(UHMWPE)/graphene oxide - Wiley Online Library

Research article Received: 12 September 2015,

Revised: 08 January 2016,

Accepted: 19 January 2016,

Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI: 10.1002/pat.3779

Preparation and investigation of tribological properties of ultra-high molecular weight polyethylene (UHMWPE)/graphene oxide Hiva Bahrami, Ahmad Ramazani S. A.*, Mojtaba Shafiee and Amanj Kheradmand This article has been devoted to investigation of the tribological properties of ultra-high molecular polyethylene/graphene oxide nanocomposite. The nanocomposite of ultra-high molecular polyethylene/graphene oxide was prepared with 0.5, 1.5, and 2.5 wt% of graphene oxide and with a molecular weight of 3.7 × 106 by in-situ polymerization using Ziegler–Natta catalyst. In this method, graphene oxide was used along with magnesium ethoxide as a novel bi-support of the Ziegler–Natta catalyst. Analyzing the pin-on-disk test, the tribological properties of the nanocomposite, such as wear rate and mean friction coefficient, were investigated under the mentioned contents of graphene oxide. The results showed that an increase in graphene oxide content causes a reduction in both wear rate and mean coefficient friction. For instance, by adding only 5 wt% graphene oxide to the polymeric matrix, the wear rate and mean coefficient friction decreased about 34% and 3.8%, respectively. Also, the morphological properties of the nanocomposite were investigated by using X-ray diffraction and scanning electron microscopy. In addition, thermal properties of the nanocomposite were analyzed using differential scanning calorimetry, under various contents of graphene oxide. The results of the morphological test indicated that the graphene oxide was completely exfoliated into the polymeric matrix without any agglomeration. Copyright © 2016 John Wiley & Sons, Ltd. Keywords: nanocomposite; UHMWPE; graphene oxide; tribological properties

INTRODUCTION

EXPERIMENTAL

Ultra-high molecular weight polyethylene (UHMWPE) is one of the thermoplastic high-density polyethylene groups that have been at the focus of many researchers and of great use, especially in artificial bodies, because of its extraordinary properties such as high strength and stiffness, good fatigue, and appropriate wear characteristics.[1–4] Today, for many complex applications of polymer, reinforcement is necessary, and UHMWPE is not an exception. A way to improve tribological properties of UHMWPE for both industrial and orthopedic applications is to add nano particles,[5,6] nanofibers,[7] and nanotubes[8,9] to the matrix. The discovery of graphene, which has some extraordinary properties like Young’s modulus of 1 TPa and ultimate strength of 130 GPa, has caused this single layer to be known as the strongest material ever measured. Therefore, the combination of extraordinary physical properties and the ability to be dispersed in various polymer matrices have resulted in a new class of polymer nanocomposites.[10,11] In this work, UHMWPE/graphene oxide (GO) nanocomposite has been prepared via in-situ Ziegler–Natta polymerization, which is a successful method for the preparation of exfoliated nanoparticles in polymeric matrices. Then, the tribological properties have been investigated, including wear rate, mean friction coefficient, and morphological properties of the prepared samples.

Materials

Polym. Adv. Technol. (2016)

Graphite powder was purchased from Dae-Jung Chemicals (Shiheung-city, Gyeonggi-do, Korea) and was used to synthesize GO. The ethylene monomer (industrial grade) was supplied by Amir Kabir Petrochemical (Mahshahr, Iran) and was further purified by being passed through molecular sieve columns (4A type having a 4-A pore size). High-purity argon (_99.999%) was supplied by Arkan Gas (Tehran, Iran) and was used in creating an inert atmosphere for the catalytic system preparation as well as the polymerization. Industrial grade n-hexane was purchased from Amir Kabir Petrochemical and was purified by a sodium benzophenole complex (diphenyl ketyl). Titanium tetrachloride (TiCl4; Riedel-de Ha¨en as a catalyst, Berlin, Germany), graphene oxide, and magnesium ethoxide were used as support of catalyst and dibutyl phthalate (DBP; Merck, Kenilworth, NJ, USA) were used as an internal donor. Triisobutylaluminium (TIBA; Aldrich, St. Louis, MO, USA) acted as a co-catalyst. * Correspondence to: Ramazani S. A. Ahmad, Sharif University of Technology, P.O. Box: 11365-8639, Tehran, Iran. E-mail: [email protected] H. Bahrami, A. Ramazani S. A., M. Shafiee, A. Kheradmand Department of Chemical and Petroleum Engineering, Sharif University of technology, Tehran, Iran

Copyright © 2016 John Wiley & Sons, Ltd.

