Study of Tensile Properties of Multiwalled Carbon Nanotube/Polyether Ether Ketone Polymer Composites at the Nanoscale.
Single-walled and multiwalled carbon nanotubes (SWCNTs and MWCNTs) have been widely studied as excellent nanofillers for polymers, metals, and ceramics due to their outstanding thermal, electrical, and mechanical properties. Various thermoplastic polymers including poly-ethylene (PE), poly-styrene (PS), poly-methylmethacrylate (PMMA), and poly-ether-ether-ketone (PEEK) have been successfully dispersed with CNTs [1-7]. Among these polymer materials, PEEK is a high-performance semicrystalline thermoplastic polymer with a great mechanical properties, thermal stability, and chemical resistance [6, 8-10] and has been widely used in aerospace, electronics, and chemical industries.
PEEK reinforced with MWCNTs exhibits improved mechanical properties even at low CNT dispersion fraction. Rong et al. reported that, by adding 5% CNT nanofillers, the Young's modulus and tensile strength of composites were increased by 20% and 3%, respectively . Davies et al. studied CNTs/PEEK composites fabricated through extrusion deposition method and reported the tensile strength of 5 wt% CNTs reinforced composites was enhanced by 21% . Ogasawara et al. concluded that the dispersion of MWCNTs aligned longitudinally led to an improvement of initial Young's modulus by 98% for a 15 wt% CNT loading . Moreover, similar results were found in CNT/PEEK composite even at elevated temperatures by Deng et al . The reported Young's modulus was enhanced by 70% and 162% for a 15 wt% CNT dispersion at 100[degrees]C and 200[degrees]C, respectively. These mechanical property improvements indicate that the CNT fillers, usually stiffer than matrix materials, are able to effectively transfer load and the composite can absorb more energy than the matrix.
It is generally difficult to extrapolate the nanoscale material properties from the bulk values due to a fact that the nanoscale properties are critically influenced by material interfaces and other microstructures or nanostructures not accounted for in sufficient accuracy compared with the bulk measurements . Halpin-Tsai model is commonly used for predicting the tensile properties of continuous fiber-reinforced composites. Based on this model, the tensile strength and elastic modulus of MWCNTs/PEEK composites for 15 wt% of MWCNT dispersion, for example, could reach up to 1.4 and 43 GPa [15-17]. However, the experiment results of 15 wt% MWCNTs/PEEK showed a maximum strength and modulus of 0.12 and 8.3 GPa, respectively . The gap between the Halpin-Tsai model and experiment results suggests that there are unfavorable factors deviating from ideal CNTs dispersion, alignment, and interfacial interactions between CNTs and the polymer matrix . The influence from these factors is essential to the bulk composite performance and can only be better understood by systematic experiments using microscale to nanoscale specimens.
To conduct a small-scale tensile testing, photolithography and chemical etching are the frequent methods used for the preparation and the release of submicron tensile specimens for in situ mechanical tests. These approaches not only have complicated sample preparation procedures but also significantly affect the reliability of the testing results. Read  and Hoffman  used wet etching to release Ti-Cu-Ti and aluminum tensile specimens. However, a challenge for wet etching method is to create stress-free specimens. The stress introduced from thin-film preparation process can detrimentally affect the experiment and the results. Haque and Saif  chose reactive ion etching (RIE) technique to prepare and release ~100nm-thick Al specimen from silicon wafer. Although stress-free samples can be made by the dry etching approach, a freestanding specimen is likely to be damaged while transferring and loading into SEM and TEM systems. The use of photolithography not only complicates the specimen fabrication process but also limits the testing materials to those of limited candidates that can be deposited on a silicon or glass substrate. Thus, a main challenge remains as how to create a nanoscale sample from those materials that are unable to be prepared through physical or chemical deposition, such as polymers and polymer-based composites.
There have been limited publications focusing on the nano-scale tensile testing and properties of CNT-reinforced nanocomposites utilizing in situ experiment or simulation approaches [21-23]. Direct measurement and observation have hardly been conducted on the interaction between CNTs and polymer matrix under uniaxial loads from microscale to nanoscale specimens. To gain a better understanding of nanocomposite performance at the nanoscale, authors developed an in situ tensile testing approach and tested MWCNTs/PEEK nanocomposites using the nanoscale specimens prepared by FIB.
