3 mL) in an ultrasonic bath for 30 min at room temperature to yield a homogeneous dispersion. Poly(ethylene oxide) (PEO) (1 g) was dissolved in distilled water (7 mL) at 80 °C until clear viscous solution has been formed. This solution was left to be cooled until reaches 40 °C, the graphene aqueous dispersion was gradually added to the PEO solution and sonicated for 30 min at room temperature. Finally, this homogeneous PEO/graphene solution was poured into a Teflon Petri dish and kept at 60 °C for 6 h for film formation. A series of PEO/graphene nanocomposite films with different graphene loadings (0.1–2 vol.%) were prepared by the same procedure.
2.3. Characterization and measurements 2.3.1. Characterization Typical tapping-mode atomic-force microscopy (AFM) measurements were performed using Multimode SPM
Fig. 1. The mechanism of preparation of graphene sheets from graphite stack.
MACROMOLECULAR NANOTECHNOLOGY
2.1. Synthesis of graphene
2.2. Synthesis of PEO/graphene composite films
1536
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
MACROMOLECULAR NANOTECHNOLOGY
from Digital Instruments with a Nanoscope IIIa Controller. Samples for AFM images were prepared by depositing a dispersed graphene/H2O solution (0.1 mg mL1) onto a freshly cleaved mica surface and allowing them to dry in air. Transmission electron microscopy (TEM) was performed using a JEOL 2010 electron microscope at an acceleration voltage of 200 kV. The graphene/H2O solution was dropped onto carbon-coated copper grids (mesh size 300) and allowed to dry under ambient conditions. The X-ray measurements were performed using a Philips X’pert diffractometer supplied with copper X-ray tube (kka1 = 1.5406 Å), nickel filter, graphite crystal monochromator, proportional counter detector, divergence slit 1° and 0.1 mm receiving slit. The working conditions were 40 kV and 30 mA for the X ray tube, scan speed 0.05° and 2 s measuring time per step. For each measurement, a complete scan was made between 5° and 40° (2h). To calibrate the measured Bragg 2h-angles, a standard reference material (SRM 640a) of pure Si powder was used. 2.3.2. Measurements Transmission (T) spectra of nanocomposite films were measured at normal incidence in the spectral ranging from 200 to 800 nm using a double-beam spectrophotometer (JASCO model V-570 UV–vis–NIR). The stress–strain behavior was measured at room temperature by using a material tester (AMETEK, USA), which connected by a digital force gage (Hunter Spring ACCU Force II, 0.01 N resolutions, USA) to measure stress forces. The force gage interfaced with computer to record the obtained data. The stress–strain behavior was measured at strain rate 0.01 mm/s. The samples were in form of strips (length 2 cm, width 2 mm, and thickness 1.2 mm). For electrical measurements, digital electrometer (616 Keithly, USA) was used. The samples were in the form of disc with cross-sectional area 1 104 m2 and 1.2 mm height. 3. Results and discussion 3.1. Characterization of graphene and PEO/graphene composites Representative contact mode height profile of FGS lying on mica is shown in Fig. 1a. Typical lateral size of FGS determined from AFM is 650–700 nm. Considering the diameter of the starting graphite material (100 lm), significant size reduction during ultrasonic is evident. Fig. 1b represents height scan of graphene sheets along the straight line drawn in Fig. 1a. Height differences of FGS plates are in multiples of 1.43 nm. This means that the graphene consists of few-layers. HRTEM is also used to characterize the degree of exfoliation of graphite. Fig. 2c presents typical TEM images of FGS, from which the thickness as well as the number of layers can be estimated. From the typical image of a FGS, approximately four to five layers of FGS were identified. The interlayer spacing was measured to be 0.34 nm, which is consistent with the theoretical value of 0.335 nm. X-ray diffraction (XRD) is an important tool for determining whether graphene-based sheets are indeed present
Fig. 2. Contact mode AFM scan of (a) FGS on mica and height profiles of (b) FGS along the straight white lines and (c) HRTEM of FGS.
as individual graphene sheets in the nanocomposites. Fig. 3 shows the XRD patterns of GO, graphene (FGS) and pristine graphite. A sharp reflection at 2h = 26.4° in the X-ray scattering pattern of pristine graphite originates from the inter-layer (0 0 2) spacing (d = 0.34 nm). The typical diffraction peak of GO was observed at about 2h = 10.9° with spacing (d = 0.91 nm). However, after thermal reduction treatment of GO, the diffraction peak of GO disappeared. The XRD results clearly demonstrate that GO was fully exfoliated into individual graphene sheets (FGS) and that the regular and periodic structure of graphene disappeared [21–23]. Transmission electron microscopy (TEM) was also employed to investigate the influence of mixing techniques on the foliated graphene sheets in the polymer matrix. The TEM image for PEO loaded 0.3 vol.% of graphene prepared by solvent method is shown in Fig. 4a. The graphene sheets shown at a low magnification in Fig. 4a appeared
θ
Fig. 3. XRD intensity profiles of (a) Graphite, (b) GO and (c) FGS.