H. BAHRAMI ET AL.

Catalyst preparation The preparation of catalyst was carried out according to the method developed by our group.[12–14] Pristine graphite powder was dried by heating for 6 h at 200°C under vacuum. Treatment of graphite with mineral acids is the most widely used technique to impart acidity to the graphenic surface. Oxidation of natural graphite flakes has been carried out according to the modified Hummers method.[15] Then, graphite oxide was fed into a triple-necked, round-bottom flask, which was equipped with a magnetic stirrer and immersed in an oil bath. Two openings of the flask were equipped with gas valves. One was connected to a vacuum pump while the other was connected to the argon cylinder; the third opening was sealed with a rubber septum for injecting further chemicals. The inert atmosphere was achieved by three times degassing and backfilling the flask with the high-purity argon to assure a completely inert condition. After applying an inert atmosphere, 150 ml of n-hexane/toluene (50/ 50 wt%) was injected into the flask. Temperature was gradually increased to 80° C while the flask contents were stirring vigorously. After reaching the desired temperature, TiCl4 (8 ml) and dibutyl phthalate (1 ml) were injected into the flask, and its contents were stirred for 2 h at constant temperature. Finally, after cooling the flask contents to 50° C, they were washed 10 times with n-hexane under argon atmosphere to assure complete removal of the unreacted residuals from graphene-catalyst complex. Finally, the flask was cooled down to the room temperature; the graphene-catalyst was dissolved in 100 ml of n-hexane, and the produced graphene-catalyst complex was stored under argon atmosphere for further application. Polymerization A 1-l pressure reactor (versoclave; Buchi AG, Flawil, Switzerland) equipped with a mechanical stirrer, a temperature controller, and a pressure control system was purged with high-purity argon at 1.1 bar at 90° C to assure a completely dry and inert atmosphere. Then, 400 ml of degassed n-hexane was transferred to the reactor. Subsequently, while stirring, degassed triisobutylaluminium (TIBA) was injected into the reactor. After 5 min, 10 ml of the graphene-catalyst complex produced beforehand was added to the mixture, and instantly ethylene monomer was introduced into the reactor at different pressure. Finally, after the polymerization was carried out for the desired time, the ethylene inlet was closed; reactor was purged with air, and 10 ml of concentrated HCl was injected into the reactor to deactivate the catalyst and terminate the polymerization. After bringing the reactor temperature down to the room temperature, the contents of the reactor were vacuum filtered, washed with ethanol and acetone, and subsequently dried under vacuum at 80° C for 24 hr. Synthesis of pure UHMWPE was carried out with the same procedure mentioned earlier for production of nanocomposites with this exception that graphene was not used in the catalytic system.

MEASUREMENTS AND CHARACTERIZATION X-ray diffraction The X-ray diffract meter was used to measure the basal spacing between GO layers in the nanocomposites. The X-ray diffraction (XRD) was performed by an expert PRO MRD (Philips, Eindhoven, the Netherlands) diffract meter using Cu Kα radiation

wileyonlinelibrary.com/journal/pat

(λ = 0.15406 nm). The samples were scanned in 2 hr ranges from 1° to 10° at a rate of 1°/sec. The generator was operated at 40 kV and 40 mA. The interlayer spacing (d001) of the graphene was calculated following the Bragg eqn (1): 2dsin θ ¼ λ

(1)

