American Institute of Aeronautics and Astronautics
1
Effect of Carbon Nanotube Growth Conditions on Strength and Stiffness of Carbon and Glass Fiber Polymer
Composites
J. Dai*, E. Soliman†, M. Safdari‡, M. Al-Haik§ and M. M. Reda Taha§§ University of New Mexico, Albuquerque, New Mexico 87131-0001, USA
Carbon nanotubes (CNTs) and Carbon nanofilaments (CNFs) were grown on the surface of carbon and glass sheets from fuel rich ethylene/oxygen combustion mixtures on certain catalytic metals. Employing a well-known catalyst deposition method, incipient wetness allowed growing CNTs/CNFs on the surface of carbon and fiber glass bidirectional sheets. Two-layer fiber reinforced polymer (FRP) composite plates were fabricated using a vacuum assisted hand lay-up technique of the carbon and glass fiber sheets after CNTs/CNFs were grown using 1.0% Nickel deposits. In this paper we report on the growth process and we examine the significance of the CNTs/CNFs growth conditions including the sizing burning temperature (250 and 500 ºC), the growth time period and the fiber base type (carbon or glass) on the strength and stiffness of this new multi-scale FRP composite. The ultimate tensile strength, tensile modulus (stiffness) and ultimate strain at failure was determined using ASTM D3039 test. It is shown that the sizing burning temperature has a significant effect on the strength and strain at failure of the new carbon fiber reinforced polymer (CFRP) and glass fiber reinforced polymer (GFRP) composites. Little effect on the composite stiffness was observed. Microstructural investigations of the failed specimens shed light on the fracture surface of the multi-scale CFRP and GFRP composites.
I. Introduction iber reinforced polymer (FRP) composites are structural materials that are load bearing and designed mainly based on mechanical properties. FRP composites are increasingly being used in civil infrastructure in a range of
applications such as reinforcing rods and tendons, wraps for seismic retrofit of columns, bonded reinforcement for strengthening of walls, beams and slabs and composite bridge decks1. It has been well established that the inclusion of nanomaterials within FRP composites has the promise of expanding their applications beyond load bearing. The attractive properties of carbon nanotubes (CNTs)2 might be attributed to their unique and defect free nanostructure. Single wall carbon nanotubes (SWCNT) possess exceptional mechanical3-4, thermal and electric properties5 compared to graphite, Kevlar, SiC and alumina fibers. The strength, elastic modulus and fracture properties of CNTs are an order of magnitude higher than most common composites used in civilian and military applications6-9. Moreover, CNTs reinforcement increases the toughness of the polymers and therefore they become impact resistance.
Most research to date focuses on using CNTs as a reinforcement, or as a filler, in a polymeric matrix by dispersing and perhaps subsequently aligning single- or multi-walled CNTs in the matrix10-11. Alignment and dispersion are critical factors that are difficult to control experimentally using typical mixing methods. CNTs embedded in a polymeric matrix form aggregates of themselves that are not only poorly adhered to in the matrix but also concentrate stresses that compromise the effect of the CNTs as reinforcement. Sonication12 and calendaring13 have been used to mitigate this problem but are not effective beyond 3% of the CNTs’ volume fraction due to the formation of agglomerated CNT aggregates in the polymer matrix14. The difficulty in uniformly dispersing CNTs in
* Post-Doctoral Research Fellow, Department of Mechanical Engineering. † PhD Student, Department of Civil Engineering. ‡ PhD Student, Department of Mechanical Engineering. § Assistant Professor, Department of Mechanical Engineering §§ Associate Professor, and Regents’ Lecturer, Department of Civil Engineering
F
51st AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference<BR> 18th12 - 15 April 2010, Orlando, Florida
American Institute of Aeronautics and Astronautics
2
polymer matrices arises from the large surface area of CNTs15 that is associated with its nano size. Dispersion and extrusion techniques have been reported in the literature for producing CNTs composites 16. The authors utilized high magnetic fields to process nanocomposites based on SWCNTs11. However, in both dispersion and extrusion techniques, producing uniform and well-dispersed CNTs composite is difficult because of the small amount of solid ‘powder’ (carbon) compared with the large amount of liquid polymer (matrix) in early mixing stages and the relatively high attractive forces between the CNTs particles. This often leads to phase separation due to the strong van der Waals forces between CNTs compared with CNTs and polymer 11.