MWCNTs/PEEK nanocomposites were fabricated by dispersing 6.5 wt% MWCNTs in the PEEK polymer matrix. MWCNTs were synthesized by chemical vapor deposition and obtained by Bussan Nanotech Research Institute Inc. (Tokyo, Japan). The diameter and average length of the MWCNTs were in the range of 20-100 nm and a few micrometers, respectively. Two-axis extruder was used to mix PEEK and MWCNTs and fabricate the thin nanocomposite films.
The microstructure and surface morphology of MWCNTs/PEEK were examined with a ZEISS Auriga 60 focused ion beam/field emission scanning electron microscope (FIB/FE-SEM). Secondary electron images were taken at an acceleration voltage of 3 kV. TEM images were taken using a FEI Talos F200C field emission transmission electron microscope (FE-TEM) operating at an accelerating voltage of 200 kV. Tensile tests were conducted on a Shimadzu AGS-X Universal Tester with IN load cell at a constant displacement rate of 8 [micro]m/s at room temperature. Thin film specimens were cut to have a dimension of 0.6 mm in width and 20 mm in overall length. Two sets of high-resolution thermogravimetric analyses (HR-TGA) were carried out on a TA Q600 HT, one from room temperature to 800[degrees]C at a heating rate of 2[degrees]C [min.sup.-1] under a nitrogen protection (50 mL [min.sup.-1]) and the other from room temperature to 600[degrees]C at a heating rate of 2[degrees]C min- in air (50 mL [min.sup.-1]).
A universal in situ tensile testing platform was accommodated and performed inside the FIB/FE-SEM. This setup contains three piezo stages (x-, y-, and z-direction) from Super C Inc. (Delaware). Each piezo stage has capacitive position feedback to enable accurate movement with a resolution down to 0.5 nm. Nanoscale dog-bone specimens of 20 [micro]m x 4 [micro]m in length and width and ~800 nm in thickness were prepared from the MWCNTs/PEEK nanocomposite with nanopatterning and visualization engine (NPVE, Zeiss) using Ga ion beam and transferred by a tungsten manipulator.
The tensile test specimen was then mounted between the force sensor and a silicon wafer fixture. Both of the force sensor and the silicon wafer (a nanoscale dog-bone holder) are fixed on the piezo stages as shown in Fig. 1, which exhibits a setting before the test. A nanosize dog-bone of MWCNTs/PEEK composite bridges between the force sensor and Si wafer fixture (insert in Fig. lb). The silicon wafer can move away from the force sensor at a speed of 0.1 [micro]m/s in this study, which applies a constant tension force to the specimen until fracture. During the test, video records the progress of the specimen elongation and meanwhile, force variation is obtained by a LabVIEW program. The cross-sectional area of fracture surface is imaged and measured with SEM and the displacement is extracted by the video analysis software. As a result, the tensile properties of the nanoscale specimen are determined.
Figure 2 presents the time-position curves of the reference and elongation points from recorded video. The reference and elongation points, as shown in Fig. 2a, are defined as the edge of silicon wafer and force sensor, respectively. It is assumed that both of the reference and elongation points move linearly (Fig. 2b). Therefore, the actual elongation of the dog-bone in each video frame is the difference of the reference and elongation points.
RESULTS AND DISCUSSION
PEEK is a high-performance thermoplastic polymer and its glass transition and melting temperatures are ~150[degrees]C and ~343[degrees]C, respectively. Figure 3 shows a typical TGA curve of the MWCNTs/PEEK composite in nitrogen, where the weight starts to drop at ~480[degrees]C and keeps up to 54% of the initial weight at 800[degrees]C due to the high temperature tolerance of ether and aromatic structures in the residual [4, 24, 25]. In air, however, the weight starts to drop at ~410[degrees]C while the flow temperature of PEEK is ~390[degrees]C, and the film is burned completely before 600[degrees]C. The first-order derivative of TGA curves shown in Fig. 3b exhibits the maximum rate of weight loss in nitrogen flow appears around 526[degrees]C, which is attributable to the decomposition of PEEK in nitrogen [4, 25]. In contrast, two peaks of maximum rate of weight loss were observed at 536[degrees]C and 565[degrees]C for the in-air experiment which represents the loss of PEEK and MWCNTs, respectively . TGA results indicate the MWCNTs/PEEK composite starts degrading at ~480[degrees]C in nitrogen and ~410[degrees]C in air. These temperatures are higher than that of the pure PEEK polymer due to the MWCNTs reinforcement, an enhancement of the composite thermal stability through the dispersion of MWCNTs, which may effectively transport at ballistic speed and dissipate heat uniformly throughout the composite matrix, a result consistent with other studies [4, 5, 26].