Scanning electron microscopy For scanning electron microscopy (SEM) analysis, samples were cut in liquid nitrogen, and the images were taken from the cross section of the gold-coated specimens. The analysis was performed with a scanning electron microscope Tuscan, Vega series 2007 (TESCAN Company, Brno, Czech Republic). Determine of average molecular weight In order to determine the average molecular weight of UHMWPE, the Mark–Houwink’s equation was used. For this purpose, UHMWPE with different concentration was dissolved in solvent of decalin at 135°C. Then with extrapolation of viscosityconcentration’s diagram, intrinsic viscosity has been derived and with inserting this value in March–Houwink’s eqn (2); average molecular weight of UHMWPE has been obtained. ½η¼ 5:37M1:49

(2)

Wear and friction test Pin on disk is essentially a material tribology test and in general engineering testing that is frequently used to screen materials. This test provides a reciprocating linear motion or curvilinear motion where a disk is rotated under a stationary pin with some compression applied and, in order to obtain the coefficient of friction-resisting force, the pin is measured as well as the compressive force.[15] So pin-on-disk devices were used in order to measure the wear rate and friction coefficient that this test has been carried out on ASTM G55 (West Conshohocken, PA, USA).

RESULTS AND DISCUSSION Morphological studies With the help of the SEM images, a uniform distribution of the reinforcing agent could be concluded, and as it is evident from Fig. 1, there is no agglomeration in the SEM images. The results confirm a uniform distribution of graphene oxide into the polymeric matrix using in-situ polymerization method. But the inflation rate or disintegration of layers could not be analyzed, and this led us to utilize XRD method. Information obtainable from diffraction of a crystal includes peak angle, relative intensity of peaks, and the width of each peak. This information forms the basis of XRD and has yielded many applications for it. As pointed out earlier, interlayer distance could be computed using Bragg law. In fact, Bragg law explains the reason behind the peaks of an X-ray diffraction. There are many sheets imaginable for each crystal with different inter-sheets. The Bragg’s conditions occur in a specific angle for each sheet. In an XRD diagram, each peak is related to a sheet, which is characterized by Miller indices via crystallography. The XRD diagram of the synthesized polymers is shown in Fig. 2. In this image, two sharp peaks could be seen at angles about 2θ = 21.4 and 2θ = 23.9, which relate to the reflection of



PREPARATION AND INVESTIGATION OF UHMWPE

Figure 1. Scanning electron microscopy (SEM) of nanocomposite with (a) 0.5% and (b) 2.5% graphene oxide.

Figure 2. The X-ray diffraction diagram of the synthesized polymers.

sheets 110 and 200 of orthorhombic crystalline phase (OR110 and OR200) of UHMWPE. It should be mentioned that in some references, monoclinic crystalline phase has been reported, which depicts the presence of such peak in samples under stress. But the presence of such crystalline phase is unusual for unstressed UHMWPE samples.[16] According to the mentioned text earlier, in-situ polymerization is an efficient way of producing fully exfoliated UHMWPE/GO nanocomposites. The synthesis of the nanocomposites with an exfoliated structure is much more feasible when in-situ polymerization is applied when compared with other methods, even at higher percentages of the reinforcing agent. The reason could be ascribed to the ease of the monomer entry into the interlayer space and the simultaneous growth of polymeric chains on the graphene’s surface in in-situ polymerization compared with the entry of long polymeric chains in other methods.[17,18] Exfoliation and separation of graphite sheets into very thin isolated layers (graphene sheets) could have an outstanding effect on the final properties of the produced composite. In past, using graphite in polymers as a reinforcing agent had not used to cause considerable improvement in their properties because researchers yet had not obtained exfoliated graphite. This meant


that graphite, with a low surface area, could just improve polymers’ properties a bit. The important point is that under exfoliation of the layers inside a polymeric matrix, improvement of the properties is achievable even at low amounts of graphene.