Alternatively, to eliminate the dispersion problem, CNTs can be controlled-grown in places where they are needed and CNTs can be grown on most substrates such as silicon, silica and alumina14,17-18. However, there are fewer reports discussing CNTs growth on carbon materials, in particular yarns and fabrics19-20. Two challenges in CNTs growth on carbon substrates are (i) transition metals are easily diffused into the carbon substrates and (ii) different phases of carbon materials are able to form on graphite substrates because growth conditions are similar to the diamond or diamond-like carbon growth 21. Recently, Makris et al.22 grew aligned CNTs on carbon fibers, both PAN- and Pitch-based, by hot filament chemical vapor deposition (HFCVD) using H2 and CH4 as precursors. Nickel clusters were electrodeposited on the fiber surfaces to catalyze the growth, and uniform CNTs coatings were obtained on both PAN- and pitch-based carbon fibers. Qu et al23 reported a new method for uniform deposition of CNTs on carbon fibers, however, it requires processing at 1100 C in the presence of oxygen and such high temperature is anticipated to induce damage in the carbon fibers’ surface.
To this end, in order to minimize the carbon fibers’ surface damage due to high temperature exposure during
CNTs/CNFs growth, we propose to utilize the state-of-the-art technique named “ Graphitic Structures by Design (GSD)” to grow carbon nanofilaments on the surface of micro carbon fibers and glass fibers at a temperature as low as 550 C. By using the unique characteristics of the GSD approach (relatively low temperature, standard atmospheric pressure), a theoretical assumption of prior work is challenged which is that synthesis of carbon nanofibers (nanotubes or otherwise) occurs due to thermal decomposition of molecules. Furthermore, a new theory is formulated that the actual driving force of nanofiber synthesis is the creation of radical species created by the combustion process. Phillips et al.24 showed that different graphitic structures can be grown under different conditions and thus graphitic structures can be grown ‘by design’. The proposed GSD process consists of four main steps: (i) removal of the sizing from the surface of the carbon fiber, (ii) ‘activation by burning’ to form nucleation sites for a metal catalyst to adhere to, (iii) loading of the metal and (iv) synthesis of the CNTs/CNFs on the fabric surface. Note that all of these steps are accomplished with basic laboratory equipment (small furnace, gas bottles, etc.) unlike some of the methods discussed earlier24 and at temperatures (c.a. 550 °C) where there is no chance of carbon degradation, which cannot be said of other methods involving HFCVD. This article briefly discusses the synthesis of CNFs on carbon fibers and glass fibers and provides initial results of the mechanical behavior of the hybrid composites produced via GSD technique.
II. Experimental Procedures Growth of carbon nanotubes
Several 5.0 inch × 6.0 inch size pith-based carbon fiber sheets were obtained from Cytec Industries, Inc. and the same size fiber glass sheets were purchased from BFG Industries, Inc. The average diameters of the fibers are around 8 microns for the carbon fiber and 15 microns for the fiber glass respectively. In order to expose the actual surface of the carbon fiber, it is necessary to remove the sizing from the carbon fiber sheets. Specifically, the sheets were heated in an air environment under 500 ºC or 250 ºC for one hour to decompose the sizing, soaked in acetone for 1 hour to dissolve the sizing from the fibers, and then followed with an ethyl alcohol rinse. Finally, the sheets were air dried for 1 hour at 100 ºC.
A catalyst was prepared by incipient wetness impregnation of aqueous solutions of metal salt. In this study,
Ni(NO3)2•6H2O (supplied by Aldrich) was used as a metal source with the loading amount of Nickel adjusted to 1.0 wt. % for all the catalyzed samples. After the samples underwent drop by drop incipient wetting via a syringe pump, they were dried in an oven at 100 ºC overnight and then calcined at 250 ºC under a N2 (106 sccm) atmosphere for 4 hours to decompose the nickel salt. The synthesis of the carbon nanotubes was carried out in a Lindberg Heavy-Duty (one-zone) furnace (diameter and length of the horizontal reactor were 2 inch and 12 inch respectively). The samples were inserted into the center of the reactor where an inert gas (N2) was introduced at a flow rate of 600 sccm for 20 minutes while the temperature was being raised to 550 ºC. A reduction took place in 50 sccm diluted hydrogen by
American Institute of Aeronautics and Astronautics
3
nitrogen (N2:H2=90:10, v/v) for 1 hour at 550 ºC to reduce the already calcinated metallic oxide particles to pure nickel. The diluted hydrogen was then replaced by 600 sccm N2 flow for 20 minutes when the reactor was heated to 700 ºC. Then the N2 flow was reduced to 300 sccm. Meanwhile, 30 sccm ethylene (chemically pure) and 50 sccm diluted hydrogen (N2:H2=90:10, v/v) were introduced for 2 hours to deposit the carbon nanofilaments on the fiber sheets.