The fracture surfaces of pure PEEK and nanocomposite specimens are shown in Fig. 4. In contrast to a smooth and uniform fracture surface of pure PEEK in Fig. 4a, the MWCNTs/PEEK fracture surface shows significant amount of pull-out MWCNTs (arrows in Fig. 4b) and crack blockage in the composite. The crack deflection by the CNT pull-out mechanism improves the nanocomposite toughness resulting from an effect of more absorbance of the energy from uniaxial load [27, 28]. Meanwhile, relative clean pull-out MWCNTs with no residual PEEK attached, as confirmed by TEM in previous work by Tsuda et al. , suggests that the adhesion between MWCNTs and polymer matrix is probably poor, indicating a relatively less effective load transfer. Figure 4c shows a TEM image of the MWCNTs/PEEK nanocomposite. The arrows in Fig. 4c point to some dispersed MWCNTs in the PEEK matrix and overall, the MWCNTs are uniformly distributed with no significant bundling or agglomerations. Thus, the MWCNTs/PEEK composite mechanical properties are less likely affected by the CNTs agglomerations.
The strain-at-fracture of composite thin films is slightly reduced compared with that of the pure PEEK as shown in Fig. 5. The reduction of overall plastic deformation of composites is due to the fact that the fracture strain of individual MWCNTs (<10% ) is noticeably smaller than that of pure PEEK polymer. This CNT caused effect was also observed in similar CNT/PEEK systems for which the failure strain decreased with increasing CNT content . In addition, the strain values of both pure PEEK and MWCNTs/PEEK nanocomposite thin films at 272% and 227%, respectively, are dramatically increased compared to that of bulk composites ([epsilon] < 25%) [11,17]. The higher degree of freedom to deform parallel to the load direction for the thin-film samples suggests that the polymer chain in both pure PEEK and the nanocomposite is dominantly oriented perpendicularly to the thickness direction. Such a phenomenon was reported for the parylene N polymer thin film which demonstrated an increase of the ultimate elongation by 27.5% as the thin film thickness decreased from 25.4 [micro]m to 0.945 [micro]m .
The average tensile strengths of the MWCNTs/PEEK nano-composite and the pure PEEK are measured to be 133.71 MPa and 88.45 MPa, respectively, and the tensile strength enhancement from the dispersion of 6.5 wt% MWCNTs in the matrix reaches 51.2%, indicating the MWCNTs are capable of carrying and transferring load from matrix in small-scale samples. However, consistent with general observations [15, 29, 31-33], the experimentally measured tensile strength is much lower than that of the theoretical calculation due to the effects of structural and processing defects such as the discontinuity of CNTs or weak CNT/PEEK interfacial shear strength. Among the unique advantages of our experimental setup, the effect of specimen size on the tensile properties was systematically investigated for the MWCNTs/PEEK composite.
The SEM images in Fig. 6 show a sequence of tensile deformation of the composite during in situ test as follows:
(1) A crack initiates from the specimen surface within the gauge area (pointed by arrow in Fig. 6a).
(2) Crack grows roughly perpendicular to the force direction and crack deflections occur (Fig. 6b).
(3) MWCNTs bridge the crack (Fig. 6c).
(4) MWCNTs pull-out (Fig. 6d).
The CNT pull-out and crack deflection represent the two major strengthening mechanisms in MWCNTs/PEEK composite, confirming the similar observation reported in the literature .