Study of crystallization structure of synthesized samples In this section, crystallization behavior of pure polymers and nanocomposites is studied. As it is known, one could determine melting temperature, crystallization temperature, glass transition temperature, percentage of crystallinity, etc. by the aid of the differential scanning calorimetry (DSC) thermal test. Figure 3 illustrates, as an example, the DSC diagram of the pure polymer and the synthesized nanocomposites. As it could be seen from the DSC curve of UHMWPE, there exists an endothermic sharp single peak for the melting and an exothermic sharp single peak for the crystallinity. The DSC analysis results are presented in Table 1 for all samples. These results include melting temperature (Tm), crystallinity temperature (Tc), heat of fusion (ΔHf), and percent crystallinity (Xc). Percentage of crystallinity could be calculated by eqn (3):



H. BAHRAMI ET AL.

Figure 3. Differential scanning calorimetry diagram of pure polymer and synthesized nanocomposites.

Table 1. DSC analysis results for each sample (mean ± standard deviation) Sample

Melting temperature (Tm)

Crystallinity temperature (Tc)

Heat of fusion (ΔHf)

Percent crystallinity (Xc)

B-PGN-0 B-PGN-0.5 B-PGN-1.5 B-PGN-2.5

132.82 ± 0.41 134.33 ± 0.67 134.54 ± 0.74 133.62 ± 0.81

116.69 ± 0.59 117.28 ± 1.08 117.54 ± 0.87 117.89 ± 0.75

158.92 ± 2.63 162.16 ± 3.58 168.07 ± 3.01 158.31 ± 4.04

54.23 ± 0.89 55.62 ± 1.22 58.23 ± 1.04 55.42 ± 1.41

DSC, differential scanning calorimetry; B-PGN, polyethylene graphene nanosheet.

XC ¼

Hf 100 ð1 φÞH°

(3)

which φ is the graphene weight fraction and H is the melting temperature of 100% crystalline polyethylene (equal to 293 J/g). Figure 4 shows the changes in the percentage of the crystallinity, melting point, and crystallinity temperature of the synthesized nanocomposites respectively based on the graphene weight percent at the molecular weight of 3.7 × 106. The diagram of the changes of percent crystallinity based on graphene percent indicates an increase of percent crystallinity for nanocomposites up to a specific graphene percent and then a decrease for its higher percentages. The increase of crystallinity in low graphene percentages is because graphene surface acts as nucleation sites for the crystallization, which as a result yields an increase of the nucleation and a growth of the crystallites and in turn a boost of the percent crystallinity. Although graphene acts as nucleation sites, but on the other hand, it decreases the mobility of the polymeric chains.[19] Thus, it shows an inhibitory role against the growth of the crystallites, which is one of the reasons why crystallinity decreases in higher graphene percentages. Another reason could be attributed to the increase of cooling rate during the forming process, which causes a decrease in the time required by the chains to be re-oriented into a crystallite form.[20–23] As shown in Fig. 5, the melting temperature had an increasing trend up to a specific percentage and then a decreasing one at higher weight percentages. The reason is attributable to the increase of nucleation occurred by the increase of graphene


Figure 4. Diagram of the changes of percent crystallinity based on graphene percent.

content and crystallinity and an improved thermodynamic stability of the crystals formed.[24] As a result at higher percentages of graphene, decrease of crystallinity and increase of thermal conduct are expected, which in turn would slow down the increase of the polymer heating rate. The diagram of the changes of the crystallinity temperature based on the graphene percentage (Fig. 6) shows an increasing trend with a sharp slope for the samples holding a graphene content up to 0.5wt% and then a low slope for higher weight percentages. The reason of the crystallinity temperature increase could be attributed to the increase of the crystallinity and also the boost of the polymer thermal conductivity and as a result




Figure 5. Diagram of changes of melting point based on graphene percent.

