- Email: [email protected]
Low-temperature grafting of carbon nanotubes on carbon fibers using a bimetallic floating catalyst
Diamond & Related Materials 68 (2016) 118–126
Contents lists available at ScienceDirect
Diamond & Related Materials journal homepage: www.elsevier.com/locate/diamond
Low-temperature grafting of carbon nanotubes on carbon fibers using a bimetallic floating catalyst Geunsung Lee a, Ji Ho Youk b, Jinyong Lee c, In Hwan Sul d, Woong-Ryeol Yu a,⁎ a
Department of Materials Science and Engineering, Seoul National University, 599 Gwanangno, Gwanak-gu, Seoul 151-742, Republic of Korea Department of Advanced Fiber Engineering, Inha University, 253 Yonghyun-dong, Nam-gu, Incheon 402-751, Republic of Korea c High Temperature Composite Materials Group, The 4th R&D Institute-4, Agency for Defense Development, Daejeon 305-600, Republic of Korea d Department of Materials Design Engineering, Kumoh National Institute of Technology, Gumi 39177, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 4 April 2016 Received in revised form 24 June 2016 Accepted 24 June 2016 Available online 27 June 2016 Keywords: Carbon nanotubes Direct growth Carbon fiber Low temperature Chemical vapor deposition
a b s t r a c t The degradation mechanism of CFs experiencing the catalytic growth of CNTs on their surface through high-temperature CVD was revealed using microstructural analysis. Catalysts nanoparticles, if their energy was high enough to diffuse into CFs, generated micro-voids in CFs and reduced their tensile strength. Herein, we report a new low-temperature processing route to grow CNTs on CF surfaces without mechanically degrading the CFs. The use of a Ni–Fe bimetallic catalyst was key to achieving the growth of CNTs at low temperature, e.g., 500 °C. Catalyst diffusion into the CFs during CVD was inhibited at this temperature, which facilitated uniform growth of CNTs on only the CF surfaces and minimized internal structural changes of the CFs. © 2016 Elsevier B.V. All rights reserved.
Fig. 1. Schematic diagram of the FCVD process for a two-zone furnace. The catalyst precursors and carbon source supplied into the first zone sublime and vaporize, respectively. These then decompose into the catalyst and carbon atoms in the second zone, where CNTs grow on the substrate.
1. Introduction The hybridization of carbon nanotubes (CNTs) and carbon fibers (CFs) is a versatile method to develop new and advanced materials by hierarchically combining their excellent thermal, electrical and ⁎ Corresponding author. E-mail address: [email protected] (W.-R. Yu).
http://dx.doi.org/10.1016/j.diamond.2016.06.012 0925-9635/© 2016 Elsevier B.V. All rights reserved.
mechanical properties at the nano- and microscales [1]. CNT-grafted CFs made via direct growth have emerged as a material that can improve the reinforcing effect of CFs in composites and solve the dispersion problems of CNTs [2]. Radially grown CNTs on CFs improve the radial stiffness [3] and axial tensile strength [4,5] of CF-reinforced composites, the interfacial shear strength (IFSS) of polymer composites [6], and the electrochemical performance of CF electrodes [7]. Several issues must be resolved, however, for successful dissemination of CNT-grafted CFs. These issues include the manufacturing method, which degrades
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
119
Fig. 2. Sample preparation using focused ion beam (FIB) milling to observe the internal structure of CNT-grafted carbon fibers (CFs) by transmission electron microscopy (TEM).
the mechanical properties of the resulting CFs [1,8], and the high cost of the batch process used for their manufacture, i.e., the pre-process for introducing catalyst nanoparticles on the CF surface [9] and the main process for growing CNTs via chemical vapor deposition (CVD) [10]. Herein, we demonstrate that a bimetallic catalyst and floating catalyst chemical vapor deposition (FCVD) can solve these problems. As one of the simplest methods to graft CNTs on CFs, FCVD provides carbon sources and catalysts concurrently, inducing the deposition
and growth of CNTs on substrates. There have been several studies concerning the grafting of CNTs on various substrates using FCVD [11– 15]. When CFs are used as the substrate, the catalyst nanoparticles diffuse into the CFs and become poisoned and deactivated by carbon atoms separated from the CFs because the CFs are porous [16]. Diffused catalysts destroy the internal structure of CFs, while the poisoning and deactivation of catalyst particles hinders CNT growth [17]. Diffused catalysts depopulate the CF surfaces of active catalyst particles, also
Fig. 3. Morphologies of CNT-grafted CFs prepared from different substrate CFs, catalysts, and CVD temperatures. (a) As-received CFs, bimetallic catalyst, 500 °C; (b) oxidized CFs, bimetallic catalyst, 500 °C; (c) as-received CFs, Fe-only catalyst, 750 °C; (d) oxidized CFs, Fe-only catalyst, 750 °C.
120
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
hampering CNT growth [18]. To prevent the deactivation and diffusion of the catalysts, buffer layers have been coated on CF surfaces to inhibit inter-diffusion between carbon and catalyst particles [19]. Buffer layers and coatings increase material and manufacturing costs. An alternative surface treatment technique was developed that improves the chemical affinity of CFs and enlarges their surface roughness [20,21]. Such surface-treated CFs have a greater affinity for catalyst nanoparticles and show improved CNT growth, but the CFs still degrade during the hightemperature CVD process. Recent research has grafted CNTs onto CFs without degradation or buffer layers. Two approaches were taken to avoid the degradation. One approach was to graft CNTs onto CFs by spraying [22] or electrophoresis [6] which do not require high-temperature procedures such as CVD. These methods are easy and suitable for large-scale production but they require additional processes to stabilize the bonding between the CNTs and CF [4]. A second approach was to carefully control the CVD to inhibit catalyst diffusion. We solved the problem of mechanical degradation of CNT-grafted CF by controlling the amount of catalysts coated on the CFs [5]. This method, however, was not sufficiently versatile for large-scale production because of the need for precise control of the amount of catalyst and the relatively high cost. Inter-diffusion of catalysts and CFs and the thermal degradation of CFs are high-temperature phenomena. Thus, lowering the CVD temperature may be the most effective means of solving the catalyst poisoning and diffusion issues. This hypothesis was investigated using FCVD with Ni–Fe bimetallic catalysts. Bimetallic catalysts are known to lower the activation energy for CNT growth and thus CVD temperature [23]. In this research, we also investigated the mechanism underlying the mechanical degradation of CNT-grafted CFs prepared by high-temperature CVD through microstructural analysis.