Figure 7 shows a typical stress-strain curve of in situ tests of MWCNTs/PEEK composites. It is worth noticing that, unlike macro-scaled and microscaled specimens undergoing plastic deformation after yield point, the nanosized dog-bone specimens show no deformation and fail in a brittle manner with an average fracture strain of 6.04%. Typically, polymer materials are not well ordered and contain a large amount of amorphous regions and free space between chains resulting in permanent distortion when subjected to uniaxial loading. However, as shown in Fig. 6, necking is not observed on MWCNTs/PEEK composites nanoscale specimen, suggesting that the reduced specimen volume constrains the motion of polymer chains and the amorphous region . In addition, specimens of dramatically reduced volume may exhibit lower defects density necessaiy for the plastic deformation [5, 27].
Table 1 summarizes tensile strength of 6.5 wt% MWCNTs/PEEK composites with respect to different specimen thicknesses. The average strength and Young's modulus increase by 280% and 23%, respectively, as the specimen thickness decreases from 4 mm to 800 nm. The improvement of strength and modulus due to the size effect suggests that smaller specimen with lower defect distribution leads to an enhancement of the tensile strength, consistent with other studies [35, 36]. For the MWCNTs/PEEK composites, the defects are considered to be voids, MWCNT ends, and weak nanofiller-matrix interfacial bonding.
The uniform dispersion of MWCNTs in PEEK polymer matrix was achieved such that the MWCNTs were scattered and individual MWCNTs separated from each other. No CNT agglomeration was observed. Typical TGA curve showed the composite weight began to drop at 480[degrees]C in nitrogen and at 410[degrees]C in the air. Two peaks of maximum rate of weight loss were observed in the air environment, corresponding to the loss of PEEK and MWCNTs, respectively. Microscale tensile strength of the MWCNTs/PEEK thin film is 51% higher than that of unreinforced PEEK. The fracture strain of the thin film composites and PEEK is dramatically increased in microscale specimens. MWCNTs/PEEK nanocomposite exhibits typical deformation stages including crack initiation, crack propagation, MWCNTs bridging, and MWCNTs pull-out. In the nanoscale, the semicrystalline PEEK nanocomposites fracture in brittle mode likely due to a decrease of defects of critical size and the lack of polymer chain motion. The average tensile strength of the nano-scale specimens is about three times higher than that of the bulk composite, suggesting a strong size dependency of the property. This in situ SEM mechanical testing method represents a promising general capability for probing properties of microstructures or nanostructures.
Authors gratefully acknowledge the instrument support from the W. M. Keck Center for Advanced Microscopy and Microanalysis at the University of Delaware.
REFERENCES AND CITED WORK
[1.] B. Peng, Y. Jiang, and A. Zhu, Polym. Test., 74, 72 (2019). https://doi.org/10.1016/j.polymertesting.2018.12.018.
[2.] R.B. Mathur, S. Pande, B.P. Singh, and T.L. Dhami, Polym. Compos., 29, 717 (2008). https://doi.org/10.1002/pc.20449.
[3.] J. Sandler, P. Werner, M.S.P. Shaffer, V. Demchuk, V. Altstadt, and A.H. Windle, Compos. Part A Appl. Sci. Manuf., 33, 1033 (2002). https://doi.org/10.1016/S1359-835X(02)00084-2.
[4.] A.M. Dfez-Pascual, M. Naffakh, M.A. Gomez, C. Marco, G. Ellis, M.T. Martinez, A. Anson, J.M. Gonzalez-Domfnguez, Y. Martfnez-Rubi, and B. Simard, Carbon N. Y., 47, 3079 (2009). https://doi.org/10.1016/j.carbon.2009.07.020.
[5.] C. Rong, G. Ma, S. Zhang, L. Song, Z. Chen, G. Wang, and P. M. Ajayan, Compos. Sci. Technol., 70, 380 (2010). https://doi. org/10.1016/j.compscitech.2009.11.024.
[6.] Y. Shang, X. Wu, Y. Liu, Z. Jiang, Z. Wang, Z. Jiang, and H. Zhang, High Perform. Polym., 31, 43 (2018). https://doi.org/ 10.1177/0954008317750726.
