- Email: [email protected]
Characteristics of hydrogen plasma treated carbon nanotubes and their influence on the mechanical properties of polyetherimide-based nanocomposites
Carbon 118 (2017) 650e658
Contents lists available at ScienceDirect
Carbon journal homepage: www.elsevier.com/locate/carbon
Characteristics of hydrogen plasma treated carbon nanotubes and their influence on the mechanical properties of polyetherimide-based nanocomposites Eung-seok Lee, Young-Kyun Lim, Yoon-soo Chun, Byung-Yong Wang, Dae-Soon Lim* Department of Materials Science and Engineering, Korea University, 1-5-ga, Anam-dong, Sungbuk-gu, Seoul, 136-701, South Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 November 2016 Received in revised form 31 March 2017 Accepted 2 April 2017 Available online 8 April 2017
The structural phase transformation and etching of the sidewalls of multi-walled carbon nanotubes (MWCNTs) by hydrogen plasma and the effect on the dispersion and mechanical properties of polyetherimide (PEI) composites was investigated. Surface-modified MWCNTs with various plasma treatment times were characterized using FESEM, HRTEM, Raman spectroscopy, FT-IR, and XPS. The results showed that plasma treatment facilitates the attachment of functional groups while damaging and detaching the sidewalls of the MWCNTs and modifying the crystallinity of the graphene layer. The mechanical properties of the PEI composites with the modified MWCNTs depended on the degree of damage to the surface and crystallinity modification of the MWCNTs. Hydrogen plasma treatment led to a significant improvement in the dispersion and mechanical properties of PEI/MWCNT composites due to the surface modification of the MWCNTs. © 2017 Elsevier Ltd. All rights reserved.
1. Introduction Carbon nanotubes (CNTs) have attracted intensive attention as a reinforcement material in many engineering applications due to their high aspect ratio and good electrical, thermal and physical properties [1e3]. These properties make CNTs suitable for a broad range of applications. High-strength epoxy composite systems made of CNTs and polymer are important materials for vehicle, machine, and electrical parts as well as countless other industrial applications [4e6]. In particular, the excellent mechanical properties of CNTs make them a promising candidate for added fillers in polymer composites. Nevertheless, because of the strong intrinsic van der Waals forces, CNTs tend to aggregate, which leads to poor dispersion and chemically inert materials; thus, good adhesion with a polymer matrix is hard to achieve. Thus, an applied load cannot be efficiently transferred from the polymer matrix to the CNTs because of this weak adhesion. To solve the problem of poor CNT dispersions, many research efforts have been devote to developing a chemical or mechanical dispersion method, such as CNT dispersion by attrition milling, adding surfactant, polymeric passivation and chemically functionalizing the surface of the CNTs
* Corresponding author. E-mail address: [email protected] (D.-S. Lim). http://dx.doi.org/10.1016/j.carbon.2017.04.003 0008-6223/© 2017 Elsevier Ltd. All rights reserved.
[7e9]. In a previous study on the chemical treatment of CNT surfaces, Ruan et al. [10] used magnetic stirring and sonication to disperse the nanotubes in xylene followed by reflux to mix the CNTs into a polymer. However, this method relies on the efficient dispersion of CNTs in the relevant solvent, while CNTs cannot be well dispersed in most solvents. In addition, treatment with a mixture of concentrated sulfuric and nitric acids [11] can introduce carboxyl groups onto the CNTs' surfaces to enhance interfacial bonding. However, this treatment was found to seriously damage the structure of the CNTs and reduces the length of the CNTs, which led to a degradation in the properties of the CNTs [12]. In the case of mechanically dispersed CNTs, Thostenson et al. [13] dispersed CNTs well into epoxy using shear mixing (roll to roll), and the results showed a 60% improvement in the thermal conductivity epoxy/CNT (5 wt%) composites. However, the experimental results showed weak interfacial bonding between the CNT surfaces and the polymer matrix. Plasma treatment is an environmentally friendly and costeffective method, and it provides a highly cross-linked polymer composite while not significantly affecting the original structure of CNTs. Shi and coworkers [14] used plasma to deposit polystyrene on the surface of CNTs, and their results show that these CNTs can be dispersed well into a polymer matrix. Surface modifying the CNT
E.-s. Lee et al. / Carbon 118 (2017) 650e658
surfaces by plasma coating can significantly lower their surface energy. To take advantage of the high strength of carbon nanotubes in a polymer matrix, achieving both a good dispersion and sufficient CNT/polymer interfacial bonding is critical [15,16]. Some plasma treatments have been proposed to avoid agglomeration, including chemical treatments, controlled oxidation, etching, polymer wrapping/absorption and the adsorption of functional groups. Moreover, plasma treatment has been used to reduce agglomerations by modifying the surface of the CNTs [17,18]. Plasma treatment generally contains functional groups that can electrostatically prevent agglomeration and enhance interactions with the polymer matrix. Various plasma methods have been used for the surface modification and surface activation/grafting of CNTs [19]. Although the effect of plasma treatment on polymer composites has been studied experimentally, few details are known about the mechanism of the hydrogen plasma effect with various gases on polymer composites. Chen et al. reported that compare to the untreated MWCNTs the MWCNTs after Ar/O2 plasma treated had a very good dispersion in aqueous solution [20]. Lee et al. also showed that oxygen plasma exhibited excellent dispersibility [21]. Compared to oxygen plasma, hydrogen plasma was less effective for dispersion. Yang et al. reported that a novel method combining ultrasonication and hydrogen passivation to MWCNTs improved dispersion and mechanical properties [22]. Hydrogen plasma treatment could contribute dispersing of MWCNTs and mechanical properties of composites due to the hydrogen binding to MWCNTs and modified sp3 bonding [23]. In spite of expectation, there are no reports on plasma treated MWCNTs for improvement of mechanical properties of MWCNTs reinforced polymer composites. In this study, the effect of hydrogen plasma treatment on the surface properties of MWCNTs was investigated. The generation of nanocrystalline particles on the MWCNT sidewalls, the enhancement of the dispersion of MWCNTs in a polymer matrix and the interfacial bonding between the MWCNTs and the polymer matrix were also investigated using hydrogen plasma treatment on MWCNTs. 2. Experimental details The MWCNTs (Applied carbon nano technology Co., Ltd. Republic of Korea) were used as a filler to fabricate a polymer composite in this study. First, a plasma treatment was performed on these MWCNT specimens using PECVD. The pressure within the vacuum chamber was maintained at 128 Pa, and the temperature was raised to 300 C. As the temperature reached 300 C, 10 standards cc/min (sccm) of hydrogen gas was injected into the chamber as a plasma generating source, and 450 W of radio frequency (RF) plasma power was applied to a 13.56 MHz RF generator to generate the hydrogen plasma, which irradiated the MWCNTs. The plasma exposure time was set to 30, 60, 90 and 120 min. The microstructure of the MWCNTs was characterized by Raman spectroscopy (LabRam HR, Jobin-Yvon, France) equipped with an Ar ion laser (l ¼ 532 nm). The binding energy was analyzed via X-ray photoelectron spectroscopy (XPS, Ulvac PHI, Inc.) using a 15 kV/ 25 W monochromated Al Ka X-ray source and a take-off angle of 45 . The Raman and XPS spectra were deconvoluted into peaks using spectral analysis software (PeakFit v4.06) to calculate the intensity ratio. FT-IR spectra of the MWCNTs were recorded with KBr in the range of 4000e500 cm1 using a FT-IR Spectrometer (LabRam ARAMIS IR2). High-resolution transmission electron microscopy (HRTEM) was performed using a Titan™ 80e300 microscope (FEI, USA). The cross-section of the polymer composite was observed using a field-emission scanning electron microscope (FESEM, Hitachi- S-4700).