2.2. Characterization of CNT-grafted CFs The morphologies of the CFs after the FCVD process were observed by field-emission scanning electron microscopy (FE-SEM). Energy-dispersive X-ray spectroscopy (EDS) was conducted to verify the elemental composition of the metal nanoparticles deposited on the CFs (JEOL, JSM 7600F). To investigate the effect of catalyst diffusion into the CF on its mechanical properties, the internal microstructure of the CNTgrafted CFs was investigated using high-resolution transmission electron microscopy (HR-TEM, JEM–3000F). The specimens were prepared using a focused ion beam (FIB, SMI3050SE, SII Nanotechnology) of Ga + ions. The CNT-grafted CFs were attached to a silicon wafer. Amorphous carbon was then deposited on the surface of the CNT-grafted CFs
2. Materials and methods 2.1. Preparation of CNT-grafted CFs by FCVD FCVD using a two-zone furnace offers a simplified and continuous process to manufacture CNT-grafted CFs (Fig. 1). A mixture of catalyst precursors and a carbon source are injected into the first zone, where the precursors are sublimated or vaporized. CFs are placed in the second zone in batch or continuous processes. The sublimated catalysts and vaporized carbon sources are decomposed into catalysts and carbon atoms in the second zone, resulting in CNT growth on the CF substrates. Carbon fibers (T700SC, Toray) were spread out and used as a substrate to minimize the shade effect. For comparison, oxidized CFs were also prepared by heating as-received CFs at 500 °C for 1 h to vary the chemical affinity of the CF surface toward catalyst particles. A bimetallic catalyst system was prepared using ferrocene (Fe(η5–C5H5)2, Sigma– Aldrich) and nickelocene (Ni(η5-C5H5)2, Sigma–Aldrich). Ferrocene and nickelocene have the same structure and differ only in their metallic element. They have a similar melting point (ferrocene: 172.5 °C; nickelocene: 171 °C to 173 °C) and decomposition temperatures (around 400 °C) [24,25]. A1:1 (molar ratio) mixture of these metallocenes was prepared in toluene (99.5%, Sigma–Aldrich). The weight ratio of the mixture was 2 wt% and the procedure was conducted in an N2-purged bottle because of the air sensitivity of nickelocene. The solution was introduced into the first FCVD zone as precursors of the catalysts and as a carbon source. FCVD was conducted at 500, 600 and 750 °C to investigate the temperature dependency. To investigate the effect of heat on the diffusion of catalysts into CFs, heat treatment was applied to the CFs in the absence of catalysts at the same temperatures. Catalyst-only deposition onto the CFs was also evaluated to determine the elemental composition of the catalyst particles deposited on the CFs. The precursor mixture was located just below the CF bundles and heated to 500 and 750 °C.
Fig. 4. The effect of temperature on CNT growth on CFs during FCVD with the bimetallic catalyst. CNT growth was hindered with increasing temperature. (a) 500 °C, (b) 600 °C, (c) 750 °C.
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
121
Fig. 5. Morphology of the catalyst-only deposited CFs and their energy-dispersive X-ray spectroscopy (EDS) results. (a) Morphology. (b) Ni and (c) Fe elemental mapping results after catalyst-only deposition at 500 °C. (d) Morphology. (b) Ni and (c) Fe elemental mapping results after catalyst-only deposition at 750 °C.
to prevent damage during the FIB process. Fig. 2 schematically shows the sample preparation procedure. Wide-angle and small-angle X-ray scattering analyses (WAXS and SAXS) with Debye images (WAXS: D/ MAX-2500, Rigaku; SAXS: TVXA-ENIF1, Tech Valley) were carried out to investigate the orientation of the carbon crystals and microvoids in the CFs after the CVD processing. Raman spectroscopy (T64000, Horiba; Ar laser, 514 nm) was used to evaluate the quality of the CNTs grafted on the CFs and the carbon structure of the substrate CFs. The mechanical properties of the CNT-grafted CFs were measured using a single-fiber tensile test with a 20 mm gage length and 2 mm/min strain ratio using a universal test machine. 3. Results and discussion Fig. 6. Tensile strength and modulus of CNT-grafted CFs prepared using different FCVD conditions. The x-axis labels correspond to the as-received CFs, only heat-treated CFs without CVD, CNT-grafted CFs prepared using Fe-only catalysts (Fe-CNT) and the bimetallic catalysts (Bi-CNT).
3.1. Surface morphology of the CFs after FCVD processing The effect of temperature and catalyst (bimetallic or Fe-only catalysts) on CNT growth on CF surfaces was investigated by observing the
122
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
chemical affinity of the oxidized CFs helped to prevent the diffusion of catalysts into the CF. Note that catalyst nanoparticles at high temperatures are prone to diffusing into the substrate CFs or becoming poisoned and deactivated [16]. The behavior of the bimetallic catalyst as a function of temperature was also investigated. Fig. 4 shows that CNTs grew uniformly on the surface of CFs during FCVD at 500 °C but CNT growth was hindered as the temperature increased. It is known that the speed of CNT growth increases as the CVD temperature increases [26]. The trend observed with the CFs above implies that the hindrance of CNT growth was caused not by a kinetic process but by deterioration of the catalysts, e.g., from inter-diffusion between carbon atoms and the catalysts. We also clarified the composition of the metal particles in FCVD by SEM and EDS by examining catalyst-only deposited CFs. Mapping of the Ni and Fe elements on the SEM images established that the catalyst particles were composed of both Ni and Fe when a bimetallic mixture was used (Fig. 5). Damaged catalyst particles were observed after processing at 750 °C while CFs after processing at 500 °C showed uniform surfaces with catalyst particles. Fig. 7. Wide-angle X-ray scattering WAXS profiles of CFs after heat treatment. Bi-500 and Fe-750 represent bimetallic (Ni–Fe) catalyst-deposited CFs after heat treatment at 500 °C for 1 h and Fe catalyst-deposited CFs after heat treatment at 750 °C for 1 h (Fe-750), respectively. Heat-CF means CF samples with only heat treatment and without catalyst.
morphologies of the CFs after the FCVD process. Fig. 3(a) and (b) show the morphologies of the as-received and oxidized CFs, respectively, demonstrating that CNTs grew at low temperature (500 °C) in the presence of the bimetallic catalyst. The chemical affinity and roughness of the substrate CFs had little effect on the bimetallic catalyst. Notably, CNT growth was not observed at such a low temperature with the single metallic catalyst (Fe-only) (Fig. 3(c)). In the Fe-only catalyst cases, CNT growth was highly dependent on the surface condition (Fig. 3(c) and (d)). At 750 °C, CNTs grew only on oxidized CFs, indicating that the
3.2. Mechanical properties and microstructure of the CNT-grafted CFs The effects of the CVD process on the mechanical properties of the CNT-grafted CFs were investigated (Fig. 6). For comparison, as-received CFs and CFs heat-treated at 750 °C without CVD processing were also characterized. The CVD temperatures for preparing the CNT-grafted CFs using the Fe-only and the bimetallic catalysts were 750 and 500 ° C, respectively. Two notable features were identified. The elastic moduli of all of the samples were similar regardless of the treatment conditions, whereas the tensile strength of the CNT-grafted CFs prepared using the Fe-only catalyst was severely degraded. Because the tensile strength of materials is governed by material defects or imperfections [27], evidently some defects were generated during the Fe-only CVD process. The microstructure of the CFs was then examined to reveal the cause of such degradation and to explain the similar elastic modulus of all samples. WAXS analysis was carried out and Debye images of the CFs were obtained (Fig. 7). To eliminate the effect of the CNTs on the WAXS data, catalyst-only deposited CFs without growing CNTs were examined. The crystallinities of the CFs were obtained using the full width at half maximum (FWHM) of each WAXS peak. The crystalline orientation of materials has an important influence on the mechanical properties; Debye ring images were also investigated for this reason (Fig. 8). The crystal orientation of the CFs was similar for the four samples. Table 1 shows that the crystallinity and orientation indices of all of the CFs were similar, suggesting that the crystal structure and orientation of the CFs were not changed (or degraded) by the high-temperature processing. Therefore, X-ray diffraction analysis could not explain the great difference in the tensile strengths of the CFs shown in Fig. 6. WAXS analysis did not reveal any significant differences between the CFs. This was attributed to the very small size of the nanoparticles (10–20 nm); at this scale, structural differences such as lamellar structures and voids cannot be seen in WAXS. Other techniques were used to study the microstructure at the nanometer scale, i.e., Raman
Table 1 Orientation index of CFs obtained from X-ray diffraction analysis. Preferred orientation
Fig. 8. Debye diffractograms of (a) as-received CFs, (b) CFs thermally treated at 750 °C for 1 h, (c) CFs with bimetallic catalyst thermally treated at 500 °C for 1 h and (d) CFs with Fe catalyst thermally treated at 750 °C for 1 h. The first and second reflected rings represent the (200) and (310) planes, respectively.
