Pingfu Wei and
Shibing Bai*
State Key Laboratory of Polymer Materials Engineering, Polymer Research Institute of Sichuan University, Chengdu 610065, China. E-mail: baishibing@scu.edu.cn; Fax: +86-28-85405133
First published on 27th October 2015
Due to the high specific area and hydrophobic nature of graphene, it is extremely difficult to fabricate graphene polymer composites without the re-stacking and re-aggregation of graphene sheets, especially when the matrix is a nonpolar polymer, such as high-density polyethylene (HDPE). Applying the solid-state shear milling (S3M) technique in preparation of a graphene/HDPE composite, we successfully solved the problems mentioned above. Highly dispersible exfoliated superfine HDPE/graphene compounding powder was prepared by the S3M technique in the solid state. HDPE/graphene composites with high dispersion and high mechanical performance were subsequently prepared via melt-based processing using this compounding powder. Transmission electron microscopy, scanning electron microscopy and wide-angle X-ray diffraction revealed that the graphene sheets exfoliated into individual sheets inside the polymer matrix. Fourier-transform infrared spectroscopy confirmed that new chemical bonds were generated between graphene and HDPE chains after the S3M process. Thermal analysis indicated that the interaction between HDPE and graphene was strengthened. The advanced material presents significant increases in yield strength, impact strength and elongation at break compared to conventional melt blending composites and neat HDPE. Tough and strong HDPE/graphene composites are suitable for a wide range of applications. Because of this method's solid-state melting-based processing, which does not require solvent, this approach possesses great potential for large-scale application in industry.
The solid-state shear milling (S3M) technique, based on self-designed pan-milling equipment, is used to treat tough, viscoelastic, heat-sensitive polymers at room temperature.21,22 The pan-milling equipment can act as a 3-dimentional scissors, providing very strong shear forces for pulverization, dispersion, mixing and activation. The S3M can effectively cut the multi-wall carbon nanotubes and induce strong interfacial interactions with polyamide 6,23 developing the structure and changing the properties of high-density polyethylene,24 solving the compatibility and viscosity matching problems, and controlling the morphology of the nanofiller in the polymer matrix.25,26 The S3M technique therefore represents a potential new way to disperse graphene sheets in a polymer matrix in the solid state. Typically, melt blending is the only practical way to prepare high-performance polymer/graphene composites on the industrial scale, as it neither depends on large amounts of solvents nor requires strict production conditions. It has been reported that reinforcement in melt-processing could be achieved at extremely low graphene loading levels (0.01–0.1 wt%),27 but graphene will re-stack in matrix with the increase of the loading level of graphene, which causes the mechanical properties to deteriorate. S3M methods provide a convenient way to prepare high-performance polymer-based graphene composites with much higher graphene loading level (0.1–4 wt%). A well-dispersed fine powder can easily undergo post-melting processing, and it can also be used as a master batch for all types of applications.
Polyethylene is one of most commonly used non-polar polymers. To date, there have been few relative papers studying polyethylene/graphene composites with high mechanical properties via melt processing. Solution-based processing becomes impossible because polyethylene is only soluble in solvents, such as xylene and trichlorobenzene above 120 °C. It is suggested that the interfacial interaction between graphene and high-density polyethylene (HDPE) is the key factor to obtain a homogenous dispersion in the polymer matrix.15 The interaction between HDPE and graphene is limited to weak van der Waals forces and hydrophobic–hydrophobic interactions.28 Therefore, the modification of HDPE or graphene seems necessary. However, a simple method that avoids using large amounts of solvent is critical for the large-scale application of polyethylene/graphene composites in industry.
In this study, using the S3M technique to treat the reduced graphene oxide and HDPE in the solid state, we aimed to disperse and exfoliate the graphene into the matrix under a strong 3D shear force in the solid state without aggregation and tried to develop a method for the large-scale application of graphene in industry. HDPE/graphene composites prepared by melt-based S3M processing exhibit improved mechanical properties compared to conventionally blended samples, even with neat polyethylene, especially the impact strength. The finely compounded powder of graphene and HPDE can easily be further processed (extrusion, injection molding, and compressing molding). The solventless nature, extensive adaptability and industrial compatibility let it to large-scale application.