651
Both untreated MWCNTs and plasma-treated MWCNTs were used to fabricate polymer composites at 260 C using a Bench Kneader (PBV-01K, Irie shokai Co., Ltd. Japan). PEI (Ultem® 1000) was supplied by GE Plastics. The samples for the mechanical testing were fabricated using a hot-pressing method. For the mechanical property tests, Instron universal testing equipment (UTM 3367) was used. Test specimen dimensions were in accordance with the respective ASTM standards (D638). Each value obtained represented the average of six samples. 3. Results and discussion Fig. 1 shows the FT-IR measurements of the plasma-treated MWCNTs as a function of the processing temperature. The sp3 CH (2850 cm1) and a CH2 asymmetric vibration mode (2920 cm1) appear in the spectra of MWCNTs treated at 300 C. The sp2 CeH related peaks (3031 cm1) are also observed in the spectrum of untreated MWCNTs. This sp2 C-H peak was shifted toward the sp3 CH related peaks as the process temperature increased to 300 C. The FT-IR spectrum of the MWCNTs plasma treated at 300 C indicated the generation of C-H functional groups, the processing temperature was fixed at 300 C. The effect of the plasma treatment on the MWCNT sidewalls was further investigated by HRTEM. Fig. 2 shows HRTEM images of the untreated and plasma-treated MWCNTs. After a 30 min treatment, the surface of MWCNT becomes rougher, probably due to the local etching of carbon bonds compared to the untreated sample, as shown in Fig. 2(a). After a 60 min treatment, nanoparticles consisting of nanocrystalline diamond (NCD) particles appear on the exterior of the MWCNTs (Fig. 2 (c)). The NCD phase formed with a d-spacing of 0.206 nm (Fig. 2 (f)). Several groups have reported that hydrogen plasma plays an important role in the transformation from CNTs to NCD [23,24]. These NCD particles could be easily removed by prolonging the plasma treatment (Fig. 2 (d)). After longer treatments, the number of CNT layers tends to decrease, probably due to layer-by-layer etching, as shown in Fig. 2 (e). The effect of the plasma exposure time on the surface of the MWCNTs was studied by Raman spectroscopy. The Raman spectra of the MWCNTs after various plasma exposure times are shown in Fig. 3. The conventional carbon peak shape was observed for all
Fig. 1. Functionalization of the MWCNTs surface: FT-IR spectra of untreated MWCNTs and MWCNTs treated with hydrogen plasma at room temperature and 300 C. (A colour version of this figure can be viewed online.)
652
E.-s. Lee et al. / Carbon 118 (2017) 650e658
Fig. 2. HRTEM images of the MWCNTs' surfaces with and without plasma treatment; (a) untreated MWCNTs and (b)e(e) MWCNTs plasma treated for various times. (A colour version of this figure can be viewed online.)
both the untreated MWCNTs and the plasma-treated MWCNTs, the D and G peaks became narrower with increasing plasma exposure time. The ID/IG and ID,total/IG ratio was also affected by the plasma exposure time, as shown in Fig. 4 (a). The sharp increase in ID/IG with the plasma treatment time up to 60 min indicates that the generation of defect sites increased upon plasma treatment. However, for plasma treatments longer than 60 min, ID/IG decreased. The average in-plane length (La) and averaged continuous graphene length including tortuosity (Leq) of both untreated MWCNTs and plasma-treated MWCNTs are displayed in Fig. 4 (b).
Fig. 3. Variation of the Raman spectra obtained for hydrogen-plasma-treated MWCNTs as a function of plasma treatment time. (A colour version of this figure can be viewed online.)
Raman spectra of the MWCNTs. The main peaks appear at 13321343 cm1 and 1571-1575 cm1. The band at 1332-1343 cm1 is assigned to the sp3-bonded carbon in disordered carbon (D band) and the broad peak at approximately 1571e1575 cm1 is attributed to the sp2-bonded carbon (G band) [25]. The MWCNTs with different plasma exposure times show both disordered-carbon D and graphitic-carbon G peaks. The ratio of the peak intensities of the D and G bands is a measure of the amount of graphitic impurities in the MWCNT sidewalls [26]. The D and G bands and the integrated intensities of the D (ID) and G (IG) bands were measured for all MWCNTs. In the spectra of
La ðnmÞ ¼ 4:4ðID =IG Þ1
(1)
Leq ðnmÞ ¼ 8:8ðI2D =ID Þ
(2)
The La and Leq of the plasma treated MWCNTs were slightly decreased, but after 60 min of plasma treatment La was increased from 1.08 to 1.2 nm and Leq was increased from 3.27 to 3.79 nm as the plasma exposure time increased. This means the size of graphite at MWCNTs side wall was decreased. Full width at half maximum (FWHM) of D band indicates the amount of amorphous and sp3-bonded carbon and their intensities determine the degree of order of graphitic structures. As shown in Fig. 4 (c), the FWHM of D was highest at the plasma exposure time of 60 min, and this result is a good agreement with the previous results. The results also showed the variation in the G peak position as a function of hydrogen plasma treatment time. The G peak position changed to a lower wavelength with treatment times up to 60 min and then held steady up to 90 min. Then, the G peak position shifted up again, as shown in Fig. 4 (d). This implies the maximum creation of sp3 bonds at approximately 60 min. High-resolution XPS data were collected with a higher binding energy detected at the C1s peak near 285.3 eV to determine the composition of the carbon species present in the MWCNTs after various plasma exposure times. To calculate the fraction of the sp3-
E.-s. Lee et al. / Carbon 118 (2017) 650e658
653
Fig. 4. Analysis of the Raman spectra peaks as a function of treatment time shown in Fig. 3; (a) ID/IG and ID,total/IG ratio, (b) Leq and La of MWCNTs, (c) FWHM of the G, D, and D'peaks and (d) the G peak position. (A colour version of this figure can be viewed online.)
bonded carbon of untreated MWCNTs and plasma-treated MWCNTs, the C1s peaks of the samples were fitted into sp2bonded carbon (284.5 eV) and sp3-bonded carbon (285 eV) components by Gaussian and Lorentzian functions [27,28]. The XPS measurement data along with the deconvoluted peaks are shown in Fig. 5, and the calculated contents of the sp2 and sp3 bonded carbon as a function of plasma exposure time are shown in Fig. 5 (f). In the case of the untreated MWCNTs, the ratio of the sp2-bonded carbon is much higher than that of sp3-bonded carbon, as shown in Fig. 5 (a). However, for the plasma-treated MWCNTs shown in Fig. 5 (b)-(e), the amount of sp2-bonded carbon decreased and the amount of sp3-bonded carbon increased as the plasma exposure time increased up to 60 min, whereas the content of the sp2bonded carbon decreased from 69.9% to 52.3% and the content of the sp3-bonded carbon increased from 16.9% to 37.2%. However, for plasma exposure times longer than 60 min, the contents of the sp2 and sp3 bonds increased and decreased, respectively. These results showed the same trend as that of the previous Raman analysis shown in Fig. 4 (a) and (d). Additionally, minor peaks appeared with weak intensities at 286.5 eV and 288.6 eV, which might have stemmed from the CeO and C]O vibration formed at the surface of the samples due to the exposure to air. Base on the HRTEM, Raman and XPS results, hydrogen plasma treatment of the MWCNTs has the following effects, as shown in Fig. 6. First, surface defects are generated by the attachment of functional groups and the etching process (Fig. 6 (b)). The sp2
carbon bonds of the MWCNTs can change into sp3 carbon bonds due to the influence of the hydrogen plasma. Moreover, atomic hydrogen etches sp2 carbon bonds at a much higher rate than it etches sp3 carbon bonds. Thus, the stable sp3 phase can be converted into nanocrystalline diamond (NCD), as shown in Fig. 6 (c). The decreasing ID/IG and sp3 content after 60 min suggests the separation of the NCD particles (Fig. 6 (d)). The separation of the outer wall containing these defects promotes a decreasing MWCNT diameter, as shown in Fig. 6 (e) and by the trend in the Raman and XPS analysis. As shown in Fig. 7, MWCNTs that were hydrogen plasma treated for 60 min maintain their dispersion characteristics in ethanol after 24 h. In contrast, the samples treated for 30, 90 and 120 min all precipitated after 24 h. These results showed the same trend as the Raman and XPS analysis, which related to the amount of sp2 and sp3-bonded carbon. Note that after the plasma treatment, the surface energy of the MWCNTs increases due to other factors besides the generation of sp3-bonded carbon [29]. The surface energy pertains to hydrogen-bonding interactions, and as a result, functional groups are generated on the MWCNTs surface due to the hydrogen plasma treatment. The results showed that hydrogen plasma treatment for 60 min provide significant improvement in dispersibility. This behavior is related to the variation of sp3/sp2 ratio depending on plasma treatment. S.C. Lee et al. have shown the correlation between sp3/sp2 ratio and water contact angle in hydrogenated diamond like carbon film [30]. Similar behavior is
654
E.-s. Lee et al. / Carbon 118 (2017) 650e658
Fig. 5. Fitted high-resolution XPS data of carbon binding energies in MWCNTs with and without plasma treatment; (a) untreated MWCNTs and (b)e(e) plasma-treated MWCNTs for various times. (f) Comparison of the attributes of the MWCNTs surface as determined from the XPS analysis. (A colour version of this figure can be viewed online.)