Sample
FWHM (%)
Int. W (%)
Tensile strength (MPa)
As-received Heat-CF Bimetal-500 Fe-750
89.8 89.1 89.7 89.1
77.2 75.7 76.4 75.5
4002 (683) 3976 (306) 3944 (508) 1082 (74)
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
123
Fig. 9. Raman spectra of (a) CNTs grafted on CF using bimetallic catalyst FCVD at 500 °C, (b) CNTs grafted on CF using Fe-only catalyst FCVD at 750 °C, (c) as-received CFs, (d) CFs with bimetallic catalyst thermally treated (but without CVD) at 500 °C for 1 h and (e) CFs with Fe-only catalyst thermally treated (but without CVD) at 750 °C for 1 h.
spectroscopy and SAXS. Raman spectra were obtained for pristine CFs and CNTs grafted on CFs at different temperatures and catalysts (Fig. 9(a) and (b)). CNTs grafted on the CFs during the high-temperature FCVD process had a much lower ID/IG ratio (0.43) than CNTs grafted on CFs at low temperature (1.28). As reported in [29], high-temperature annealing improved the quality of the CNTs. The Raman spectra indicated that the quality of the CNTs grafted on the CFs varied with the catalyst and CVD temperature (i.e., high-temperature processing provided
better-quality CNTs). In the case of the CF substrate, however, those CFs after the high-temperature heat treatment with Fe catalysts had a higher ID/IG ratio (0.91; Fig. 9(e)) than as-received CFs (0.75; Fig. 9(c)) and CFs after low-temperature heat treatment with bimetallic catalysts (0.76; Fig. 9(d)), implying smaller crystallites on their surfaces [30]. Since lower ID/IG ratio usually means high modulus and not high strength, the Raman spectra could not explain the mechanical properties as shown in Fig. 6.
124
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
Fig. 10. Small-angle X-ray scattering (SAXS) patterns of CFs. (a) As-received CFs, (b) Ni–Fe bimetallic catalyst-only deposited CFs after heat treatment at 500 °C for 1 h (Bi-500), (c) Fe catalyst-only deposited CFs after heat treatment at 750 °C for 1 h (Fe-750) and (d) the scattering intensity of two-dimensional SAXS patterns along the vertical meridial direction.
SAXS analysis can usefully characterize microstructures [32,33] and microvoids [34,35] in fibers. Microvoids in CFs have a great influence on their mechanical properties [36]. SAXS analysis for the CFs (Fig. 10) established that the scattering intensity of the Fe–750 sample was higher than that of the other samples, which indicated a relatively larger volume of microvoids [37] between the lamellae. These microvoids acted as defects, which resulted in the mechanical degradation highlighted in Fig. 6. The internal structure of the CFs was directly observed using FIB and TEM; significantly different morphologies were observed. Fig. 11(a) clearly shows that catalyst nanoparticles had diffused into the CFs that had undergone Fe-only catalyst FCVD at 750 °C. We believe that this diffusion prevented the CNTs from growing on the CFs (Fig. 3(c)). In contrast, no catalyst nanoparticles were observed on CNT-grafted CFs prepared using bimetallic-catalyst FCVD (Fig. 11(b)). This explained the good growth of the CNTs on the CFs observed in Fig. 3(a). Some catalyst nanoparticles diffused into the CFs with bimetallic FCVD processing at 750 °C (Fig. 11(c)). This diffusion explained why there was no growth of CNTs observed with the bimetallic FCVD at this highest temperature (Fig. 4(c)). Regardless of the type of catalyst, high-temperature CVD caused the catalyst nanoparticles to diffuse into the CFs. Furthermore, this diffusion resulted in a lowered tensile strength because of the introduction of defects near the surface.
analyses. The catalysts nanoparticles, provided their energy was sufficient to diffuse into the CFs, generated microvoids near the surface that acted as stress-concentrators, thereby lowering the strength of the CFs. We demonstrated that CNT-grafted CFs can be manufactured without mechanical degradation of the CFs. This was achieved using a bimetallic catalyst, which lowered the activation energy for CNT growth, thereby enabling a lower CVD temperature to be used. Interdiffusion of CFs and catalysts, which is the main source of the mechanical degradation and reduced CNT growth, was successfully avoided with the low-temperature CVD process.
Prime novelty statement The degradation mechanism of the mechanical properties of carbon fibers (CFs) experiencing the catalytic growth of carbon nanotubes (CNTs) on their surface through high-temperature CVD was revealed using microstructural analysis, reporting a new lowtemperature processing route to grow CNTs on CF surfaces without mechanically degrading the CFs using a Ni–Fe bimetallic floating catalyst system.
Acknowledgement 4. Conclusions The degradation mechanism of CFs experiencing catalytic growth of CNTs on their surface was revealed using TEM, Raman, WAXD and SAXD
This work was supported by DAPA and ADD and by the National Research Foundation of Korea (NRF) grant funded by the Ministry of Science, ICT & Future Planning (MSIP) (NO. NRF-2015R1A5A1037627).
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
Fig. 11. The internal structure of CNT-grafted CFs prepared via different FCVD processes. (a) Fe-only catalyst FCVD at 750 °C, (b) bimetallic catalyst FCVD at 500 °C and (c) bimetallic catalyst FCVD at 750 °C.
References [1] E. Thostenson, W. Li, D. Wang, Z. Ren, T. Chou, Carbon nanotube/carbon fiber hybrid multiscale composites, J. Appl. Phys. 91 (2002) 6034–6037.