Figure 6. Diagram of crystallinity of samples based on percent of graphene oxide.

an increase in the cooling rate of the polymer. In higher percentages, the increase rate of the crystallinity temperature decreases. This was so because of the conflict of the two effects, namely, the crystallinity decrease (and subsequently, a decrease of the crystallinity temperature) and the cooling rate increase (and subsequently, an increase of the crystallinity temperature).[25] Tests of friction and wear In this section, the wear properties of the samples are investigated using pin-on-disk test. These properties include average dynamic coefficient of friction and special wear. Special wear rate (Kw) is obtained via eqn (2). Kw ¼

Δm ρF N L

(4)

In the previous equation, Δm indicates the weight loss of the sample in milligrams, ρ denotes the density of the sample in grams per cubic centimeters, FN is for vertical force in Newton (N), and L stands for the wear distance traveled in meter. In Figure 7, the dynamic coefficient of friction could be seen based


on the wear distance traveled for the samples. The results of the wear test are presented in Table 2. These results include special wear rate and average dynamic coefficient of friction. Before turning into the analysis of the results, the wear mechanism would be studied for a unidirectional motion. As a result of a unidirectional wearing, two phenomena are developed at the surface being worn: firstly, formation of some waves perpendicular to the wear direction, and secondly, formation of some fibers along the wear direction. The waves are formed as a result of the strain accumulation, or in other words because of the repetitive changes occurring in the surface. The distance between the waves depends on the roughness of the contacting surfaces and the material properties. The phenomenon of ripple surface is not unique to UHMWPE, but all ductile metal surfaces show such behavior during wear. But in contrast to metals, UHMWPE has a special molecular structure. As we know, molecular chains of UHMWPE are very large and oriented in a semicrystalline layered (lamellar) structure. Lamellar crystals are randomly oriented above glass transition temperature. During the deformation, these crystals could rotate and rearrange according to the deformation. If plastic strain is sufficiently large, the lamellar crystals could break and form fiber-like structures. It should be noted that under a unidirectional wearing, surface tensile stresses and sheer stress would be parallel to the direction of tension and re-oriented molecules. This model, which is called tensile rupture or tensile instability, points to the importance of fibril strength in a unidirectional wearing. The fibril strength depends on the degree of the orientation of polymer molecules within fibrils and the number of polymeric chains in cross-sectional area unit. The degree of the orientation of polymer molecules is dependent upon the mobility of polymeric chains at the surface. Number of molecules in cross-sectional unit area depends on both the degree of molecular orientation and chain length, i.e., polymer molecular weight.[26,27] Figure 8 depicts a considerable decrease of the wear rate in the presence of graphene as the reinforcing agent such that by a mere 0.5 wt% increase of graphene content in the polymeric matrix, the wear rate decreases by about 34%. This reduction could be attributed to factors such as improvement of mechanical properties of polymer, increase of crystallinity, and also



H. BAHRAMI ET AL.

Figure 7. Dynamic coefficient of friction based on wear distance.

Table 2. Results of wear test (mean ± standard deviation) Sample

Special wear rate average × 106

Dynamic coefficient of friction

B-PGN-0 B-PGN-0.5 B-PGN-1.5 B-PGN-2.5

1.840 ± 0.105 1.220 ± 0.079 0.910 ± 0.031 0.605 ± 0.010

0.2112 ± 0.0096 0.2032 ± 0.0072 0.1853 ± 0.0042 0.1652 ± 0.0051

Figure 9. Average change of dynamic coefficient of friction for nanocomposites based on graphene weight percent.

has decreased by increasing the percent graphene. The reason could lie in the lubricating properties of graphene.[30]

CONCLUSION

Figure 8. Changes of friction coefficient for samples based on percent of graphene oxide.

decrease of friction coefficient. It should be noted that increase of crystallinity improves the wear properties of polymers.[28–30] Figure 9 shows the average change of the dynamic coefficient of friction for the nanocomposites based on the graphene weight percent. As it could be seen, the coefficient of friction


The present work has investigated the tribological and morphological properties of UHMWPE nanocomposites. The morphological studies confirmed that the nanocomposite produced had an exfoliated structure of graphene oxide without any agglomeration in the polymeric matrix. The results showed that the tribological properties such as wear rate and mean friction coefficient decrease by adding graphene oxide and, thus, the produced nanocomposite could be an appropriate choice to be used in implant and artificial body, and also in other applications where friction is low. It can be concluded that graphene oxide is a suitable filler, which yields extraordinary improvements to the polymer and thereby could be of potential use in technical and advanced applications.