E.-s. Lee et al. / Carbon 118 (2017) 650e658
655
Fig. 6. Schematic of the hydrogen plasma effect on the MWCNT surface. (A colour version of this figure can be viewed online.)
Fig. 7. Dispersion properties of MWCNTs in ethanol without and with plasma treatment after settling for 24 h. (A colour version of this figure can be viewed online.)
expected for hydrogen plasma treated MWCNTs. So, this chemical state of MWCNTs can change the wetting behavior of plasma treated MWCNTs. The reconstruction of the high sp3 carbon bonds could make the MWCNTs to hydrophilic and enhanced dispersion behavior. Fig. 8 (a) and (b) show the tensile test results of the untreated and plasma-treated MWCNTs composite for various plasma treatment time for 2 wt% concentration of MWCNTs in the polymer. As can be seen, increasing plasma treatment time to 60 min produces the maximum tensile strength.
Based on the previous results, MWCNTs with a plasma exposure time of 60 min showed the most suitable properties for a composite filler, and a PEI/MWCNT composite was fabricated with MWCNTs treated with these conditions. The mechanical properties of the PEI/MWCNT composite were measured for both untreated and plasma-treated MWCNTs. For the plasma-treated MWCNTs used in the mechanical measurement samples, the H2 flow rate was fixed to 10 sccm during the plasma treatment, which was expected to provide the best mechanical properties in the generated nanocrystalline particles. To investigate the effect of the plasma treatment, the contents of both the untreated MWCNTs and plasmatreated MWCNTs in the composite were 2 wt%. Fig. 9 shows the stress-strain curve results for PEI, the PEI/untreated MWCNT composite and the PEI/hydrogen plasma-treated MWCNT composite. The mechanical test results of Fig. 9 are summarized in Table 1 in detail. Table 1 shows that the tensile strength of the PEI was improved when hydrogen plasma-treated MWCNTs were added. The tensile strength of the PEI alone was 104.3 MPa, compared to 120.8 MPa for the PEI/plasma-treated MWCNT composite. However, the tensile strength of the PEI/untreated MWCNT composite decreased to 81 MPa, which is lower than that of the PEI alone. Moreover, the modulus of elasticity results showed different trends between the untreated and plasma-treated MWCNT composites, similar to those of the tensile strength case. The modulus of elasticity of PEI was 4.05 GPa, but that of the PEI/untreated MWCNT composite increased to 4.77 GPa for 2 wt% of untreated MWCNTs. On the other hand, the modulus of elasticity of the PEI/hydrogen plasma-treated MWCNT composite increased greatly to 6.71 GPa for the MWCNT content of 2 wt%. Overall, the modulus of elasticity
656
E.-s. Lee et al. / Carbon 118 (2017) 650e658
Fig. 8. Comparison of mechanical properties of PEI/untreated and PEI/plasma-treated MWCNTs composite at various treatment time: (a) stress-strain curves, (b) tensile strength. (A colour version of this figure can be viewed online.)
Fig. 9. Stress-strain curves comparing the PEI/untreated MWCNT composite and the PEI/plasma-treated MWCNT composite. (A colour version of this figure can be viewed online.)
treatment improved the adhesion and dispersion of the MWCNTs, which facilitates more interpenetration between the MWCNT surface and PEI matrix [31]. The hydrogen plasma treatment was found to enhance the tensile strength, modulus of elasticity and yield strength of the PEI/hydrogen plasma-treated MWCNTs, which could be caused by a heterogeneous plasma effect. The schematic illustration of the untreated MWCNTs and hydrogen plasma-treated MWCNTs interface composites is given in Fig. 10 (a) and (b). Plasma treatment created the surface functional groups and sp3 crystalline particles that increased the mechanical interlocking and the chemical bond strength. SEM images of fracture surfaces of adding untreated and plasma-treated MWCNTs to PEI matrix are shown in Fig. 11. Untreated MWCNTs were not dispersed in the PEI matrix as shown in Fig. 11 (a). The dispersion property of hydrogen plasma-treated MWCNTs was improved compared to untreated MWCNTs (Fig. 11 (b)). This is well consistent with dispersion behaviors of MWCNTs in the solvent shown in Fig. 7. The surface modifications of MWCNTs by hydrogen plasma treatment improve the CNTs/polymer matrix interaction and thus increased mechanical properties. 4. Conclusions
of the PEI/hydrogen plasma-treated MWCNT composite was improved by 40.6% compared to that of the PEI/untreated MWCNT composite. The yield strength measurement results showed the same trend as the modulus of elasticity. The yield strength of the PEI was 62.4 MPa and that of the PEI/untreated MWCNT composite was 71.81 MPa. For the PEI/hydrogen plasma-treated MWCNT composite, the yield strength was 98.13 MPa. The yield strength of the PEI/hydrogen plasma-treated MWCNT composite showed a 36.7% improvement compared to that of the PEI/untreated MWCNT composite. The chemical and mechanical reactions during plasma
PEI/MWCNT composites were fabricated using hydrogen plasma-treated MWCNTs to investigate the effect of the treatment on the mechanical properties of the polymer composite. The surface of the MWCNTs was modified by the hydrogen plasma treatment with varying plasma exposure times from 0 to 120 min. A Raman analysis of the plasma-treated MWCNTs indicated that the sp3-bond carbons increased as the hydrogen exposure time increased to 60 min. In addition, XPS measurement results showed the same trend, in which the contents of the sp3-bonded carbonrelated deconvoluted peaks increased and those of the sp2-
Table 1 Comparison of mechanical properties.
PEI PEI/Untreated MWCNTs (2 wt%) PEI/hydrogen plasma-treated MWCNTs (Plasma exposure time 60 min, 2 wt%) (±indicate standard deviation).
Tensile strength (MPa)
Modulus of elasticity (GPa)
Yield strength (MPa)
104.3 ± 10.2 81 ± 12.6 120.8 ± 6.4
4.05 ± 0.1 4.77 ± 0.1 6.71 ± 0.2
62.3 ± 5.1 71.8 ± 6.2 98.1 ± 2.7
E.-s. Lee et al. / Carbon 118 (2017) 650e658
657
Fig. 10. Schematic images of PEI/MWCNT composites comparing (a) the PEI/untreated MWCNT composite and (b) the PEI/plasma-treated MWCNT composite. (A colour version of this figure can be viewed online.)