125
[2] Y.S. Song, J.R. Youn, Influence of dispersion states of carbon nanotubes on physical properties of epoxy nanocomposites, Carbon 43 (2005) 1378–1385. [3] K. Kepple, G. Sanborn, P. Lacasse, K. Gruenberg, W. Ready, Improved fracture toughness of carbon fiber composite functionalized with multi walled carbon nanotubes, Carbon 46 (2008) 2026–2033. [4] G. Lee, K.D. Ko, Y.C. Yu, J. Lee, W.-R. Yu, J.H. Youk, A facile method for preparing CNTgrafted carbon fibers and improved tensile strength of their composites, Compos. A Appl. Sci. Manuf. 69 (2015) 132–138. [5] K.J. Kim, J. Kim, W.-R. Yu, J.H. Youk, J. Lee, Improved tensile strength of carbon fibers undergoing catalytic growth of carbon nanotubes on their surface, Carbon 54 (2013) 258–267. [6] E. Bekyarova, E.T. Thostenson, A. Yu, H. Kim, J. Gao, J. Tang, H.T. Hahn, T.W. Chou, M.E. Itkis, R.C. Haddon, Multiscale carbon nanotube − carbon fiber reinforcement for advanced epoxy composites, Langmuir 23 (2007) 3970–3974. [7] H. Qian, A. Bismarck, E.S. Greenhalgh, G. Kalinka, M.S. Shaffer, Hierarchical composites reinforced with carbon nanotube grafted fibers: the potential assessed at the single fiber level, Chem. Mater. 20 (2008) 1862–1869. [8] E.T. Thostenson, T.-W. Chou, Aligned multi-walled carbon nanotube-reinforced composites: processing and mechanical characterization, J. Phys. D Appl. Phys. 35 (2002) L77. [9] K.J. Kim, W.-R. Yu, J.H. Youk, J. Lee, Factors governing the growth mode of carbon nanotubes on carbon-based substrates, Phys. Chem. Chem. Phys. 14 (2012) 14041–14048. [10] Q. Zhang, J. Liu, R. Sager, L. Dai, J. Baur, Hierarchical composites of carbon nanotubes on carbon fiber: influence of growth condition on fiber tensile properties, Compos. Sci. Technol. 69 (2009) 594–601. [11] C. Singh, M.S. Shaffer, A.H. Windle, Production of controlled architectures of aligned carbon nanotubes by an injection chemical vapour deposition method, Carbon 41 (2003) 359–368. [12] G. Lee, K.J. Kim, W.-R. Yu, J.H. Youk, The effect of the surface roughness of carbon fibres on CNT growth by floating-catalyst chemical vapour deposition, Int. J. Nanotechnol. 10 (2013) 800–810. [13] H. Liu, Y. Zhang, D. Arato, R. Li, P. Mérel, X. Sun, Aligned multi-walled carbon nanotubes on different substrates by floating catalyst chemical vapor deposition: critical effects of buffer layer, Surf. Coat. Technol. 202 (2008) 4114–4120. [14] S. Talapatra, S. Kar, S. Pal, R. Vajtai, L. Ci, P. Victor, M. Shaijumon, S. Kaur, O. Nalamasu, P. Ajayan, Direct growth of aligned carbon nanotubes on bulk metals, Nat. Nanotechnol. 1 (2006) 112–116. [15] R.G. de Villoria, S. Figueredo, A. Hart, S. Steiner Iii, A. Slocum, B. Wardle, High-yield growth of vertically aligned carbon nanotubes on a continuously moving substrate, Nanotechnology 20 (2009) 405611. [16] K. Kuwana, H. Endo, K. Saito, D. Qian, R. Andrews, E.A. Grulke, Catalyst deactivation in CVD synthesis of carbon nanotubes, Carbon 43 (2005) 253–260. [17] K.J. Kim, W.-R. Yu, J.H. Youk, J. Lee, Degradation and healing mechanisms of carbon fibers during the catalytic growth of carbon nanotubes on their surfaces, ACS Appl. Mater. Interfaces 4 (2012) 2250–2258. [18] A.A. Puretzky, D.B. Geohegan, S. Jesse, I.N. Ivanov, G. Eres, In situ measurements and modeling of carbon nanotube array growth kinetics during chemical vapor deposition, Appl. Phys. A Mater. Sci. Process. 81 (2005) 223–240. [19] W. Hu, D. Gong, Z. Chen, L. Yuan, K. Saito, C.A. Grimes, P. Kichambare, Growth of well-aligned carbon nanotube arrays on silicon substrates using porous alumina film as a nanotemplate, Appl. Phys. Lett. 79 (2001) 3083–3085. [20] J.S. Lee, T.J. Kang, Changes in physico-chemical and morphological properties of carbon fiber by surface treatment, Carbon 35 (1997) 209–216. [21] C. Masarapu, B. Wei, Direct growth of aligned multiwalled carbon nanotubes on treated stainless steel substrates, Langmuir 23 (2007) 9046–9049. [22] D.C. Davis, J.W. Wilkerson, J. Zhu, V.G. Hadjiev, A strategy for improving mechanical properties of a fiber reinforced epoxy composite using functionalized carbon nanotubes, Compos. Sci. Technol. 71 (2011) 1089–1097. [23] W.H. Chiang, R.M. Sankaran, Synergistic effects in bimetallic nanoparticles for low temperature carbon nanotube growth, Adv. Mater. 20 (2008) 4857–4861. [24] L. Brissonneau, R. Sahnoun, C. Mijoule, C. Vahlas, Investigation of nickelocene decomposition during chemical vapor deposition of nickel, J. Electrochem. Soc. 147 (2000) 1443–1448. [25] M. Bernhauer, M. Braun, K.J. Hüttinger, Kinetics of mesophase formation in a stirred tank reactor and properties of the products—V. Catalysis by ferrocene, Carbon 32 (1994) 1073–1085. [26] K.-E. Kim, K.-J. Kim, W.S. Jung, S.Y. Bae, J. Park, J. Choi, J. Choo, Investigation on the temperature-dependent growth rate of carbon nanotubes using chemical vapor deposition of ferrocene and acetylene, Chem. Phys. Lett. 401 (2005) 459–464. [27] W.D. Callister, D.G. Rethwisch, Materials science and engineering: an introduction, Wiley, New York, 2007. [29] F. Huang, K.T. Yue, P. Tan, S.-L. Zhang, Z. Shi, X. Zhou, Z. Gu, Temperature dependence of the Raman spectra of carbon nanotubes, J. Appl. Phys. 84 (1998) 4022–4024. [30] F. Tuinstra, J.L. Koenig, Raman spectrum of graphite, J. Chem. Phys. 53 (1970) 1126–1130. [32] B.R. Pauw, M.E. Vigild, K. Mortensen, J.W. Andreasen, E.A. Klop, D.W. Breiby, O. Bunk, Strain-induced internal fibrillation in looped aramid filaments, Polymer 51 (2010) 4589–4598. [33] T. Kobayashi, K. Sumiya, Y. Fukuba, M. Fujie, T. Takahagi, K. Tashiro, Structural heterogeneity and stress distribution in carbon fiber monofilament as revealed by synchrotron micro-beam X-ray scattering and micro-Raman spectral measurements, Carbon 49 (2011) 1646–1652. [34] A.F. Thünemann, W. Ruland, Microvoids in polyacrylonitrile fibers: a small-angle Xray scattering study, Macromolecules 33 (2000) 1848–1852.
126
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
[35] W. Ruland, Carbon fibers, Adv. Mater. 2 (1990) 528–536. [36] Z. Hua, Y.J. Zhong, D.F. Li, Pores in PAN-based carbon fiber and effects of pore defects on mechanical properties, Materials Science Forum, Trans Tech Publ. 2007, pp. 1665–1668.
[37] C. Zhu, X. Liu, X. Yu, N. Zhao, J. Liu, J. Xu, A small-angle X-ray scattering study and molecular dynamics simulation of microvoid evolution during the tensile deformation of carbon fibers, Carbon 50 (2012) 235–243.