[7.] J.C. Moses, A. Gangrade, and B.B. Mandal, "Chapter 5 Carbon Nanotubes and Their Polymer Nanocomposites," in Nanomaterials and Polymer Nanocomposite, P.N. Karak, Ed., Elsevier, Amsterdam, 145 (2019).
[8.] D. Jones, D. Leach, and D. Moore, Polymer (Guildf)., 26, 1385 (1985). https://doi.org/10.1016/0032-3861 (85)90316-7.
[9.] Z.A. Znc, Polym. Eng. Sci., 25, 474 (1985).
[10.] P. Cebe, S.Y. Chung, and S.-D. Hong, J. Appl. Polym. Sci., 33, 487 (1987). https://doi.org/10.1002/app.1987.070330217.
[11.] R. Davis, Y. Shyng, Y. Wang, and O. Ghita, "Extrusion deposition of carbon nanotubes (CNT)/poly ether ether ketone (PEEK)," in Proceedings of the 20th International Conference on Composite Mater (2015).
[12.] T. Ogasawara, T. Tsuda, and N. Takeda, Compos. Sci. Technol., 71, 73 (2011). https://doi.org/10.1016/j.compscitech.2010.10.001.
[13.] F. Deng, T. Ogasawara, and N. Takeda, Compos. Sci. Technol., 67, 2959 (2007). https://doi.org/10.1016/j.compscitech.2007.05.014.
[14.] W. Nix, Metall. Mater. Trans. A., 20, 2217 (1989). https://doi. org/10.1002/sca.20196.
[15.] J.C.H. Affdl and J.L. Kardos, Polym. Eng. Sci., 16, 344 (1976). https://doi.org/l 0.1002/pen.760160512.
[16.] M. Yu, Science, 287, 637 (2000). https://doi.org/10.1126/science. 287.5453.637.
[17.] A.M. Dfez-Pascual, G. Martinez, M.T. Martinez, and M. A. Gomez, J. Mater. Chem., 20, 8247 (2010). https://doi.org/10. 1039/c0jm01531h.
[18.] D.T. Read, J. Test. Eval., 26. 255 (1998).
[19.] R.W. Hoffman, MRS Proc., 130, 295 (1988). https://doi.org/10. 1557/PROC-130-295.
[20.] M.A. Haque and M.T.A. Saif, Exp. Mech., 42, 123 (2002). https://doi.org/10.1177/0018512002042001797.
[21.] X. Li, H. Gao, W.A. Scrivens, D. Fei, X. Xu, M.A. Sutton, A. P. Reynolds, and M.L. Myrick, Nanotechnology, 15, 1416 (2004). https://doi.org/10.1088/0957-4484/15/l 1/005.
[22.] A.V. Desai and M.A. Haque, Thin-Walled Struct., 43, 1787 (2005). https://doi.org/10.1016/j.tws.2005.07.003.
[23.] S. Roozpeikar and A.M. Fattahi, SN Appl. Sci., 1, 17 (2018). https://doi.org/10.1007/s42452-018-0022-y.
[24.] H. Wang, G. Wang, W. Li, Q. Wang, W. Wei, Z. Jiang, and S. Zhang, J. Mater. Chem., 22, 21232 (2012). https://doi.org/10. 1039/c2jm35129c.
[25.] M. Naffakh, A.M. Di'ez-Pascual, and M.A. Gomez-Fatou, J. Mater. Chem., 21, 7425 (2011). https://doi.org/10.1039/cljml0441a.
[26.] S. Samie, A. Nejati, M.R. Avazfard, and S. Amini, AIP Conf. Proc., 1779, 040009 (2016). https://doi.org/10.1063/L4965500.
[27.] V. Mirjalili and P. Hubert, Compos. Sci. Technol., 70, 1537 (2010). https://doi.org/10.1016/j.compscitech.2010.05.016.
[28.] F.H. Gojny, M.H.G. Wichmann, B. Fiedler, and K. Schulte, Compos. Sci. Technol., 65, 2300 (2005). https://doi.org/10.1016/j. compscitech.2005.04.021.
[29.] T. Tsuda, T. Ogasawara, F. Deng, and N. Takeda, Compos. Sci. Technol., 71, 1295 (2011). https://doi.org/10.1016/j.compscitech. 2011.04.014.