Fig. 11. Cross-sectional SEM images of PEI/MWCNT composites comparing (a) the PEI/untreated MWCNT composite and (b) the PEI/plasma-treated MWCNT composite.
bonded carbon-related deconvoluted peaks decreased. These results demonstrated not only the modification of the carbon bonding structure but also the generation of carbon nanoparticles or other carbon clusters, which results in improved chemical and mechanical bonding strengths. Therefore, the mechanical properties of the PEI/MWCNT composites could be enhanced using hydrogen plasma-treated MWCNTs as the reinforcement material. Most mechanical properties of PEI/plasma-treated MWCNT composites, including the tensile strength, yield strength, modulus of elasticity and compressive stress, were improved as the MWCNT content increased up to 3 wt% due to the effect of the increased functional groups and the modified carbon bonding structure of the MWCNTs. The results showed that polymer composites with 2 wt% MWCNTs showed the best mechanical properties. Finally, the hydrogen plasma treatment enhanced the dispersion of MWCNTs due to the change in sp3 bonding, and at the same time, carbon nanoparticles were formed at the surface of MWCNTs, which improved the filler effect of the MWCNTs in the polymer composite.
[7]
[8]
[9] [10]
[11] [12]
[13]
[14]
[15]
References [1] R.S. Ruoff, D.C. Lorents, Mechanical and thermal-properties of carbon nanotubes, Carbon 33 (7) (1995) 925e930. [2] W.A. Deheer, A. Chatelain, D. Ugarte, A carbon nanotube field-emission electron source, Science 270 (5239) (1995) 1179e1180. [3] P.M. Ajayan, O. Stephan, P. Redlich, C. Colliex, Carbon nanotubes as removable templates for metal-oxide nanocomposites and nanostructures, Nature 375 (6532) (1995) 564e567. [4] M. Moniruzzaman, K.I. Winey, Polymer nanocomposites containing carbon nanotubes, Macromolecules 39 (16) (2006) 5194e5205. [5] J.N. Coleman, U. Khan, Y.K. Gun'ko, Mechanical reinforcement of polymers using carbon nanotubes, Adv. Mater. 18 (6) (2006) 689e706. [6] J.N. Coleman, U. Khan, W.J. Blau, Y.K. Gun'ko, Small but strong: a review of the
[16]
[17]
[18]
[19]
mechanical properties of carbon nanotube-polymer composites, Carbon 44 (9) (2006) 1624e1652. F. Hussain, M. Hojjati, M. Okamoto, R.E. Gorga, Review article: polymer-matrix nanocomposites, processing, manufacturing, and application: an overview, J. Compos Mater. 40 (17) (2006) 1511e1575. J. Zhao, G.Z. Piao, X. Wang, J. Lian, Z.B. Wang, H.Q. Hu, et al., Improved dispersion of PEG-functionalized carbon nanofibers in toluene, Mat. Sci. Eng. C-Bio S 29 (3) (2009) 742e745. J.L. Bahr, J.M. Tour, Covalent chemistry of single-wall carbon nanotubes, J. Mater Chem. 12 (7) (2002) 1952e1958. S.L. Ruan, P. Gao, X.G. Yang, T.X. Yu, Toughening high performance ultrahigh molecular weight polyethylene using multiwalled carbon nanotubes, Polymer 44 (19) (2003) 5643e5654. I.D. Rosca, F. Watari, M. Uo, T. Akaska, Oxidation of multiwalled carbon nanotubes by nitric acid, Carbon 43 (15) (2005) 3124e3131. C.B. Dong, A.S. Campell, R. Eldawud, G. Perhinschi, Y. Rojanasakul, C.Z. Dinu, Effects of acid treatment on structure, properties and biocompatibility of carbon nanotubes, Appl. Surf. Sci. 264 (2013) 261e268. E.T. Thostenson, T.W. Chou, Processing-structure-multi-functional property relationship in carbon nanotube/epoxy composites, Carbon 44 (14) (2006) 3022e3029. D.L. Shi, J. Lian, P. He, L.M. Wang, F. Xiao, L. Yang, et al., Plasma coating of carbon nanofibers for enhanced dispersion and interfacial bonding in polymer composites, Appl. Phys. Lett. 83 (25) (2003) 5301e5303. Q.D. Chen, L.M. Dai, M. Gao, S.M. Huang, A. Mau, Plasma activation of carbon nanotubes for chemical modification, J. Phys. Chem. B 105 (3) (2001) 618e622. Y. Gao, P. He, J. Lian, L.M. Wang, D. Qian, J. Zhao, et al., Improving the mechanical properties of polycarbonate nanocomposites with plasma-modified carbon nanofibers, J. Macromol. Sci. B 45 (2006) 671e679. C.A. Avila-Orta, V.J. Cruz-Delgado, M.G. Neira-Velazquez, E. Hernandez-Hernandez, M.G. Mendez-Padilla, F.J. Medellin-Rodriguez, Surface modification of carbon nanotubes with ethylene glycol plasma, Carbon 47 (8) (2009) 1916e1921. M.V. Naseh, A.A. Khodadadi, Y. Mortazavi, F. Pourfayaz, O. Alizadeh, M. Maghrebi, Fast and clean functionalization of carbon nanotubes by dielectric barrier discharge plasma in air compared to acid treatment, Carbon 48 (5) (2010) 1369e1379. X.M. Ren, D.D. Shao, G.X. Zhao, G.D. Sheng, J. Hu, S.T. Yang, et al., Plasma
658
[20]
[21]
[22] [23] [24] [25] [26]
E.-s. Lee et al. / Carbon 118 (2017) 650e658 induced multiwalled carbon nanotube grafted with 2-vinylpyridine for preconcentration of Pb(II) from aqueous solutions, Plasma Process Polym. 8 (7) (2011) 589e598. C.L. Chen, B. Liang, A. Ogino, X.K. Wang, M. Nagatsu, Oxygen functionalization of multiwall carbon nanotubes by microwave-excited surface-wave plasma treatment, J. Phys. Chem. C 113 (18) (2009) 7659e7665. J. Lee, D. Lim, W. Choi, S. Dimitrijev, Influence of various plasma treatment on the properties of carbon nanotubes for composite applications, J. Nanosci. Nanotechno 12 (2) (2012) 1507e1512. Y.C. Yang, Z.H. Xu, Z.W. Pan, X.D. Li, Hydrogen passivation induced dispersion of multi-walled carbon nanotubes, Adv. Mater 24 (7) (2012) 881. L.T. Sun, J.L. Gong, Z.Y. Zhu, D.Z. Zhu, S.X. He, Z.X. Wang, et al., Nanocrystalline diamond from carbon nanotubes, Appl. Phys. Lett. 84 (15) (2004) 2901e2903. L.T. Sun, J.L. Gong, D.Z. Zhu, Z.Y. Zhu, S.X. He, Diamond nanorods from carbon nanotubes, Adv. Mater 16 (20) (2004) 1849. A.C. Ferrari, J. Robertson, Interpretation of Raman spectra of disordered and amorphous carbon, Phys. Rev. B 61 (20) (2000) 14095e14107. M.G. Jeong, S.H. Yoon, Y.S. Chun, E.S. Lee, D.S. Lim, Effect of lattice structure of
[27]
[28]
[29]
[30]
[31]
silicon carbide on crystal formation of carbide-derived carbon, Carbon 79 (2014) 19e27. T.I.T. Okpalugo, P. Papakonstantinou, H. Murphy, J. McLaughlin, N.M.D. Brown, High resolution XPS characterization of chemical functionalised MWCNTs and SWCNTs, Carbon 43 (1) (2005) 153e161. V. Datsyuk, M. Kalyva, K. Papagelis, J. Parthenios, D. Tasis, A. Siokou, et al., Chemical oxidation of multiwalled carbon nanotubes, Carbon 46 (6) (2008) 833e840. P.C. Ma, S.Y. Mo, B.Z. Tang, J.K. Kim, Dispersion, interfacial interaction and reagglomeration of functionalized carbon nanotubes in epoxy composites, Carbon 48 (6) (2010) 1824e1834. S.C. Lee, F.C. Tai, C.H. Wei, Correlation between sp(2)/sp(3) ratio or hydrogen content and water contact angle in hydrogenated DLC film, Mater. Trans. 48 (9) (2007) 2534e2538. J.A. Kim, D.G. Seong, T.J. Kang, J.R. Youn, Effects of surface modification on rheological and mechanical properties of CNT/epoxy composites, Carbon 44 (10) (2006) 1898e1905.