Contents lists available at ScienceDirect
Diamond & Related Materials journal homepage: www.elsevier.com/locate/diamond
Low-temperature grafting of carbon nanotubes on carbon fibers using a bimetallic floating catalyst Geunsung Lee a, Ji Ho Youk b, Jinyong Lee c, In Hwan Sul d, Woong-Ryeol Yu a,⁎ a
Department of Materials Science and Engineering, Seoul National University, 599 Gwanangno, Gwanak-gu, Seoul 151-742, Republic of Korea Department of Advanced Fiber Engineering, Inha University, 253 Yonghyun-dong, Nam-gu, Incheon 402-751, Republic of Korea c High Temperature Composite Materials Group, The 4th R&D Institute-4, Agency for Defense Development, Daejeon 305-600, Republic of Korea d Department of Materials Design Engineering, Kumoh National Institute of Technology, Gumi 39177, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 4 April 2016 Received in revised form 24 June 2016 Accepted 24 June 2016 Available online 27 June 2016 Keywords: Carbon nanotubes Direct growth Carbon fiber Low temperature Chemical vapor deposition
a b s t r a c t The degradation mechanism of CFs experiencing the catalytic growth of CNTs on their surface through high-temperature CVD was revealed using microstructural analysis. Catalysts nanoparticles, if their energy was high enough to diffuse into CFs, generated micro-voids in CFs and reduced their tensile strength. Herein, we report a new low-temperature processing route to grow CNTs on CF surfaces without mechanically degrading the CFs. The use of a Ni–Fe bimetallic catalyst was key to achieving the growth of CNTs at low temperature, e.g., 500 °C. Catalyst diffusion into the CFs during CVD was inhibited at this temperature, which facilitated uniform growth of CNTs on only the CF surfaces and minimized internal structural changes of the CFs. © 2016 Elsevier B.V. All rights reserved.
Fig. 1. Schematic diagram of the FCVD process for a two-zone furnace. The catalyst precursors and carbon source supplied into the first zone sublime and vaporize, respectively. These then decompose into the catalyst and carbon atoms in the second zone, where CNTs grow on the substrate.
1. Introduction The hybridization of carbon nanotubes (CNTs) and carbon fibers (CFs) is a versatile method to develop new and advanced materials by hierarchically combining their excellent thermal, electrical and ⁎ Corresponding author. E-mail address: [email protected] (W.-R. Yu).
http://dx.doi.org/10.1016/j.diamond.2016.06.012 0925-9635/© 2016 Elsevier B.V. All rights reserved.
mechanical properties at the nano- and microscales [1]. CNT-grafted CFs made via direct growth have emerged as a material that can improve the reinforcing effect of CFs in composites and solve the dispersion problems of CNTs [2]. Radially grown CNTs on CFs improve the radial stiffness [3] and axial tensile strength [4,5] of CF-reinforced composites, the interfacial shear strength (IFSS) of polymer composites [6], and the electrochemical performance of CF electrodes [7]. Several issues must be resolved, however, for successful dissemination of CNT-grafted CFs. These issues include the manufacturing method, which degrades
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
119
Fig. 2. Sample preparation using focused ion beam (FIB) milling to observe the internal structure of CNT-grafted carbon fibers (CFs) by transmission electron microscopy (TEM).
the mechanical properties of the resulting CFs [1,8], and the high cost of the batch process used for their manufacture, i.e., the pre-process for introducing catalyst nanoparticles on the CF surface [9] and the main process for growing CNTs via chemical vapor deposition (CVD) [10]. Herein, we demonstrate that a bimetallic catalyst and floating catalyst chemical vapor deposition (FCVD) can solve these problems. As one of the simplest methods to graft CNTs on CFs, FCVD provides carbon sources and catalysts concurrently, inducing the deposition
and growth of CNTs on substrates. There have been several studies concerning the grafting of CNTs on various substrates using FCVD [11– 15]. When CFs are used as the substrate, the catalyst nanoparticles diffuse into the CFs and become poisoned and deactivated by carbon atoms separated from the CFs because the CFs are porous [16]. Diffused catalysts destroy the internal structure of CFs, while the poisoning and deactivation of catalyst particles hinders CNT growth [17]. Diffused catalysts depopulate the CF surfaces of active catalyst particles, also
Fig. 3. Morphologies of CNT-grafted CFs prepared from different substrate CFs, catalysts, and CVD temperatures. (a) As-received CFs, bimetallic catalyst, 500 °C; (b) oxidized CFs, bimetallic catalyst, 500 °C; (c) as-received CFs, Fe-only catalyst, 750 °C; (d) oxidized CFs, Fe-only catalyst, 750 °C.
120
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
hampering CNT growth [18]. To prevent the deactivation and diffusion of the catalysts, buffer layers have been coated on CF surfaces to inhibit inter-diffusion between carbon and catalyst particles [19]. Buffer layers and coatings increase material and manufacturing costs. An alternative surface treatment technique was developed that improves the chemical affinity of CFs and enlarges their surface roughness [20,21]. Such surface-treated CFs have a greater affinity for catalyst nanoparticles and show improved CNT growth, but the CFs still degrade during the hightemperature CVD process. Recent research has grafted CNTs onto CFs without degradation or buffer layers. Two approaches were taken to avoid the degradation. One approach was to graft CNTs onto CFs by spraying [22] or electrophoresis [6] which do not require high-temperature procedures such as CVD. These methods are easy and suitable for large-scale production but they require additional processes to stabilize the bonding between the CNTs and CF [4]. A second approach was to carefully control the CVD to inhibit catalyst diffusion. We solved the problem of mechanical degradation of CNT-grafted CF by controlling the amount of catalysts coated on the CFs [5]. This method, however, was not sufficiently versatile for large-scale production because of the need for precise control of the amount of catalyst and the relatively high cost. Inter-diffusion of catalysts and CFs and the thermal degradation of CFs are high-temperature phenomena. Thus, lowering the CVD temperature may be the most effective means of solving the catalyst poisoning and diffusion issues. This hypothesis was investigated using FCVD with Ni–Fe bimetallic catalysts. Bimetallic catalysts are known to lower the activation energy for CNT growth and thus CVD temperature [23]. In this research, we also investigated the mechanism underlying the mechanical degradation of CNT-grafted CFs prepared by high-temperature CVD through microstructural analysis.