1537
uniform and the graphene sheets looked more like a lump with an irregular shape than a flat sheet, and exhibited rough surface and rounded edges, which may be caused by the evaporation of water upon drying/heating. Further examination using a high resolution image (inset of Fig. 4b) gave a deeper insight into understanding of the exfoliation and morphology of graphene layers. The sheets are found to consist of four graphene layers and the thickness of graphene sheets in this case was found 1.4 nm, like the FGS seen in (Fig. 2c). According to TEM and HRTEM measurements, one can propose the following the mechanism. When the FGS incorporated into polymer under sonication in the solvent method, the polymer solution tends to diffuse between graphene sheets. This results in a further exfoliation of the graphene sheets to bilayers and in turn a reduction of graphene thickness. This made the graphene bilayers tends to align parallel to the polymer surface when the solution cast into petri dish, which results in transparent nanocomposites. This mechanism is shown in Fig. 5a. For melt blend method, the FGS added to polymer powder in the mixer and then this mixture compression molded under heating press. In this case, the polymer tends to intercalated with FGS instead of to diffuse between graphene layers. So, the thickness of FGS did not change and tends to wrinkle and randomly distributed into polymer matrix. This mechanism is shown in Fig. 5b. 3.2. UV–vis spectra of PEO/graphene nanocomposite films The UV–vis spectra of PEO and PEO loaded 0.3 vol.% graphene nanocomposites films prepared by solvent and melt blended techniques were recorded in transmittance mode as shown in Fig. 6. It is clear that the prepared PEO/graphene nanocomposite film by solvent method is highly transparent (87.6%) in the visible range. Comparing with the neat PEO (95%), the loss of transparency did not exceed 9%. For PEO/graphene nanocomposite film prepared by melt blended method, the transparency decreased to 66%, with loss around 30%. This drop in transparency is most likely to be due to the scattering of the light by the large aggregates formed in the nanocomposite film. This finding agrees with the TEM observation which indicated the random dispersion of graphene into polymer matrix by melt blended method. 3.3. Electrical conductivity
Fig. 4. TEM micrographs of (a) solvent blended and (b) melt blended PEO/ graphene composite and HRTEM inset of each image.
flat and transparent, with thickness 0.7 nm. Further examination using a high resolution image (inset of Fig. 4a) gave a deeper insight into understanding of the exfoliation and morphology of graphene layers. The sheets are found to consist of two graphene layers, unlike the FGS seen in (Fig. 2c). In contrary, TEM image for PEO loaded 0.3 vol.% of graphene prepared by melt method (Fig. 4b) showed that the distribution of graphene in the PEO matrix is not
Fig. 7 shows the variation of electrical conductivity of the prepared composites with filler volume fraction. A typical S-shaped curve is observed that separates three regions: insulating, transition, and conductive. It is clear that, the electrical conductivity of PEO is dramatically influenced by graphene addition. Both graphite and FGS effectively reduce the surface resistance, but they differ greatly in the onset concentration for electrical percolation. The model that is most often used to quantify the changes in the transition and conductive regions is the so called statistical percolation model [24]. Proposed by Kirkpatrick [25] and Zallen [26], this model predicts the electrical conductivity of an insulator–conductor binary
MACROMOLECULAR NANOTECHNOLOGY
Intensity (a.u.)
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
Fig. 5. The mechanism of FGS incorporation into polymer for (a) solvent technique and (b) melt blend technique.
mixture. The result is a power-law variation of the conductivity (r):
r ¼ ro
λ Fig. 6. Transmittance spectroscopy of (a) neat PEO, (b) solvent blended and (c) melt blended PEO/0.3 vol.% graphene composite.