Contents lists available at ScienceDirect
Carbon journal homepage: www.elsevier.com/locate/carbon
Characteristics of hydrogen plasma treated carbon nanotubes and their influence on the mechanical properties of polyetherimide-based nanocomposites Eung-seok Lee, Young-Kyun Lim, Yoon-soo Chun, Byung-Yong Wang, Dae-Soon Lim* Department of Materials Science and Engineering, Korea University, 1-5-ga, Anam-dong, Sungbuk-gu, Seoul, 136-701, South Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 November 2016 Received in revised form 31 March 2017 Accepted 2 April 2017 Available online 8 April 2017
The structural phase transformation and etching of the sidewalls of multi-walled carbon nanotubes (MWCNTs) by hydrogen plasma and the effect on the dispersion and mechanical properties of polyetherimide (PEI) composites was investigated. Surface-modified MWCNTs with various plasma treatment times were characterized using FESEM, HRTEM, Raman spectroscopy, FT-IR, and XPS. The results showed that plasma treatment facilitates the attachment of functional groups while damaging and detaching the sidewalls of the MWCNTs and modifying the crystallinity of the graphene layer. The mechanical properties of the PEI composites with the modified MWCNTs depended on the degree of damage to the surface and crystallinity modification of the MWCNTs. Hydrogen plasma treatment led to a significant improvement in the dispersion and mechanical properties of PEI/MWCNT composites due to the surface modification of the MWCNTs. © 2017 Elsevier Ltd. All rights reserved.
1. Introduction Carbon nanotubes (CNTs) have attracted intensive attention as a reinforcement material in many engineering applications due to their high aspect ratio and good electrical, thermal and physical properties [1e3]. These properties make CNTs suitable for a broad range of applications. High-strength epoxy composite systems made of CNTs and polymer are important materials for vehicle, machine, and electrical parts as well as countless other industrial applications [4e6]. In particular, the excellent mechanical properties of CNTs make them a promising candidate for added fillers in polymer composites. Nevertheless, because of the strong intrinsic van der Waals forces, CNTs tend to aggregate, which leads to poor dispersion and chemically inert materials; thus, good adhesion with a polymer matrix is hard to achieve. Thus, an applied load cannot be efficiently transferred from the polymer matrix to the CNTs because of this weak adhesion. To solve the problem of poor CNT dispersions, many research efforts have been devote to developing a chemical or mechanical dispersion method, such as CNT dispersion by attrition milling, adding surfactant, polymeric passivation and chemically functionalizing the surface of the CNTs
* Corresponding author. E-mail address: [email protected] (D.-S. Lim). http://dx.doi.org/10.1016/j.carbon.2017.04.003 0008-6223/© 2017 Elsevier Ltd. All rights reserved.
[7e9]. In a previous study on the chemical treatment of CNT surfaces, Ruan et al. [10] used magnetic stirring and sonication to disperse the nanotubes in xylene followed by reflux to mix the CNTs into a polymer. However, this method relies on the efficient dispersion of CNTs in the relevant solvent, while CNTs cannot be well dispersed in most solvents. In addition, treatment with a mixture of concentrated sulfuric and nitric acids [11] can introduce carboxyl groups onto the CNTs' surfaces to enhance interfacial bonding. However, this treatment was found to seriously damage the structure of the CNTs and reduces the length of the CNTs, which led to a degradation in the properties of the CNTs [12]. In the case of mechanically dispersed CNTs, Thostenson et al. [13] dispersed CNTs well into epoxy using shear mixing (roll to roll), and the results showed a 60% improvement in the thermal conductivity epoxy/CNT (5 wt%) composites. However, the experimental results showed weak interfacial bonding between the CNT surfaces and the polymer matrix. Plasma treatment is an environmentally friendly and costeffective method, and it provides a highly cross-linked polymer composite while not significantly affecting the original structure of CNTs. Shi and coworkers [14] used plasma to deposit polystyrene on the surface of CNTs, and their results show that these CNTs can be dispersed well into a polymer matrix. Surface modifying the CNT
E.-s. Lee et al. / Carbon 118 (2017) 650e658
surfaces by plasma coating can significantly lower their surface energy. To take advantage of the high strength of carbon nanotubes in a polymer matrix, achieving both a good dispersion and sufficient CNT/polymer interfacial bonding is critical [15,16]. Some plasma treatments have been proposed to avoid agglomeration, including chemical treatments, controlled oxidation, etching, polymer wrapping/absorption and the adsorption of functional groups. Moreover, plasma treatment has been used to reduce agglomerations by modifying the surface of the CNTs [17,18]. Plasma treatment generally contains functional groups that can electrostatically prevent agglomeration and enhance interactions with the polymer matrix. Various plasma methods have been used for the surface modification and surface activation/grafting of CNTs [19]. Although the effect of plasma treatment on polymer composites has been studied experimentally, few details are known about the mechanism of the hydrogen plasma effect with various gases on polymer composites. Chen et al. reported that compare to the untreated MWCNTs the MWCNTs after Ar/O2 plasma treated had a very good dispersion in aqueous solution [20]. Lee et al. also showed that oxygen plasma exhibited excellent dispersibility [21]. Compared to oxygen plasma, hydrogen plasma was less effective for dispersion. Yang et al. reported that a novel method combining ultrasonication and hydrogen passivation to MWCNTs improved dispersion and mechanical properties [22]. Hydrogen plasma treatment could contribute dispersing of MWCNTs and mechanical properties of composites due to the hydrogen binding to MWCNTs and modified sp3 bonding [23]. In spite of expectation, there are no reports on plasma treated MWCNTs for improvement of mechanical properties of MWCNTs reinforced polymer composites. In this study, the effect of hydrogen plasma treatment on the surface properties of MWCNTs was investigated. The generation of nanocrystalline particles on the MWCNT sidewalls, the enhancement of the dispersion of MWCNTs in a polymer matrix and the interfacial bonding between the MWCNTs and the polymer matrix were also investigated using hydrogen plasma treatment on MWCNTs. 2. Experimental details The MWCNTs (Applied carbon nano technology Co., Ltd. Republic of Korea) were used as a filler to fabricate a polymer composite in this study. First, a plasma treatment was performed on these MWCNT specimens using PECVD. The pressure within the vacuum chamber was maintained at 128 Pa, and the temperature was raised to 300 C. As the temperature reached 300 C, 10 standards cc/min (sccm) of hydrogen gas was injected into the chamber as a plasma generating source, and 450 W of radio frequency (RF) plasma power was applied to a 13.56 MHz RF generator to generate the hydrogen plasma, which irradiated the MWCNTs. The plasma exposure time was set to 30, 60, 90 and 120 min. The microstructure of the MWCNTs was characterized by Raman spectroscopy (LabRam HR, Jobin-Yvon, France) equipped with an Ar ion laser (l ¼ 532 nm). The binding energy was analyzed via X-ray photoelectron spectroscopy (XPS, Ulvac PHI, Inc.) using a 15 kV/ 25 W monochromated Al Ka X-ray source and a take-off angle of 45 . The Raman and XPS spectra were deconvoluted into peaks using spectral analysis software (PeakFit v4.06) to calculate the intensity ratio. FT-IR spectra of the MWCNTs were recorded with KBr in the range of 4000e500 cm1 using a FT-IR Spectrometer (LabRam ARAMIS IR2). High-resolution transmission electron microscopy (HRTEM) was performed using a Titan™ 80e300 microscope (FEI, USA). The cross-section of the polymer composite was observed using a field-emission scanning electron microscope (FESEM, Hitachi- S-4700).