2.2. Characterization of CNT-grafted CFs The morphologies of the CFs after the FCVD process were observed by field-emission scanning electron microscopy (FE-SEM). Energy-dispersive X-ray spectroscopy (EDS) was conducted to verify the elemental composition of the metal nanoparticles deposited on the CFs (JEOL, JSM 7600F). To investigate the effect of catalyst diffusion into the CF on its mechanical properties, the internal microstructure of the CNTgrafted CFs was investigated using high-resolution transmission electron microscopy (HR-TEM, JEM–3000F). The specimens were prepared using a focused ion beam (FIB, SMI3050SE, SII Nanotechnology) of Ga + ions. The CNT-grafted CFs were attached to a silicon wafer. Amorphous carbon was then deposited on the surface of the CNT-grafted CFs
2. Materials and methods 2.1. Preparation of CNT-grafted CFs by FCVD FCVD using a two-zone furnace offers a simplified and continuous process to manufacture CNT-grafted CFs (Fig. 1). A mixture of catalyst precursors and a carbon source are injected into the first zone, where the precursors are sublimated or vaporized. CFs are placed in the second zone in batch or continuous processes. The sublimated catalysts and vaporized carbon sources are decomposed into catalysts and carbon atoms in the second zone, resulting in CNT growth on the CF substrates. Carbon fibers (T700SC, Toray) were spread out and used as a substrate to minimize the shade effect. For comparison, oxidized CFs were also prepared by heating as-received CFs at 500 °C for 1 h to vary the chemical affinity of the CF surface toward catalyst particles. A bimetallic catalyst system was prepared using ferrocene (Fe(η5–C5H5)2, Sigma– Aldrich) and nickelocene (Ni(η5-C5H5)2, Sigma–Aldrich). Ferrocene and nickelocene have the same structure and differ only in their metallic element. They have a similar melting point (ferrocene: 172.5 °C; nickelocene: 171 °C to 173 °C) and decomposition temperatures (around 400 °C) [24,25]. A1:1 (molar ratio) mixture of these metallocenes was prepared in toluene (99.5%, Sigma–Aldrich). The weight ratio of the mixture was 2 wt% and the procedure was conducted in an N2-purged bottle because of the air sensitivity of nickelocene. The solution was introduced into the first FCVD zone as precursors of the catalysts and as a carbon source. FCVD was conducted at 500, 600 and 750 °C to investigate the temperature dependency. To investigate the effect of heat on the diffusion of catalysts into CFs, heat treatment was applied to the CFs in the absence of catalysts at the same temperatures. Catalyst-only deposition onto the CFs was also evaluated to determine the elemental composition of the catalyst particles deposited on the CFs. The precursor mixture was located just below the CF bundles and heated to 500 and 750 °C.
Fig. 4. The effect of temperature on CNT growth on CFs during FCVD with the bimetallic catalyst. CNT growth was hindered with increasing temperature. (a) 500 °C, (b) 600 °C, (c) 750 °C.
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
121
Fig. 5. Morphology of the catalyst-only deposited CFs and their energy-dispersive X-ray spectroscopy (EDS) results. (a) Morphology. (b) Ni and (c) Fe elemental mapping results after catalyst-only deposition at 500 °C. (d) Morphology. (b) Ni and (c) Fe elemental mapping results after catalyst-only deposition at 750 °C.
to prevent damage during the FIB process. Fig. 2 schematically shows the sample preparation procedure. Wide-angle and small-angle X-ray scattering analyses (WAXS and SAXS) with Debye images (WAXS: D/ MAX-2500, Rigaku; SAXS: TVXA-ENIF1, Tech Valley) were carried out to investigate the orientation of the carbon crystals and microvoids in the CFs after the CVD processing. Raman spectroscopy (T64000, Horiba; Ar laser, 514 nm) was used to evaluate the quality of the CNTs grafted on the CFs and the carbon structure of the substrate CFs. The mechanical properties of the CNT-grafted CFs were measured using a single-fiber tensile test with a 20 mm gage length and 2 mm/min strain ratio using a universal test machine. 3. Results and discussion Fig. 6. Tensile strength and modulus of CNT-grafted CFs prepared using different FCVD conditions. The x-axis labels correspond to the as-received CFs, only heat-treated CFs without CVD, CNT-grafted CFs prepared using Fe-only catalysts (Fe-CNT) and the bimetallic catalysts (Bi-CNT).
3.1. Surface morphology of the CFs after FCVD processing The effect of temperature and catalyst (bimetallic or Fe-only catalysts) on CNT growth on CF surfaces was investigated by observing the
122
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
chemical affinity of the oxidized CFs helped to prevent the diffusion of catalysts into the CF. Note that catalyst nanoparticles at high temperatures are prone to diffusing into the substrate CFs or becoming poisoned and deactivated [16]. The behavior of the bimetallic catalyst as a function of temperature was also investigated. Fig. 4 shows that CNTs grew uniformly on the surface of CFs during FCVD at 500 °C but CNT growth was hindered as the temperature increased. It is known that the speed of CNT growth increases as the CVD temperature increases [26]. The trend observed with the CFs above implies that the hindrance of CNT growth was caused not by a kinetic process but by deterioration of the catalysts, e.g., from inter-diffusion between carbon atoms and the catalysts. We also clarified the composition of the metal particles in FCVD by SEM and EDS by examining catalyst-only deposited CFs. Mapping of the Ni and Fe elements on the SEM images established that the catalyst particles were composed of both Ni and Fe when a bimetallic mixture was used (Fig. 5). Damaged catalyst particles were observed after processing at 750 °C while CFs after processing at 500 °C showed uniform surfaces with catalyst particles. Fig. 7. Wide-angle X-ray scattering WAXS profiles of CFs after heat treatment. Bi-500 and Fe-750 represent bimetallic (Ni–Fe) catalyst-deposited CFs after heat treatment at 500 °C for 1 h and Fe catalyst-deposited CFs after heat treatment at 750 °C for 1 h (Fe-750), respectively. Heat-CF means CF samples with only heat treatment and without catalyst.
morphologies of the CFs after the FCVD process. Fig. 3(a) and (b) show the morphologies of the as-received and oxidized CFs, respectively, demonstrating that CNTs grew at low temperature (500 °C) in the presence of the bimetallic catalyst. The chemical affinity and roughness of the substrate CFs had little effect on the bimetallic catalyst. Notably, CNT growth was not observed at such a low temperature with the single metallic catalyst (Fe-only) (Fig. 3(c)). In the Fe-only catalyst cases, CNT growth was highly dependent on the surface condition (Fig. 3(c) and (d)). At 750 °C, CNTs grew only on oxidized CFs, indicating that the
3.2. Mechanical properties and microstructure of the CNT-grafted CFs The effects of the CVD process on the mechanical properties of the CNT-grafted CFs were investigated (Fig. 6). For comparison, as-received CFs and CFs heat-treated at 750 °C without CVD processing were also characterized. The CVD temperatures for preparing the CNT-grafted CFs using the Fe-only and the bimetallic catalysts were 750 and 500 ° C, respectively. Two notable features were identified. The elastic moduli of all of the samples were similar regardless of the treatment conditions, whereas the tensile strength of the CNT-grafted CFs prepared using the Fe-only catalyst was severely degraded. Because the tensile strength of materials is governed by material defects or imperfections [27], evidently some defects were generated during the Fe-only CVD process. The microstructure of the CFs was then examined to reveal the cause of such degradation and to explain the similar elastic modulus of all samples. WAXS analysis was carried out and Debye images of the CFs were obtained (Fig. 7). To eliminate the effect of the CNTs on the WAXS data, catalyst-only deposited CFs without growing CNTs were examined. The crystallinities of the CFs were obtained using the full width at half maximum (FWHM) of each WAXS peak. The crystalline orientation of materials has an important influence on the mechanical properties; Debye ring images were also investigated for this reason (Fig. 8). The crystal orientation of the CFs was similar for the four samples. Table 1 shows that the crystallinity and orientation indices of all of the CFs were similar, suggesting that the crystal structure and orientation of the CFs were not changed (or degraded) by the high-temperature processing. Therefore, X-ray diffraction analysis could not explain the great difference in the tensile strengths of the CFs shown in Fig. 6. WAXS analysis did not reveal any significant differences between the CFs. This was attributed to the very small size of the nanoparticles (10–20 nm); at this scale, structural differences such as lamellar structures and voids cannot be seen in WAXS. Other techniques were used to study the microstructure at the nanometer scale, i.e., Raman
Table 1 Orientation index of CFs obtained from X-ray diffraction analysis. Preferred orientation
Fig. 8. Debye diffractograms of (a) as-received CFs, (b) CFs thermally treated at 750 °C for 1 h, (c) CFs with bimetallic catalyst thermally treated at 500 °C for 1 h and (d) CFs with Fe catalyst thermally treated at 750 °C for 1 h. The first and second reflected rings represent the (200) and (310) planes, respectively.