The carbon nanotubes and CNTs/CNFs grown on the fiber sheets were analyzed by a scanning electron
microscopy (SEM) (Hitachi S5200 Nano SEM) and a high resolution transmission electron microscope (HRTEM). Fig. 1 and Fig. 2 show SEM micrographs of CNTs/CNFs growth on carbon fiber sheets and glass fiber sheets respectively. Both figures show the fiber microstructure with burning at 500 oC and 250 oC. Fig. 3 and Fig. 4 show HRTEM micrographs of CNTs/CNFs growth on carbon fiber sheets and glass fiber sheets respectively. Both figures show the fiber microstructure with burning at 500 oC and 250 oC.
Figure 1. SEM of CNTs/CNFs on carbon fiber sheets burn at 500 ºC (a-b) and at 250 ºC (c-d)
(d)
(c)
(b)
(a)
American Institute of Aeronautics and Astronautics
4
Figure 2. SEM of CNTs/CNFs on glass fiber sheets burn at 500 ºC (a-c) and at 250 ºC (d-f)
American Institute of Aeronautics and Astronautics
5
Figure 3. TEM of CNTs/CNFs on carbon fiber sheets burn at 500 ºC (a-c) and at 250 ºC (d-f)
American Institute of Aeronautics and Astronautics
6
Figure 4. HRTEM of CNTs/CNFs on glass fiber sheets burn at 500 ºC (a-c) and at 250 ºC (d-f).
Fabrication of composite plate The FRP composite plates were fabricated using vacuum assisted hand lay-up technique. The carbon fiber fabric
used is 5.4 oz plain weave 3k tow size, 0.01 inch thick while the glass fiber used is 4 oz plain weave E glass with thickness of 0.0059 inch. The epoxy used is 635 thin epoxy system, resin:hardner mixing ratio of 2:1, pot life of 30-45 min. at 80 ºF, set time of 5-6 hours and drying time of 24-28 hours. In the fabrication process, a non-porous release film was attached over an aluminum plate and a peel ply was placed over the release film. The first layer of the fiber fabric was then placed and the epoxy was applied using a roller. The second layer was placed over the first layer and more epoxy was added to the second layer. The roller was used to compact the two layers together. Another peel ply was placed over the laminate in order to protect the laminate from contamination. A porous release film was laid over the peel ply, which serves as a barrier between the composite and the subsequent layer and allows for the suction of the air bubbles. A breather ply was placed over the film to form the vacuum path to the laminate. The entire system was covered with a Naylon bag and sealed by a sealant tape placed over the non-porous release film. A vacuum pump provided by Ideal Vacuum Products, LLC was connected to the nylon bag through a vacuum port. A pressure level of 2.3 × 10-2 Torr was applied and the bag was checked for any leakage. The vaccum
American Institute of Aeronautics and Astronautics
7
fabrication is shown in Fig. 5 and the fabricated carbon fiber reinforced polymer (CFRP) and glass fiber reinforced polymer (GFRP) composite plates are shown in Fig. 6.
Figure 5. Fabrication of carbon fiber fabric using the vacuum assisted hand lay-up technique
(a) CFRP (b) GFRP Figure 6. Fabrication of carbon fiber fabric using the vacuum assisted hand lay-up technique
III. Mechanical Characterization The experimental plan and all parameters examined in the experimental plan are summarized in Fig. 7. Three groups of fabrics were made; Group I: the original fabrics wihout any processing, Group II: the fabrics with the sizing removed and Group III: the fabrics with the sizing removed and CNTs growth. These groups were fabricated and tested for every set. Two temperatures were examined for removing the sizing, 250 ºC and 500 ºC. Four sets were examined by alternating the two temperatures and the two types of FRPs. All CNTs/CNFs growth were performed using Nickel at 700 oC. Two layer composite plate was fabricated for every specimen. Five specimens were created from each group and a tension test was performed on these specimens. The significance of fabrication parameters on the tensile strength of the CNTs-carbon and CNTs-glass FRP fabrics was examined. All specimens were prepared for the tension test in accordance with ASTM D3903 testing standards. The tension specimens were prepared using tapered G10 ends to allow uniform pressure distribution on the specimen at the grips location and to prevent slip and possible failure at the grips. The test specimens were left to cure for 24 hours and then were attached in the tension testing frame as shown in Fig. 8. The tension test was performed in accordance with ASTM D3903 and the load and displacement of the specimens were recorded using a standard data acquisition system.