651
Both untreated MWCNTs and plasma-treated MWCNTs were used to fabricate polymer composites at 260 C using a Bench Kneader (PBV-01K, Irie shokai Co., Ltd. Japan). PEI (Ultem® 1000) was supplied by GE Plastics. The samples for the mechanical testing were fabricated using a hot-pressing method. For the mechanical property tests, Instron universal testing equipment (UTM 3367) was used. Test specimen dimensions were in accordance with the respective ASTM standards (D638). Each value obtained represented the average of six samples. 3. Results and discussion Fig. 1 shows the FT-IR measurements of the plasma-treated MWCNTs as a function of the processing temperature. The sp3 CH (2850 cm1) and a CH2 asymmetric vibration mode (2920 cm1) appear in the spectra of MWCNTs treated at 300 C. The sp2 CeH related peaks (3031 cm1) are also observed in the spectrum of untreated MWCNTs. This sp2 C-H peak was shifted toward the sp3 CH related peaks as the process temperature increased to 300 C. The FT-IR spectrum of the MWCNTs plasma treated at 300 C indicated the generation of C-H functional groups, the processing temperature was fixed at 300 C. The effect of the plasma treatment on the MWCNT sidewalls was further investigated by HRTEM. Fig. 2 shows HRTEM images of the untreated and plasma-treated MWCNTs. After a 30 min treatment, the surface of MWCNT becomes rougher, probably due to the local etching of carbon bonds compared to the untreated sample, as shown in Fig. 2(a). After a 60 min treatment, nanoparticles consisting of nanocrystalline diamond (NCD) particles appear on the exterior of the MWCNTs (Fig. 2 (c)). The NCD phase formed with a d-spacing of 0.206 nm (Fig. 2 (f)). Several groups have reported that hydrogen plasma plays an important role in the transformation from CNTs to NCD [23,24]. These NCD particles could be easily removed by prolonging the plasma treatment (Fig. 2 (d)). After longer treatments, the number of CNT layers tends to decrease, probably due to layer-by-layer etching, as shown in Fig. 2 (e). The effect of the plasma exposure time on the surface of the MWCNTs was studied by Raman spectroscopy. The Raman spectra of the MWCNTs after various plasma exposure times are shown in Fig. 3. The conventional carbon peak shape was observed for all
Fig. 1. Functionalization of the MWCNTs surface: FT-IR spectra of untreated MWCNTs and MWCNTs treated with hydrogen plasma at room temperature and 300 C. (A colour version of this figure can be viewed online.)
652
E.-s. Lee et al. / Carbon 118 (2017) 650e658
Fig. 2. HRTEM images of the MWCNTs' surfaces with and without plasma treatment; (a) untreated MWCNTs and (b)e(e) MWCNTs plasma treated for various times. (A colour version of this figure can be viewed online.)
both the untreated MWCNTs and the plasma-treated MWCNTs, the D and G peaks became narrower with increasing plasma exposure time. The ID/IG and ID,total/IG ratio was also affected by the plasma exposure time, as shown in Fig. 4 (a). The sharp increase in ID/IG with the plasma treatment time up to 60 min indicates that the generation of defect sites increased upon plasma treatment. However, for plasma treatments longer than 60 min, ID/IG decreased. The average in-plane length (La) and averaged continuous graphene length including tortuosity (Leq) of both untreated MWCNTs and plasma-treated MWCNTs are displayed in Fig. 4 (b).
Fig. 3. Variation of the Raman spectra obtained for hydrogen-plasma-treated MWCNTs as a function of plasma treatment time. (A colour version of this figure can be viewed online.)
Raman spectra of the MWCNTs. The main peaks appear at 13321343 cm1 and 1571-1575 cm1. The band at 1332-1343 cm1 is assigned to the sp3-bonded carbon in disordered carbon (D band) and the broad peak at approximately 1571e1575 cm1 is attributed to the sp2-bonded carbon (G band) [25]. The MWCNTs with different plasma exposure times show both disordered-carbon D and graphitic-carbon G peaks. The ratio of the peak intensities of the D and G bands is a measure of the amount of graphitic impurities in the MWCNT sidewalls [26]. The D and G bands and the integrated intensities of the D (ID) and G (IG) bands were measured for all MWCNTs. In the spectra of
La ðnmÞ ¼ 4:4ðID =IG Þ1
(1)
Leq ðnmÞ ¼ 8:8ðI2D =ID Þ
(2)
The La and Leq of the plasma treated MWCNTs were slightly decreased, but after 60 min of plasma treatment La was increased from 1.08 to 1.2 nm and Leq was increased from 3.27 to 3.79 nm as the plasma exposure time increased. This means the size of graphite at MWCNTs side wall was decreased. Full width at half maximum (FWHM) of D band indicates the amount of amorphous and sp3-bonded carbon and their intensities determine the degree of order of graphitic structures. As shown in Fig. 4 (c), the FWHM of D was highest at the plasma exposure time of 60 min, and this result is a good agreement with the previous results. The results also showed the variation in the G peak position as a function of hydrogen plasma treatment time. The G peak position changed to a lower wavelength with treatment times up to 60 min and then held steady up to 90 min. Then, the G peak position shifted up again, as shown in Fig. 4 (d). This implies the maximum creation of sp3 bonds at approximately 60 min. High-resolution XPS data were collected with a higher binding energy detected at the C1s peak near 285.3 eV to determine the composition of the carbon species present in the MWCNTs after various plasma exposure times. To calculate the fraction of the sp3-
E.-s. Lee et al. / Carbon 118 (2017) 650e658
653
Fig. 4. Analysis of the Raman spectra peaks as a function of treatment time shown in Fig. 3; (a) ID/IG and ID,total/IG ratio, (b) Leq and La of MWCNTs, (c) FWHM of the G, D, and D'peaks and (d) the G peak position. (A colour version of this figure can be viewed online.)
bonded carbon of untreated MWCNTs and plasma-treated MWCNTs, the C1s peaks of the samples were fitted into sp2bonded carbon (284.5 eV) and sp3-bonded carbon (285 eV) components by Gaussian and Lorentzian functions [27,28]. The XPS measurement data along with the deconvoluted peaks are shown in Fig. 5, and the calculated contents of the sp2 and sp3 bonded carbon as a function of plasma exposure time are shown in Fig. 5 (f). In the case of the untreated MWCNTs, the ratio of the sp2-bonded carbon is much higher than that of sp3-bonded carbon, as shown in Fig. 5 (a). However, for the plasma-treated MWCNTs shown in Fig. 5 (b)-(e), the amount of sp2-bonded carbon decreased and the amount of sp3-bonded carbon increased as the plasma exposure time increased up to 60 min, whereas the content of the sp2bonded carbon decreased from 69.9% to 52.3% and the content of the sp3-bonded carbon increased from 16.9% to 37.2%. However, for plasma exposure times longer than 60 min, the contents of the sp2 and sp3 bonds increased and decreased, respectively. These results showed the same trend as that of the previous Raman analysis shown in Fig. 4 (a) and (d). Additionally, minor peaks appeared with weak intensities at 286.5 eV and 288.6 eV, which might have stemmed from the CeO and C]O vibration formed at the surface of the samples due to the exposure to air. Base on the HRTEM, Raman and XPS results, hydrogen plasma treatment of the MWCNTs has the following effects, as shown in Fig. 6. First, surface defects are generated by the attachment of functional groups and the etching process (Fig. 6 (b)). The sp2
carbon bonds of the MWCNTs can change into sp3 carbon bonds due to the influence of the hydrogen plasma. Moreover, atomic hydrogen etches sp2 carbon bonds at a much higher rate than it etches sp3 carbon bonds. Thus, the stable sp3 phase can be converted into nanocrystalline diamond (NCD), as shown in Fig. 6 (c). The decreasing ID/IG and sp3 content after 60 min suggests the separation of the NCD particles (Fig. 6 (d)). The separation of the outer wall containing these defects promotes a decreasing MWCNT diameter, as shown in Fig. 6 (e) and by the trend in the Raman and XPS analysis. As shown in Fig. 7, MWCNTs that were hydrogen plasma treated for 60 min maintain their dispersion characteristics in ethanol after 24 h. In contrast, the samples treated for 30, 90 and 120 min all precipitated after 24 h. These results showed the same trend as the Raman and XPS analysis, which related to the amount of sp2 and sp3-bonded carbon. Note that after the plasma treatment, the surface energy of the MWCNTs increases due to other factors besides the generation of sp3-bonded carbon [29]. The surface energy pertains to hydrogen-bonding interactions, and as a result, functional groups are generated on the MWCNTs surface due to the hydrogen plasma treatment. The results showed that hydrogen plasma treatment for 60 min provide significant improvement in dispersibility. This behavior is related to the variation of sp3/sp2 ratio depending on plasma treatment. S.C. Lee et al. have shown the correlation between sp3/sp2 ratio and water contact angle in hydrogenated diamond like carbon film [30]. Similar behavior is
654
E.-s. Lee et al. / Carbon 118 (2017) 650e658
Fig. 5. Fitted high-resolution XPS data of carbon binding energies in MWCNTs with and without plasma treatment; (a) untreated MWCNTs and (b)e(e) plasma-treated MWCNTs for various times. (f) Comparison of the attributes of the MWCNTs surface as determined from the XPS analysis. (A colour version of this figure can be viewed online.)