σ
MACROMOLECULAR NANOTECHNOLOGY
1538
u uc 1 uc
t ð1Þ
where ro is the conductivity of unfilled polymer, u the volume fraction of filler, uc percolation threshold and t is a universal exponent (t P 2 for a randomly 3D dispersion of filler and t < 2 for 2D dispersion) [27]. According to this model the percolation concentrations were 1.8 vol.% for pristine graphite versus 0.5 vol.% for FGS melt blended and 0.3 vol.% for solvent blended. The exponent t equals 2.1 for pristine graphite, 3.6 for FGS melt blended, and 1.2 for solvent blended. Conductivity measurements indicate that the percolation threshold is inversely proportional to aspect ratio Af. Among FGS composites, conductivity of solvent blended samples is higher than melt blended ones at the same filler volume fraction. Resistance decrease takes place at even less than 0.3 vol.% of FGS in case of solvent mixed samples while it required more than 0.5 vol.% for melt intercalation. The values of exponent t indicate that both of pristine graphite and graphene melt blended are randomly distributed into PEO matrix, while graphene solvent blended is distributed parallel to the surface of PEO film. This trend agrees with the TEM observation which indicated the dispersion of graphene into polymer matrix. 3.4. Mechanical properties
Fig. 7. Conductivity vs. volume fraction for (a) melt blended PEO/ graphene, (b) solvent blended PEO/graphene and (c) pristine graphite/ PEO.
Typical stress–strain curves of PEO/graphene composites are shown in Fig. 8. The mechanical performance of the graphene/PEO nanocomposites (solvent blended) was significantly increased as compared to that of the melt blended method. However, in case of solvent blended (Fig. 8a), composite films all broke at higher elongation
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
1539
(a)
Fig. 9. Young modulus as a function of graphene volume fraction.
(b) Ejj ¼ Em
1 þ ML f uc 1 ML uc
ð3Þ
0 E g
Em ML ¼ @ Eg Em
0 E g
Em M T ¼ @ Eg Em
F¼ Fig. 8. Stress–strain behavior for (a) solvent blended and (b) melt blended PEO/graphene composite.
than melt blended (Fig. 8b). For example, with only 0.3 vol.% FGS, the tensile strength increased by 189% against 104% for melt blended, the elongation at break increased by 111% against a decrease by 70% for melt blended, the Young’s modulus increased by 200% against 120% for melt blended. This may be due to the aspect ratio and dispersion state of graphene by solvent blended method were similar in nanocomposites with different FGS loading, the enhanced mechanical properties are believed to arise from the strong interaction between graphene and the PEO matrix due to high aspect ratio of graphene [24], while less well dispersed morphology of graphene into PEO by melt blended may lead to smaller interfacial area and may reduce the adhesion. To confirm such assumption, Halpin–Tsai model [28–30] is selected for predicting the modulus of unidirectional (E||) as well as the modulus of randomly distributed (Er) filler-reinforced nanocomposites. The Young’s modulus of the nanocomposite for FGS randomly dispersed as 3D network throughout the polymer matrix is defined as
Er ¼ Em
3 1 þ ML f uc 5 1 þ 2M T uc þ 8 1 ML f uc 8 1 M T uc
1 1 A þf
ð4Þ
1 1 A þ2
ð5Þ
2Af 2lg ¼ 3 3t g
ð6Þ
Eg and Em are Young’s modulus of the FGS and the polymer matrix. Af, lg, and tg represent the aspect ratio, length and thickness of the graphene oxide sheet, and uc is the volume fraction of FGS in the nanocomposites. Fig. 9 depicts the young modulus as a function of graphene volume fraction. It is clear that, there is a good agreement between the experimental data obtained for graphene/PEO nanocomposites, solvent blended, and the calculated results under the hypothesis that FGS is aligned parallel to the surface of the nanocomposite film. This consistency indicates that external tensile loads were successfully transmitted to the FGS filler across the graphene–PEO interface via strong interfacial interactions, [15,18] and presumably the graphene sheets prefer to align parallel to the surface of the sample film within the nanocomposites [18]. In contrary, there is a good agreement between the experimental data obtained for graphene/PEO nanocomposites, melt blended, and the calculated results under the hypothesis that FGS is randomly dispersed as 3D network throughout the polymer matrix. This trend agrees with the TEM observation which indicated the dispersion of graphene into polymer matrix. 4. Conclusion