American Institute of Aeronautics and Astronautics
8
Figure 7. Experimental plan to examine the significance of CNTs-composite fabrication process on the tensile strength of CNTs-composites.
(a) Dimensions (b) Testing frame
Figure 8. Tension testing frame (a) dimensions and (b) test set-up
IV. Results and Discussions Fig. 9 shows a comparison for the stress-strain curves of the carbon and glass fiber composites. The average
ultimate tensile strengths without any treatment denoted (R-R) were 56,565 and 11,995 psi for carbon and glass fibers respectively. These results show that the carbon fiber ultimate strength is approximately five times that of the glass fibers. The CFRP composite has showed stiffness and strain at failure about 50-70% higher than that of the GFRP composites. The effect of temperature level of the sizing removal can be deduced from Fig. 10. The average ultimate tensile strengths for the sizing removal (R-S) at 250 ºC were 43,094 and 9,747 psi for carbon and glass fibers respectively. The results here show that the carbon fiber experienced a slight increase in the average ultimate
American Institute of Aeronautics and Astronautics
9
strength (9%) in the case of 250 ºC while the strength decreased by 25% in the case of 500 ºC. In the case of glass fibers, the average ultimate strength decreased by 37% and 50% in the case of 250 ºC and 500 ºC respectively. It is worth noting that the strength reduction was always accompanied with a reduction in strain at failure in both carbon and glass fiber composites. The increase of the carbon fiber strength with sizing removal at 250 oC might be attributed to an enhanced frictional/bond strength between the carbon fiber surface and the epoxy matrix when the sizing material was removed. However, it is obvious that temperatures higher than 250 oC such as the case of sizing removal at 500 oC might have an adverse effect and result in damaging the fiber material leading to a reduced composite strength. It can also be observed that glass fiber is much more sensitive to such thermal treatment than carbon fibers with strength loss at both 250 oC and 500 oC. Damage at growth might be attributed to the existence of Oxygen. Therefore, the removal of Oxygen during the preprocessing and growth process might be necessary. Finally, no significant effect of either sizing removal or CNTs/CNFs growth can be observed on the carbon or glass fiber composite stiffness. This may indicate that damage due to thermal treatment is basically a surface damage of the fiber materials that does not allow for development of good bond with the matrix.
0
10000
20000
30000
40000
50000
60000
70000
0.0 0.5 1.0 1.5 2.0 2.5
Strain, %
Str
ess,
psi
Carbon (R-R)
Glass (R-R)
Figure 9. Comparison between the carbon and glass fiber composites produced using raw fibers.
0
10000
20000
30000
40000
50000
60000
70000
80000
0.0 0.5 1.0 1.5 2.0Strain, %
Str
ess,
psi
Carbon (R-R)
Carbon (R-S)-250
Carbon (R-S)-500
0
5000
10000
15000
20000
25000
0.0 0.5 1.0 1.5 2.0Strain, %
Str
ess
, p
si
Glass (R-R)Glass (R-S)-250Glass (R-S)-500
(a) CFRP (b) GFRP
Figure 10. Effect of sizing removal temperature on final strength of composites.