Sample
FWHM (%)
Int. W (%)
Tensile strength (MPa)
As-received Heat-CF Bimetal-500 Fe-750
89.8 89.1 89.7 89.1
77.2 75.7 76.4 75.5
4002 (683) 3976 (306) 3944 (508) 1082 (74)
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
123
Fig. 9. Raman spectra of (a) CNTs grafted on CF using bimetallic catalyst FCVD at 500 °C, (b) CNTs grafted on CF using Fe-only catalyst FCVD at 750 °C, (c) as-received CFs, (d) CFs with bimetallic catalyst thermally treated (but without CVD) at 500 °C for 1 h and (e) CFs with Fe-only catalyst thermally treated (but without CVD) at 750 °C for 1 h.
spectroscopy and SAXS. Raman spectra were obtained for pristine CFs and CNTs grafted on CFs at different temperatures and catalysts (Fig. 9(a) and (b)). CNTs grafted on the CFs during the high-temperature FCVD process had a much lower ID/IG ratio (0.43) than CNTs grafted on CFs at low temperature (1.28). As reported in [29], high-temperature annealing improved the quality of the CNTs. The Raman spectra indicated that the quality of the CNTs grafted on the CFs varied with the catalyst and CVD temperature (i.e., high-temperature processing provided
better-quality CNTs). In the case of the CF substrate, however, those CFs after the high-temperature heat treatment with Fe catalysts had a higher ID/IG ratio (0.91; Fig. 9(e)) than as-received CFs (0.75; Fig. 9(c)) and CFs after low-temperature heat treatment with bimetallic catalysts (0.76; Fig. 9(d)), implying smaller crystallites on their surfaces [30]. Since lower ID/IG ratio usually means high modulus and not high strength, the Raman spectra could not explain the mechanical properties as shown in Fig. 6.
124
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
Fig. 10. Small-angle X-ray scattering (SAXS) patterns of CFs. (a) As-received CFs, (b) Ni–Fe bimetallic catalyst-only deposited CFs after heat treatment at 500 °C for 1 h (Bi-500), (c) Fe catalyst-only deposited CFs after heat treatment at 750 °C for 1 h (Fe-750) and (d) the scattering intensity of two-dimensional SAXS patterns along the vertical meridial direction.
SAXS analysis can usefully characterize microstructures [32,33] and microvoids [34,35] in fibers. Microvoids in CFs have a great influence on their mechanical properties [36]. SAXS analysis for the CFs (Fig. 10) established that the scattering intensity of the Fe–750 sample was higher than that of the other samples, which indicated a relatively larger volume of microvoids [37] between the lamellae. These microvoids acted as defects, which resulted in the mechanical degradation highlighted in Fig. 6. The internal structure of the CFs was directly observed using FIB and TEM; significantly different morphologies were observed. Fig. 11(a) clearly shows that catalyst nanoparticles had diffused into the CFs that had undergone Fe-only catalyst FCVD at 750 °C. We believe that this diffusion prevented the CNTs from growing on the CFs (Fig. 3(c)). In contrast, no catalyst nanoparticles were observed on CNT-grafted CFs prepared using bimetallic-catalyst FCVD (Fig. 11(b)). This explained the good growth of the CNTs on the CFs observed in Fig. 3(a). Some catalyst nanoparticles diffused into the CFs with bimetallic FCVD processing at 750 °C (Fig. 11(c)). This diffusion explained why there was no growth of CNTs observed with the bimetallic FCVD at this highest temperature (Fig. 4(c)). Regardless of the type of catalyst, high-temperature CVD caused the catalyst nanoparticles to diffuse into the CFs. Furthermore, this diffusion resulted in a lowered tensile strength because of the introduction of defects near the surface.
analyses. The catalysts nanoparticles, provided their energy was sufficient to diffuse into the CFs, generated microvoids near the surface that acted as stress-concentrators, thereby lowering the strength of the CFs. We demonstrated that CNT-grafted CFs can be manufactured without mechanical degradation of the CFs. This was achieved using a bimetallic catalyst, which lowered the activation energy for CNT growth, thereby enabling a lower CVD temperature to be used. Interdiffusion of CFs and catalysts, which is the main source of the mechanical degradation and reduced CNT growth, was successfully avoided with the low-temperature CVD process.
Prime novelty statement The degradation mechanism of the mechanical properties of carbon fibers (CFs) experiencing the catalytic growth of carbon nanotubes (CNTs) on their surface through high-temperature CVD was revealed using microstructural analysis, reporting a new lowtemperature processing route to grow CNTs on CF surfaces without mechanically degrading the CFs using a Ni–Fe bimetallic floating catalyst system.
Acknowledgement 4. Conclusions The degradation mechanism of CFs experiencing catalytic growth of CNTs on their surface was revealed using TEM, Raman, WAXD and SAXD
This work was supported by DAPA and ADD and by the National Research Foundation of Korea (NRF) grant funded by the Ministry of Science, ICT & Future Planning (MSIP) (NO. NRF-2015R1A5A1037627).
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
Fig. 11. The internal structure of CNT-grafted CFs prepared via different FCVD processes. (a) Fe-only catalyst FCVD at 750 °C, (b) bimetallic catalyst FCVD at 500 °C and (c) bimetallic catalyst FCVD at 750 °C.
References [1] E. Thostenson, W. Li, D. Wang, Z. Ren, T. Chou, Carbon nanotube/carbon fiber hybrid multiscale composites, J. Appl. Phys. 91 (2002) 6034–6037.