E.-s. Lee et al. / Carbon 118 (2017) 650e658
655
Fig. 6. Schematic of the hydrogen plasma effect on the MWCNT surface. (A colour version of this figure can be viewed online.)
Fig. 7. Dispersion properties of MWCNTs in ethanol without and with plasma treatment after settling for 24 h. (A colour version of this figure can be viewed online.)
expected for hydrogen plasma treated MWCNTs. So, this chemical state of MWCNTs can change the wetting behavior of plasma treated MWCNTs. The reconstruction of the high sp3 carbon bonds could make the MWCNTs to hydrophilic and enhanced dispersion behavior. Fig. 8 (a) and (b) show the tensile test results of the untreated and plasma-treated MWCNTs composite for various plasma treatment time for 2 wt% concentration of MWCNTs in the polymer. As can be seen, increasing plasma treatment time to 60 min produces the maximum tensile strength.
Based on the previous results, MWCNTs with a plasma exposure time of 60 min showed the most suitable properties for a composite filler, and a PEI/MWCNT composite was fabricated with MWCNTs treated with these conditions. The mechanical properties of the PEI/MWCNT composite were measured for both untreated and plasma-treated MWCNTs. For the plasma-treated MWCNTs used in the mechanical measurement samples, the H2 flow rate was fixed to 10 sccm during the plasma treatment, which was expected to provide the best mechanical properties in the generated nanocrystalline particles. To investigate the effect of the plasma treatment, the contents of both the untreated MWCNTs and plasmatreated MWCNTs in the composite were 2 wt%. Fig. 9 shows the stress-strain curve results for PEI, the PEI/untreated MWCNT composite and the PEI/hydrogen plasma-treated MWCNT composite. The mechanical test results of Fig. 9 are summarized in Table 1 in detail. Table 1 shows that the tensile strength of the PEI was improved when hydrogen plasma-treated MWCNTs were added. The tensile strength of the PEI alone was 104.3 MPa, compared to 120.8 MPa for the PEI/plasma-treated MWCNT composite. However, the tensile strength of the PEI/untreated MWCNT composite decreased to 81 MPa, which is lower than that of the PEI alone. Moreover, the modulus of elasticity results showed different trends between the untreated and plasma-treated MWCNT composites, similar to those of the tensile strength case. The modulus of elasticity of PEI was 4.05 GPa, but that of the PEI/untreated MWCNT composite increased to 4.77 GPa for 2 wt% of untreated MWCNTs. On the other hand, the modulus of elasticity of the PEI/hydrogen plasma-treated MWCNT composite increased greatly to 6.71 GPa for the MWCNT content of 2 wt%. Overall, the modulus of elasticity
656
E.-s. Lee et al. / Carbon 118 (2017) 650e658
Fig. 8. Comparison of mechanical properties of PEI/untreated and PEI/plasma-treated MWCNTs composite at various treatment time: (a) stress-strain curves, (b) tensile strength. (A colour version of this figure can be viewed online.)
Fig. 9. Stress-strain curves comparing the PEI/untreated MWCNT composite and the PEI/plasma-treated MWCNT composite. (A colour version of this figure can be viewed online.)
treatment improved the adhesion and dispersion of the MWCNTs, which facilitates more interpenetration between the MWCNT surface and PEI matrix [31]. The hydrogen plasma treatment was found to enhance the tensile strength, modulus of elasticity and yield strength of the PEI/hydrogen plasma-treated MWCNTs, which could be caused by a heterogeneous plasma effect. The schematic illustration of the untreated MWCNTs and hydrogen plasma-treated MWCNTs interface composites is given in Fig. 10 (a) and (b). Plasma treatment created the surface functional groups and sp3 crystalline particles that increased the mechanical interlocking and the chemical bond strength. SEM images of fracture surfaces of adding untreated and plasma-treated MWCNTs to PEI matrix are shown in Fig. 11. Untreated MWCNTs were not dispersed in the PEI matrix as shown in Fig. 11 (a). The dispersion property of hydrogen plasma-treated MWCNTs was improved compared to untreated MWCNTs (Fig. 11 (b)). This is well consistent with dispersion behaviors of MWCNTs in the solvent shown in Fig. 7. The surface modifications of MWCNTs by hydrogen plasma treatment improve the CNTs/polymer matrix interaction and thus increased mechanical properties. 4. Conclusions
of the PEI/hydrogen plasma-treated MWCNT composite was improved by 40.6% compared to that of the PEI/untreated MWCNT composite. The yield strength measurement results showed the same trend as the modulus of elasticity. The yield strength of the PEI was 62.4 MPa and that of the PEI/untreated MWCNT composite was 71.81 MPa. For the PEI/hydrogen plasma-treated MWCNT composite, the yield strength was 98.13 MPa. The yield strength of the PEI/hydrogen plasma-treated MWCNT composite showed a 36.7% improvement compared to that of the PEI/untreated MWCNT composite. The chemical and mechanical reactions during plasma
PEI/MWCNT composites were fabricated using hydrogen plasma-treated MWCNTs to investigate the effect of the treatment on the mechanical properties of the polymer composite. The surface of the MWCNTs was modified by the hydrogen plasma treatment with varying plasma exposure times from 0 to 120 min. A Raman analysis of the plasma-treated MWCNTs indicated that the sp3-bond carbons increased as the hydrogen exposure time increased to 60 min. In addition, XPS measurement results showed the same trend, in which the contents of the sp3-bonded carbonrelated deconvoluted peaks increased and those of the sp2-
Table 1 Comparison of mechanical properties.
PEI PEI/Untreated MWCNTs (2 wt%) PEI/hydrogen plasma-treated MWCNTs (Plasma exposure time 60 min, 2 wt%) (±indicate standard deviation).
Tensile strength (MPa)
Modulus of elasticity (GPa)
Yield strength (MPa)
104.3 ± 10.2 81 ± 12.6 120.8 ± 6.4
4.05 ± 0.1 4.77 ± 0.1 6.71 ± 0.2
62.3 ± 5.1 71.8 ± 6.2 98.1 ± 2.7
E.-s. Lee et al. / Carbon 118 (2017) 650e658
657
Fig. 10. Schematic images of PEI/MWCNT composites comparing (a) the PEI/untreated MWCNT composite and (b) the PEI/plasma-treated MWCNT composite. (A colour version of this figure can be viewed online.)