REFERENCES [1] M. K. Steven, UHMWPE Biomaterials Handbook, 2nd edn. . Academic Press, Boston, 2009. [2] B. M. Amoli, A. Ramazani S.A, H. Izadi, J. Appl. Polym. Sci. 2012, 125, 453. [3] A. Ramazani S.A, M. G. Saremi, B. M. Amoli, H. Izadi, Polym. Compos. 2012, 33, 1858. [4] J. P. Davim, N. Marques, Tribol. Lett. 2001, 11, 91. [5] D. S. Xiong, J. M. Lin, D. L. Fan, Biomed. Mater. 2006, 1, 175. [6] D. Xiong, J. Lin, D. Fan, Z. Jin, J. Mater. Sci. 2007, 18, 2131. [7] X. Ren, X. Q. Wang, G. Sui, W. H. Zhong, M. A. Fuqua, C. A. Ulven, J. Appl. Polym. Sci. 2008, 107, 2837. [8] Y. Xi, A. Yamanaka, Y. Bin, M. Matsuo, J. Appl. Polym. Sci. 2007, 10, 2868. [9] J. F. Gao, Z. M. Li, Q. J. Meng, Q. Yang, Mater. Lett. 2008, 62, 3530. [10] K. Hyunwoo, A. A. Abdala, W. Christopher, Macromolecules 2010, 43, 6515. [11] C. Lee, X. Wei, J. W. Kysar, J. Hone, Science 2008, 321, 385. [12] A. Ramazani S.A, F. Tavakolzadeh, H. Baniasadi, Polym. Compos. 2009, 30, 1388. [13] S. J. Nikkhah, A. Ramazani S.A, H. Baniasadi, F. Tavakolzadeh, Mater. Des. 2009, 30, 2309. [14] A. Ramazani S.A, M. G. Saremi, B. M. Amoli, H. Izadi, Polym. Compos. 2012, 33, 1858.


[15] W. S. Hummers, R. E. Offeman, J. Am. Chem. Soc. 1958, 80, 1339. [16] G. Zamfirova, J. M. Perena, Polym. J. 2002, 43, 125. [17] M. A. Milania, D. Gonzálezb, R. Quijadab, N. R. S. Bassoc, M. L. Cerradad, D. S. Azambujaa, G. B. Galland, Compos. Sci. Technol. 2013, 84, 1. [18] Z. Xu, C. Gao, Macromolecules 2010, 43, 6716. [19] J. Rong, M. Sheng, H. Li, E. Ruckenstein, Polym. Compos. 2002, 23, 658. [20] Y. Chen, Y. Qi, Z. Tai, X. Yan, F. Zhu, Q. Xue, Eur. Polym. J. 2012, 48, 1026. [21] S. R. Bakshi, J. E. Tercero, A. Agarwal, A: Appl. Sci. Manuf. 2007, 38, 2493. [22] X. Wanga, H. Yang, L. Song, Y. Hu, W. Xin, H. Lu, Compos. Sci. Technol. 2011, 72, 1. [23] Y. Li, J. Zhu, S. Wei, J. Ryu, L. Sun, Z. Guo, Macromol. Chem. Phys. 2011, 212, 195. [24] D. -H. Lee, H. -S. Kim, K. -B. Yoon, K. E. Min, K. H. Seo, S. K. Noh, Sci. Technol. Adv. Mater. 2005, 6, 457. [25] J. Xian, T. D. Lawrence, Polym. Compos. 2012, 33, 636. [26] A. Wang, C. Stark, J. H. Dumbleton, J. Eng. Med. 1996, 210, 141. [27] B. J. Briscoe, T. A. Stolarski, J. Tribol. 1981, 103, 503. [28] Z. Tai, Y. Chen, Y. An, X. Yan, Tribol. Lett. 2012, 46, 55–63. [29] Y. -S. Zoo, J. -W. An, D. –. P. Lim, Tribol. Lett. 2004, 16, 305. [30] Z. Wei, Y. –. P. Zhao, S. L. Ruan, P. Gao, T. X. Yu, Surf. Interface Anal. 2006, 38, 883.