To investigate the interaction between graphene and HDPE after solid state shear milling, compounding powder wrapped in filter paper was Soxhlet extracted by xylene at 140 °C for 48 hours. HDPE was extracted from the compounding powder and dissolved in xylene while insoluble free graphene stayed in the tube. By adding ethanol to the solution, grey flocculent precipitates emerges from the transparent liquid. The grey flocculent precipitates was suction filtrated from the solution and washed by ethanol and pure water 3 times, respectively. Then it was dried at 120 °C for 4 hours. We note the product as graphene modified HDPE (MG-HDPE). The conventional blended composite was selected as the control group which went through the same extracting process as mentioned above. The product was noted as the UM-HDPE.
Scanning electron microscopy (SEM) was performed on a Inspect F field-emission electron microscopy (FEI, Eindhoven, Netherlands).
Fourier-transform infrared (FT-IR) spectra were recorded on a Nicolet 6700 FT-IR spectrometer (Thermo Nicolet Ltd, Vernon Hills, IL, USA).
Differential scanning calorimetric (DSC) analysis was performed on a TA Q20 differential scanning calorimeter (TA Instruments, USA). In DSC analysis, dried samples under vacuum were heated from 40 °C to 180 °C and cooled slowly to room temperature in the first cycle to remove thermal history. The data were collected using a heating rate of 10 °C min−1 in a nitrogen atmosphere in the second cycle.
Thermal gravimetric analysis (TGA) was done in a TGA-Q50 (TA Instruments Co. Ltd, New Castle, DE, USA), using a heat rate 20 °C min−1 in a nitrogen atmosphere from 40 °C to 600 °C.
Wide-angle X-ray diffraction (WAXD) was performed using a DX-1000 diffractometer (Dandong Fangyuan Instrument Co., Ltd, China). The CuKα generator system was operated at 40 kV and 25 mA, and the scanning 2θ ranged from 5° to 40°, with a scanning rate of 1° min−1.
Mechanical characterization was performed on an RG-L-10 (Reger Instrument Co. Ltd., China) using a load cell of 10 kN at ambient temperature with a crosshead speed of 50 mm min−1 according to the ISO 527 and ISO 178 standards. Charpy notched impact tests were conducted at ambient temperature on a ZBC-4B impact tester (Shenzhen Xinsansi Measurement Technology Co., Ltd., China) according to the ISO 179 standard. For all samples, 5 strips were mechanically tested, and the averaged values of the five measurements were used.
TEM analysis (Fig. 2a and b) shows the graphene used in this study does not achieve a high monolayer content. However, it is reported that few-layer graphene is more efficient than monolayer graphene as a reinforcing agent in a polymer matrix,30 so at that point this is advantageous. The compounding powder sample for TEM test was ultrasonically treated in ethanol and then deposited to remove the free HDPE particles. The TEM analysis of the compounding powder indicates that after milling with graphene sheets, some HDPE particles are completely coated (Fig. 2d), while others are partially attached to the edges of graphene sheets (Fig. 2c). Hence, it is indicated that pre-treating with the S3M process may effectively strengthens the interaction between the polymer matrix and graphene under a strong shear force and even mechanical chemical reaction may occurred, which is crucial for the graphene not re-stacking during the post-melting process.
![]() | ||
Fig. 2 TEM images of pristine graphene (a and b), 4 wt% HDPE/graphene compounding powder with 30 cycles S3M (c and d). |
The morphology of graphene in the S3M-processed composites with a very low content (0.1 wt%) of graphene is shown in Fig. 3a and b. After undergoing the melting-diluting process, a visible “boundary” between the neat polyethylene and compounding powder can be observed. The dark area shown in Fig. 3a is the low content master batch, which was evenly although non-continuously distributed in the polymer matrix. According to previous work, the structure of polyethylene does not changed to the degree at which the phase separation will happen, but the molecular weight slightly decreased with the change of the crystallization behavior after milling.24 Additionally, graphene were fully exfoliated into the individual sheets (Fig. 3b). Because of low content of graphene in HDPE matrix, it is difficult to find the graphene sheets under TEM. But the morphology of graphene in matrix is basically as same as what showed in Fig. 3 which we believe can be represented for the whole sample. By increasing the content of the master batch, the “boundary” disappeared (actually it is disappeared when graphene content reaches to 0.5 wt%) and the dark master batch area became continuous (Fig. 3c and d). TEM images (Fig. 3c and d) verify that the graphene sheets were exfoliated into few-layer sheets in the polyethylene matrix, basically keeping the original morphology before the S3M process. It is shown that well exfoliated several-layer graphene effectively serves as reinforcing units within composites, while conventionally blended composites (Fig. 4) do not achieve the same effect because the graphene sheets re-stack into dozes of layers during the melting process.