[30.] M.A. Spivack, Rev. Sci. Instrum., 43, 985 (1972). https://doi.org/ 10.1063/1.1685843.
[31.] J.N. Coleman, U. Khan, W.J. Blau, and Y.K. Gun'ko, Carbon N. Y., 44, 1624 (2006). https://doi.org/10.1016/jxarbon.2006. 02.038.
[32.] V. Mittal, Polymer Nanotubes Nanocomposites: Synthesis, Properties and Applications, Scrivener Publishing LLC, Salem, Massachusetts (2010). https://doi.org/10.1002/9780470 905647.
[33.] F. Deng, T. Ogasawara, and N. Takada, Key Eng. Mater., 334-335, 721 (2007). https://doi.org/10.4028/www.scientific.net/ KEM.334-335.721.
[34.] E.T. Thostenson and T.-W. Chou, J. Phys. D. Appl. Phys., 35, L77 (2002). https://doi.org/10.1088/0022-3727/35/16/103.
[35.] Z.P. Bazant, Arch. Appl. Mech., 69, 703 (1999). https://doi.org/ 10.1007/s004190050252.
[36.] Z.P. Bazant, Int. J. Solids Struct., 37, 69 (2000). https://doi.org/ 10.1016/S0020-7683(99)00077-3.
Chun-Yen Hsu(ID), (1,2,3) Kathryn Scrafford, (1) Chaoying Ni (ID), (3) Fei Deng (1,2)
(1) Super C Inc., 1352 Marrows Road, Newark, Delaware, 19711
(2) Shenzhen CONE Technology Co., Ltd., 1201, National Engineering Laboratory Building B, No. 20 Gaoxin S. Seventh Road, Shenzhen 518057, China
(3) Department of Materials Science and Engineering, University of Delaware, Newark, Delaware, 19711
Correspondence to: F. Deng; e-mail: firstname.lastname@example.org
Published online in Wiley Online Library (wileyonlinelibrary.com).
Caption: FIG. 1. (a) In situ tensile testing setup including a universal piezo stage, force sensor, and silicon wafer as a specimen holder, (b) SEM images before the test showing tensile specimen bridging between force sensor and silicon holder.
Caption: FIG. 2. (a) SEM image of MWCNTs/PEEK specimen from recorded video and (b) time--position curve of the reference and elongation points.
Caption: FIG. 3. (a) HR-TGA measurements of MWCNTs/PEEK composites in nitrogen and in air and (b) the first-order derivative of weight loss curves in (a).
Caption: FIG. 4. Fracture surface of (a) pure PEEK, (b) 6.5 wt% MWCNTs/PEEK thin film and (c) TEM image of MWCNTs/PEEK nanocomposite.
Caption: FIG. 5. Stress-strain curve of pure PEEK and composite thin films.
Caption: FIG. 6. Deformation process of in situ tensile test of MWCNTs/PEEK composite. 1212 POLYMER ENGINEERING AND SCIENCE-2019
Caption: FIG. 7. Stress-strain curve of nanoscale MWCNTs/PEEK composites.
TABLE 1. Tensile properties of 6.5 wt% MWCNTs/PEEK with different specimen thickness. Specimen scale Specimen Tensile Fracture Young's thickness strength strain (%) modulus (mm) (MPa) (GPa) Bulk  4.0 102.15 12.49 5.32 Nanoscale I 7.64 x [10.sup.-4] 369.17 4.78 7.36 Nanoscale II 7.75 x [10.sup.-4] 379.64 6.33 6.20 Nanoscale III 8.50 x [10.sup.-4] 417.17 7.02 5.99
|Printer friendly Cite/link Email Feedback|
|Author:||Hsu, Chun-Yen; Scrafford, Kathryn; Ni, Chaoying; Deng, Fei|
|Publication:||Polymer Engineering and Science|
|Date:||Jun 1, 2019|
|Previous Article:||Effect of Graphene Oxide on the Interlaminar Fracture Toughness of Carbon Fiber/Epoxy Composites.|
|Next Article:||Hybrid-Permeability Model Evaluation Through Concepts of Tortuosity and Resistance Rate: Properties of Manufactured Hybrid Laminate.|