Fig. 11. Cross-sectional SEM images of PEI/MWCNT composites comparing (a) the PEI/untreated MWCNT composite and (b) the PEI/plasma-treated MWCNT composite.
bonded carbon-related deconvoluted peaks decreased. These results demonstrated not only the modification of the carbon bonding structure but also the generation of carbon nanoparticles or other carbon clusters, which results in improved chemical and mechanical bonding strengths. Therefore, the mechanical properties of the PEI/MWCNT composites could be enhanced using hydrogen plasma-treated MWCNTs as the reinforcement material. Most mechanical properties of PEI/plasma-treated MWCNT composites, including the tensile strength, yield strength, modulus of elasticity and compressive stress, were improved as the MWCNT content increased up to 3 wt% due to the effect of the increased functional groups and the modified carbon bonding structure of the MWCNTs. The results showed that polymer composites with 2 wt% MWCNTs showed the best mechanical properties. Finally, the hydrogen plasma treatment enhanced the dispersion of MWCNTs due to the change in sp3 bonding, and at the same time, carbon nanoparticles were formed at the surface of MWCNTs, which improved the filler effect of the MWCNTs in the polymer composite.
[7]
[8]
[9] [10]
[11] [12]
[13]
[14]
[15]
References [1] R.S. Ruoff, D.C. Lorents, Mechanical and thermal-properties of carbon nanotubes, Carbon 33 (7) (1995) 925e930. [2] W.A. Deheer, A. Chatelain, D. Ugarte, A carbon nanotube field-emission electron source, Science 270 (5239) (1995) 1179e1180. [3] P.M. Ajayan, O. Stephan, P. Redlich, C. Colliex, Carbon nanotubes as removable templates for metal-oxide nanocomposites and nanostructures, Nature 375 (6532) (1995) 564e567. [4] M. Moniruzzaman, K.I. Winey, Polymer nanocomposites containing carbon nanotubes, Macromolecules 39 (16) (2006) 5194e5205. [5] J.N. Coleman, U. Khan, Y.K. Gun'ko, Mechanical reinforcement of polymers using carbon nanotubes, Adv. Mater. 18 (6) (2006) 689e706. [6] J.N. Coleman, U. Khan, W.J. Blau, Y.K. Gun'ko, Small but strong: a review of the
[16]
[17]
[18]
[19]
mechanical properties of carbon nanotube-polymer composites, Carbon 44 (9) (2006) 1624e1652. F. Hussain, M. Hojjati, M. Okamoto, R.E. Gorga, Review article: polymer-matrix nanocomposites, processing, manufacturing, and application: an overview, J. Compos Mater. 40 (17) (2006) 1511e1575. J. Zhao, G.Z. Piao, X. Wang, J. Lian, Z.B. Wang, H.Q. Hu, et al., Improved dispersion of PEG-functionalized carbon nanofibers in toluene, Mat. Sci. Eng. C-Bio S 29 (3) (2009) 742e745. J.L. Bahr, J.M. Tour, Covalent chemistry of single-wall carbon nanotubes, J. Mater Chem. 12 (7) (2002) 1952e1958. S.L. Ruan, P. Gao, X.G. Yang, T.X. Yu, Toughening high performance ultrahigh molecular weight polyethylene using multiwalled carbon nanotubes, Polymer 44 (19) (2003) 5643e5654. I.D. Rosca, F. Watari, M. Uo, T. Akaska, Oxidation of multiwalled carbon nanotubes by nitric acid, Carbon 43 (15) (2005) 3124e3131. C.B. Dong, A.S. Campell, R. Eldawud, G. Perhinschi, Y. Rojanasakul, C.Z. Dinu, Effects of acid treatment on structure, properties and biocompatibility of carbon nanotubes, Appl. Surf. Sci. 264 (2013) 261e268. E.T. Thostenson, T.W. Chou, Processing-structure-multi-functional property relationship in carbon nanotube/epoxy composites, Carbon 44 (14) (2006) 3022e3029. D.L. Shi, J. Lian, P. He, L.M. Wang, F. Xiao, L. Yang, et al., Plasma coating of carbon nanofibers for enhanced dispersion and interfacial bonding in polymer composites, Appl. Phys. Lett. 83 (25) (2003) 5301e5303. Q.D. Chen, L.M. Dai, M. Gao, S.M. Huang, A. Mau, Plasma activation of carbon nanotubes for chemical modification, J. Phys. Chem. B 105 (3) (2001) 618e622. Y. Gao, P. He, J. Lian, L.M. Wang, D. Qian, J. Zhao, et al., Improving the mechanical properties of polycarbonate nanocomposites with plasma-modified carbon nanofibers, J. Macromol. Sci. B 45 (2006) 671e679. C.A. Avila-Orta, V.J. Cruz-Delgado, M.G. Neira-Velazquez, E. Hernandez-Hernandez, M.G. Mendez-Padilla, F.J. Medellin-Rodriguez, Surface modification of carbon nanotubes with ethylene glycol plasma, Carbon 47 (8) (2009) 1916e1921. M.V. Naseh, A.A. Khodadadi, Y. Mortazavi, F. Pourfayaz, O. Alizadeh, M. Maghrebi, Fast and clean functionalization of carbon nanotubes by dielectric barrier discharge plasma in air compared to acid treatment, Carbon 48 (5) (2010) 1369e1379. X.M. Ren, D.D. Shao, G.X. Zhao, G.D. Sheng, J. Hu, S.T. Yang, et al., Plasma
658
[20]
[21]
[22] [23] [24] [25] [26]
E.-s. Lee et al. / Carbon 118 (2017) 650e658 induced multiwalled carbon nanotube grafted with 2-vinylpyridine for preconcentration of Pb(II) from aqueous solutions, Plasma Process Polym. 8 (7) (2011) 589e598. C.L. Chen, B. Liang, A. Ogino, X.K. Wang, M. Nagatsu, Oxygen functionalization of multiwall carbon nanotubes by microwave-excited surface-wave plasma treatment, J. Phys. Chem. C 113 (18) (2009) 7659e7665. J. Lee, D. Lim, W. Choi, S. Dimitrijev, Influence of various plasma treatment on the properties of carbon nanotubes for composite applications, J. Nanosci. Nanotechno 12 (2) (2012) 1507e1512. Y.C. Yang, Z.H. Xu, Z.W. Pan, X.D. Li, Hydrogen passivation induced dispersion of multi-walled carbon nanotubes, Adv. Mater 24 (7) (2012) 881. L.T. Sun, J.L. Gong, Z.Y. Zhu, D.Z. Zhu, S.X. He, Z.X. Wang, et al., Nanocrystalline diamond from carbon nanotubes, Appl. Phys. Lett. 84 (15) (2004) 2901e2903. L.T. Sun, J.L. Gong, D.Z. Zhu, Z.Y. Zhu, S.X. He, Diamond nanorods from carbon nanotubes, Adv. Mater 16 (20) (2004) 1849. A.C. Ferrari, J. Robertson, Interpretation of Raman spectra of disordered and amorphous carbon, Phys. Rev. B 61 (20) (2000) 14095e14107. M.G. Jeong, S.H. Yoon, Y.S. Chun, E.S. Lee, D.S. Lim, Effect of lattice structure of
[27]
[28]
[29]
[30]
[31]
silicon carbide on crystal formation of carbide-derived carbon, Carbon 79 (2014) 19e27. T.I.T. Okpalugo, P. Papakonstantinou, H. Murphy, J. McLaughlin, N.M.D. Brown, High resolution XPS characterization of chemical functionalised MWCNTs and SWCNTs, Carbon 43 (1) (2005) 153e161. V. Datsyuk, M. Kalyva, K. Papagelis, J. Parthenios, D. Tasis, A. Siokou, et al., Chemical oxidation of multiwalled carbon nanotubes, Carbon 46 (6) (2008) 833e840. P.C. Ma, S.Y. Mo, B.Z. Tang, J.K. Kim, Dispersion, interfacial interaction and reagglomeration of functionalized carbon nanotubes in epoxy composites, Carbon 48 (6) (2010) 1824e1834. S.C. Lee, F.C. Tai, C.H. Wei, Correlation between sp(2)/sp(3) ratio or hydrogen content and water contact angle in hydrogenated DLC film, Mater. Trans. 48 (9) (2007) 2534e2538. J.A. Kim, D.G. Seong, T.J. Kang, J.R. Youn, Effects of surface modification on rheological and mechanical properties of CNT/epoxy composites, Carbon 44 (10) (2006) 1898e1905.