125
[2] Y.S. Song, J.R. Youn, Influence of dispersion states of carbon nanotubes on physical properties of epoxy nanocomposites, Carbon 43 (2005) 1378–1385. [3] K. Kepple, G. Sanborn, P. Lacasse, K. Gruenberg, W. Ready, Improved fracture toughness of carbon fiber composite functionalized with multi walled carbon nanotubes, Carbon 46 (2008) 2026–2033. [4] G. Lee, K.D. Ko, Y.C. Yu, J. Lee, W.-R. Yu, J.H. Youk, A facile method for preparing CNTgrafted carbon fibers and improved tensile strength of their composites, Compos. A Appl. Sci. Manuf. 69 (2015) 132–138. [5] K.J. Kim, J. Kim, W.-R. Yu, J.H. Youk, J. Lee, Improved tensile strength of carbon fibers undergoing catalytic growth of carbon nanotubes on their surface, Carbon 54 (2013) 258–267. [6] E. Bekyarova, E.T. Thostenson, A. Yu, H. Kim, J. Gao, J. Tang, H.T. Hahn, T.W. Chou, M.E. Itkis, R.C. Haddon, Multiscale carbon nanotube − carbon fiber reinforcement for advanced epoxy composites, Langmuir 23 (2007) 3970–3974. [7] H. Qian, A. Bismarck, E.S. Greenhalgh, G. Kalinka, M.S. Shaffer, Hierarchical composites reinforced with carbon nanotube grafted fibers: the potential assessed at the single fiber level, Chem. Mater. 20 (2008) 1862–1869. [8] E.T. Thostenson, T.-W. Chou, Aligned multi-walled carbon nanotube-reinforced composites: processing and mechanical characterization, J. Phys. D Appl. Phys. 35 (2002) L77. [9] K.J. Kim, W.-R. Yu, J.H. Youk, J. Lee, Factors governing the growth mode of carbon nanotubes on carbon-based substrates, Phys. Chem. Chem. Phys. 14 (2012) 14041–14048. [10] Q. Zhang, J. Liu, R. Sager, L. Dai, J. Baur, Hierarchical composites of carbon nanotubes on carbon fiber: influence of growth condition on fiber tensile properties, Compos. Sci. Technol. 69 (2009) 594–601. [11] C. Singh, M.S. Shaffer, A.H. Windle, Production of controlled architectures of aligned carbon nanotubes by an injection chemical vapour deposition method, Carbon 41 (2003) 359–368. [12] G. Lee, K.J. Kim, W.-R. Yu, J.H. Youk, The effect of the surface roughness of carbon fibres on CNT growth by floating-catalyst chemical vapour deposition, Int. J. Nanotechnol. 10 (2013) 800–810. [13] H. Liu, Y. Zhang, D. Arato, R. Li, P. Mérel, X. Sun, Aligned multi-walled carbon nanotubes on different substrates by floating catalyst chemical vapor deposition: critical effects of buffer layer, Surf. Coat. Technol. 202 (2008) 4114–4120. [14] S. Talapatra, S. Kar, S. Pal, R. Vajtai, L. Ci, P. Victor, M. Shaijumon, S. Kaur, O. Nalamasu, P. Ajayan, Direct growth of aligned carbon nanotubes on bulk metals, Nat. Nanotechnol. 1 (2006) 112–116. [15] R.G. de Villoria, S. Figueredo, A. Hart, S. Steiner Iii, A. Slocum, B. Wardle, High-yield growth of vertically aligned carbon nanotubes on a continuously moving substrate, Nanotechnology 20 (2009) 405611. [16] K. Kuwana, H. Endo, K. Saito, D. Qian, R. Andrews, E.A. Grulke, Catalyst deactivation in CVD synthesis of carbon nanotubes, Carbon 43 (2005) 253–260. [17] K.J. Kim, W.-R. Yu, J.H. Youk, J. Lee, Degradation and healing mechanisms of carbon fibers during the catalytic growth of carbon nanotubes on their surfaces, ACS Appl. Mater. Interfaces 4 (2012) 2250–2258. [18] A.A. Puretzky, D.B. Geohegan, S. Jesse, I.N. Ivanov, G. Eres, In situ measurements and modeling of carbon nanotube array growth kinetics during chemical vapor deposition, Appl. Phys. A Mater. Sci. Process. 81 (2005) 223–240. [19] W. Hu, D. Gong, Z. Chen, L. Yuan, K. Saito, C.A. Grimes, P. Kichambare, Growth of well-aligned carbon nanotube arrays on silicon substrates using porous alumina film as a nanotemplate, Appl. Phys. Lett. 79 (2001) 3083–3085. [20] J.S. Lee, T.J. Kang, Changes in physico-chemical and morphological properties of carbon fiber by surface treatment, Carbon 35 (1997) 209–216. [21] C. Masarapu, B. Wei, Direct growth of aligned multiwalled carbon nanotubes on treated stainless steel substrates, Langmuir 23 (2007) 9046–9049. [22] D.C. Davis, J.W. Wilkerson, J. Zhu, V.G. Hadjiev, A strategy for improving mechanical properties of a fiber reinforced epoxy composite using functionalized carbon nanotubes, Compos. Sci. Technol. 71 (2011) 1089–1097. [23] W.H. Chiang, R.M. Sankaran, Synergistic effects in bimetallic nanoparticles for low temperature carbon nanotube growth, Adv. Mater. 20 (2008) 4857–4861. [24] L. Brissonneau, R. Sahnoun, C. Mijoule, C. Vahlas, Investigation of nickelocene decomposition during chemical vapor deposition of nickel, J. Electrochem. Soc. 147 (2000) 1443–1448. [25] M. Bernhauer, M. Braun, K.J. Hüttinger, Kinetics of mesophase formation in a stirred tank reactor and properties of the products—V. Catalysis by ferrocene, Carbon 32 (1994) 1073–1085. [26] K.-E. Kim, K.-J. Kim, W.S. Jung, S.Y. Bae, J. Park, J. Choi, J. Choo, Investigation on the temperature-dependent growth rate of carbon nanotubes using chemical vapor deposition of ferrocene and acetylene, Chem. Phys. Lett. 401 (2005) 459–464. [27] W.D. Callister, D.G. Rethwisch, Materials science and engineering: an introduction, Wiley, New York, 2007. [29] F. Huang, K.T. Yue, P. Tan, S.-L. Zhang, Z. Shi, X. Zhou, Z. Gu, Temperature dependence of the Raman spectra of carbon nanotubes, J. Appl. Phys. 84 (1998) 4022–4024. [30] F. Tuinstra, J.L. Koenig, Raman spectrum of graphite, J. Chem. Phys. 53 (1970) 1126–1130. [32] B.R. Pauw, M.E. Vigild, K. Mortensen, J.W. Andreasen, E.A. Klop, D.W. Breiby, O. Bunk, Strain-induced internal fibrillation in looped aramid filaments, Polymer 51 (2010) 4589–4598. [33] T. Kobayashi, K. Sumiya, Y. Fukuba, M. Fujie, T. Takahagi, K. Tashiro, Structural heterogeneity and stress distribution in carbon fiber monofilament as revealed by synchrotron micro-beam X-ray scattering and micro-Raman spectral measurements, Carbon 49 (2011) 1646–1652. [34] A.F. Thünemann, W. Ruland, Microvoids in polyacrylonitrile fibers: a small-angle Xray scattering study, Macromolecules 33 (2000) 1848–1852.
126
G. Lee et al. / Diamond & Related Materials 68 (2016) 118–126
[35] W. Ruland, Carbon fibers, Adv. Mater. 2 (1990) 528–536. [36] Z. Hua, Y.J. Zhong, D.F. Li, Pores in PAN-based carbon fiber and effects of pore defects on mechanical properties, Materials Science Forum, Trans Tech Publ. 2007, pp. 1665–1668.
[37] C. Zhu, X. Liu, X. Yu, N. Zhao, J. Liu, J. Xu, A small-angle X-ray scattering study and molecular dynamics simulation of microvoid evolution during the tensile deformation of carbon fibers, Carbon 50 (2012) 235–243.