![]() | ||
Fig. 3 TEM images of S3M-processed composites with 0.1 wt% (a and b), 1 wt% (c and d) content of graphene. |
![]() | ||
Fig. 4 TEM images of conventional blending composites with 0.1 wt% (a and b), 1 wt% (c and d) content of graphene. |
Conventional melt-based processing is not able to avoid the re-stacking and aggregation of graphene sheets because of the strong π–π interaction and van der Waals interaction between graphene sheets that cause them to readily re-stack into graphite-like powders or films.31,32 It is verified in Fig. 4 that the graphene sheets re-stack or re-aggregate to form graphite-like structures (Fig. 4b–d) after a high-temperature melting process. This may greatly lower the high specific surface area of graphene and hinder the effective load transfer from the matrix to the fillers, leading to a significant drop in the mechanical properties, which is also confirmed in the following discussion.
To further explore the morphology of graphene sheets in matrix at a larger scale, we choose Scanning Electron Microscope (SEM) as an investigation. It is clearly showed in Fig. 5a and b that after the S3M process, HDPE particle intercalated into the graphene sheets which avoids the re-stacking of graphene sheets during melting process. While the graphene sheets in conventionally blended composite re-stacked into a buck of graphite (Fig. 5c and d). So it is confirmed that S3M technique is an effective way to disperse graphene sheets into nonpolar polymer matrix, in this case is HDPE.
![]() | ||
Fig. 5 TEM images of cross-section from S3M processing composite (a and b) and conventional blending composites (c and d) with 1 wt% graphene. |
The mechanical chemical reactions in S3M are extremely complex processes, so only can we make a rough estimate of the real reactions occurred during the S3M process. The Scheme 1 shows possible mechanical chemical reactions during the S3M process. It is deduced that on the one hand during the S3M process polyethylene chains broke and carbon free radicals were generated, in the meantime carbon–oxygen double bond also broke. So it is possible to form ether bond (C–O–C) from the combination of carbon free radical and the broken carbon–oxygen double bond. On the other hand, previous work33 showed that reactions occurred between oxygen (and carbon dioxide) in air and free radicals generated as a result of mechanochemical degradation of HDPE under pan-mill stress. As a result, the carbonyl groups (carboxyl, aldehyde) were introduced into the polyethylene chains. The carbonyl groups are able to further react with the hydroxyl in graphene. Because of its instantaneity, it is difficult to prove the existence of the carbon free radicals. But it is a reasonable inference to estimate the mechanical chemical reaction occurred.
DSC second heating curves reveal the crystallization of composites from melts. Tm of S3M processed composites increased to around 131 °C. While Tm of conventional blended composites are slightly lower than the pure HDPE (Fig. 7a and b). The reason why the Tm is higher in S3M composites can be explained to the heterogeneous nucleated, better formed PE crystals templated by the RGO. However, the random distribution of re-stacked graphene in conventional blended composites hinder the crystallization of PE which results in much more faults and imperfections. DSC cooling curves also shows the increase in Tc which is attributed to the heterogeneous nucleation induced by the 2D RGO sheets. And the half peak width of the S3M samples is smaller than that of HDPE and conventional samples (Fig. 7c). The smaller the half peak width is, the narrower the size distribution of the crystallites is. It indicates that the crystallization of S3M is more uniformly nucleated. Carbon residue of S3M processed, conventional blended composites and neat HDPE in TGA analysis are 1.6%, 1.1%, and 0.28%, respectively (Fig. 7d). Theoretically, carbon residue of HDPE/graphene composites containing 1 wt% graphene is around 1%. The 1.6% carbon residue of S3M processed composites can be explained as a result of intercalation of HDPE into the graphene sheet. The polymer between the graphene sheets are protected by the graphene sheets from the thermal decomposition, which leads to the increase in carbon residue.