ð2Þ
The Young’s modulus of the nanocomposite for FGS aligned parallel to the surface of the sample film is defined as
Graphene sheets were prepared by chemical oxidation of pristine graphite flakes followed by thermal exfoliation and reduction of graphite oxide. Typical lateral size of FGS determined from AFM is 650–700 nm and thickness
MACROMOLECULAR NANOTECHNOLOGY
where
MACROMOLECULAR NANOTECHNOLOGY
1540
W.E. Mahmoud / European Polymer Journal 47 (2011) 1534–1540
1.43 nm. The incorporation of FGS into the polymer by solvent technique results in a reduction of FGS thickness to 0.7 nm, while the FGS thickness did not change when incorporated into polymer by melt blend technique. The XRD results clearly demonstrate that GO was fully exfoliated into individual graphene sheets. The UV–vis spectra of PEO loaded 0.3 vol.% graphene nanocomposites films prepared by solvent and melt blended techniques were recorded in transmittance mode. The results showed that the prepared PEO/graphene nanocomposite film by solvent method is highly transparent (87.6%) comparing with melt blended method (66%). The electrical measurements depicted that the percolation concentration is 0.5 vol.% for by melt blended method versus 0.3 vol.% for solvent blended. The mechanical performance of the graphene/ PEO nanocomposites (solvent blended) was significantly increased as compared to that of the melt blended method. Halpin–Tsai model indicated that the graphene sheets prefer to align parallel to the surface of the nanocomposites using solvent method. In contrary, there is a good agreement between the experimental data obtained for graphene/PEO nanocomposites, melt blended, and the calculated results under the hypothesis that FGS is randomly dispersed as 3D network throughout the polymer matrix. References [1] Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y, Dubonos SV, et al. Science 2004;306:666. [2] Lee C, Wei X, Kysar JW, Hone J. Science 2008;321:385. [3] Di CA, Wei DC, Yu G, Liu YQ, Guo YL, Zhu DB. Adv Mater 2008;20:3289. [4] Liu ZF, Liu Q, Huang Y, Ma YF, Yin SG, Zhang XY, et al. Adv Mater 2008;20:3924.
[5] Ang PK, Chen W, Wee ATS, Loh KP. J Am Chem Soc 2008;130:14392–3. [6] Liu Z, Robinson JT, Sun XM, Dai HJ. J Am Chem Soc 2008;130:10876. [7] Stankovich S, Dikin DA, Dommett GHB, Kohlhaas KM, Zimney EJ, Stach EA, et al. Nature 2006;442:282. [8] Berger C, Song ZM, Li XB, Wu XS, Brown N, Naud C, et al. Science 2006;312:1191. [9] Kim KS, Zhao Y, Jang H, Lee SY, Kim JM, Ahn JH, et al. Nature 2009;457:706. [10] Si Y, Samulski ET. Nano Lett 2008;8:1679. [11] Xu YX, Bai H, Lu GW, Li C, Shi GQ. J Am Chem Soc 2008;130:5856. [12] Li XL, Wang XR, Zhang L, Lee SW, Dai HJ. Science 2008;319:1229. [13] Li XL, Zhang GY, Bai XD, Sun XM, Wang XR, Wang E, et al. Nat Nanotechnol 2008;3:538. [14] Valles C, Drummond C, Saadaoui H, Furtado CA, He M, Roubeau O, et al. J Am Chem Soc 2008;130:15802. [15] Liu N, Luo F, Wu HX, Liu YH, Zhang C, Chen J. Adv Funct Mater 2008;18:1518. [16] Viculis LM, Mack JJ, Kaner RB. Science 2003;299:1361. [17] Fan XB, Peng WC, Li Y, Li XY, Wang SL, Zhang GL, et al. Adv Mater 2008;20:4490. [18] Tung VC, Allen MJ, Yang Y, Kaner RB. Nat Nanotechnol 2009;4:25. [19] Hummers WS, Offeman RE. J Am Chem Soc 1958;80:1339. [20] Park S, An JH, Piner RD, Jung I, Yang DX, Velamakanni A, et al. Chem Mat 2008;20:6592. [21] Zhu J. Nat Nanotechnol 2008;3:528. [22] Liu W, Hoa SV, Pugh M. Compos Sci Technol 2005;65:2364. [23] Zhu J, Kim J, Peng H, Margrave JL, Khabashesku VN, Barrera EV. Nano Lett 2003;3:1107. [24] Mahmoud WE, El-Mossalamy EH, Arafa HM. J Appl Poly Sci 2011;121:502. [25] Kirkpatrick S. Rev Mod Phys 1973;45:574. [26] Zallen R. The physics of amorphous solids. New York: Wiley; 1985 [Chapter 4]. [27] Derrida B, Stauffer D, Herrmann HJ, Vannemius J. J Phys Lett 1983;44:701. [28] Kalaitzidou K, Fukushima H, Miyagawa H, Drzal LT. Polym Eng Sci 2007;47:1796. [29] Schaefer DW, Justice RS. Macromolecules 2007;40:8501. [30] Halpin JC, Kardos JL. Polym Eng Sci 1976;16:344.