![]() | ||
Fig. 8 X-ray diffraction pattern of conventionally blended composite, S3M-processed composite with 4 wt% graphene, HDPE, and graphene. |
Fig. 9 shows the XRD patterns of HDPE/graphene compounding powders subjected to different numbers of milling cycles. An extra diffraction peak (around 20°) appeared in the 1-cycle-milling compounding powder, but it vanished with the increasing of the number of cycles. It is stated that S3M is able to exfoliate graphene sheets into individual sheets and disperse them well in HDPE in the solid state. This is consistent with the results shown in Fig. 8 that the state of well exfoliated graphene in compounding powder is able to be preserved after the melting process.
![]() | ||
Fig. 9 X-ray diffraction patterns of compounding powder subjected to different numbers of milling cycles. |
![]() | ||
Fig. 10 Yield strength and elongation at break for 4 wt% graphene loading HDPE/graphene composites with different numbers of S3M milling cycles. |
The mechanical properties of polymer composites are determined by many factors, among which the morphology of the graphene, orientation of fillers, and degrees of exfoliation and dispersion are the most important.37 Fig. 11 shows that the impact strength increased from 22.6 kJ m−2 for the 0.1 wt% conventionally blended sample to 27.6 kJ m−2 for the 0.1 wt% S3M-processed sample. When the graphene content reaches 1 wt%, the impact strength of the S3M processed composites increases to 32 kJ m−2, which is a 56.1% and 25.5% improvement compared to the conventionally blended sample and neat HDPE, respectively. The impact strength reflects the ability of a material to absorb energy at fracture when exposed to a sudden impact. Different adhesions between the graphene sheets and the polymer matrix as well as the difference in the dispersion state in the polymer will result in different energy-absorbing mechanisms.38 It was reported that introducing graphene as a nanofiller into the polymer matrix usually reduced the impact strength. Reduced impact strength usually results from the incorporation of a rigid filler into a relatively tough polymer matrix.19,39 However, the impact strength improved considerably in our study. According to the SEM results, HDPE is intercalated between the graphene sheets in S3M composites. When composites exposed to the sudden impact, strong and tough graphene sheets can slowly transfer the strain to the HDPE matrix. The exfoliated graphene sheets in this case do not act as a rigid filler but rather a flexible one. Unfolded graphene sheets absorb a substantial amount of energy by changing their morphology when exposed to sudden impact.
![]() | ||
Fig. 11 Impact strength for S3M-processed and conventionally blended HDPE/graphene composites with different graphene loading levels from 0 to 4 wt%. |
Fig. 12 also shows a significant improvement in the yield strength of S3M-processed composites compared to both conventionally blended samples and neat HDPE. However, when graphene content increases to 4.0 wt%, there is a sudden yield strength increase in conventional composites. We assume that this improvement can be explained to “aggregation effect” which means re-stacked graphene can act as a particulate filler when it reaches to a certain quantity. As a result, the yield strength of conventional composite suddenly increases to a value close to that of S3M composite when graphene content reaches to 4 wt%, but the impact strength and elongation at break remain in a low level. Besides, fully exfoliated graphene sheets in S3M processed composites can effectively serve as a reinforcing agent which results in the significant improvement in the yield strength. Therefore, HDPE/graphene composites prepared by S3M achieve reinforcing and toughening effects at the same time. To our knowledge, no report has ever presented such improvements in the mechanical properties of non-polar thermoplastic polymers prepared via melt-based processing.
![]() | ||
Fig. 12 Yield strength for S3M processed and conventionally blended HDPE/graphene composites with different graphene loading levels from 0 to 4 wt%. |
This journal is © The Royal Society of Chemistry 2015 |