Fig. 11 shows that the growth of the CNTs/CNFs decreases the ultimate strength further. For example, the average ultimate strength for the carbon fibers with sizing removal at 500 ºC decreased by 26% after growing carbon nanotubes. Similarly, the average ultimate strength of glass fibers with sizing removal at 250 ºC decreased by 25%. This loss of strength after growth might be explained by the further damage anticipated in the fiber during growth at 700 oC. SEM investigation of fracture surfaces for the carbon fiber composites with and without CNTs/CNFs growth was performed. Fig. 12 and Fig. 13 show the fracture surface at different growth conditions. No significant difference between the fracture surfaces was observed under the SEM that can be correlated to the fabrication process. However, sizing material melt down can be observed in both Fig. 12 (b) and 13 (b). While the SEM micrographs do not show a major difference between sizing removal using 250 oC and 500 oC, it is obvious from the results that temperatures as high as 500 oC will have an adverse effect on the fiber itself. Moreover, it can be
American Institute of Aeronautics and Astronautics
10
deduced that fiber damage took place during the growth process resulting in a significant loss of composite strength. It is therefore obvious that successful growth of CNTs/CNFs CFRP or GFRP composites will require preparing growth conditions to occur at low temperature to prevent fiber damage.
0
10000
20000
30000
40000
50000
60000
70000
0.0 0.5 1.0 1.5 2.0 2.5Strain, %
Str
es
s, p
si
Carbon (R-R)Carbon (R-S)-500Carbon (R-G)-500
0
5000
10000
15000
20000
25000
0.0 0.5 1.0 1.5 2.0Strain, %
Str
ess,
ps
i
Glass (R-R)
Glass (R-S)-250
Glass (R-G)-250
(a) CFRP (b) GFRP
Figure 11. Effect of carbon nanotubes growth on the tensile strength of composite.
(a)
(b)
(c) Figure 12. SEM micrographs showing composite fracture surface of (a) raw fibers from specimens (R-R), (b) after
sizing removal at 250 oC from specimens (R-S) and (c) after sizing removal at 250 oC and growth at 700 oC of CNTs/CNFs from specimens (R-G).
American Institute of Aeronautics and Astronautics
11
(a)
(b)
(c) Figure 13. SEM micrographs showing composite fracture surface of (a) raw fibers from specimens (R-R), (b) after
sizing removal at 500 oC from specimens (R-S) and (c) after sizing removal at 500 oC and growth at 700 oC of CNTs/CNFs from specimens (R-G).
V. Conclusion The above experimental investigation showed sizing removal and growth typically result in reducing CFRP and
GFRP strength. Little effect on the material stiffness was observed, but a reduction of the ultimate strength and strain at failure was observed. Of special interest was the significance of sizing removal at 250 ºC where a slight increase in the composite strength was observed. The removal of Oxygen during the preprocessing and growth process seems necessary to limit damage in the composite materials. Further research to allow CNTs/CNFs growth at temperatures lower than those reported here are currently being investigated.
Acknowledgments This research is funded by four grants to the research team by Defense Threat Reduction Agency (DTRA) Grant
# HDTRA1-08-1-0017 P00001, National Science Foundation (NSF) Grant Award # CMMI-0800249 and Army Research Office (ARO) Grant # W911NF-08-1-0421. The authors extend their thanks for such great support.
References 1. Karbhari, V.M. Durability of Composites for Civil Structural Applications, (Woodhead Publishing Limited, Cambridge
England 2007). 2. Iijima, S. Helical Microtubules of Graphitic Carbon. Nature 354(1991). 3. Rinzler, A.G., Liu, J., Dai, H. & al., P.N.e. Applied Physics 67, A29 -37 (1998).
American Institute of Aeronautics and Astronautics
12
4. Dietzel, D., Faucher, M. & A. Iaia, e.a. Analysis of mechanical properties of single wall carbon nanotubes fixed at a tip apex by atomic force microscopy. Nanotechnology 16, S73 (2005).
5. Bernholc, J., Brenner, D. & M.B. Nardelli, e.a. Mechanical and electrical properties of nanotubes. Annu. Rev. Mater. Res. 32(2002).
6. ASM. Composites Handbook, , (ASM International, Materials Park, Ohio, 1987). 7. ASM. Fatigue and Fracture Handbook. (ed. Lampman, S.R.) (ASM International, Materials Park, Ohio 1996). 8. ASM. Metals Handbook Desk Edition, , (ASM International, Materials Park, Ohio 1998). 9. Thostenson, E. & Chou, W. Aligned Multi-Walled Carbon Nanotube-Reinforced Composites: Processing and
Mechanical Characterization. J. Phys. D. 35(2001). 10. Choi, E., Al-Haik, M., Garmestani, H. & al, e. Enhancement of Thermal and Electrical Properties of Carbon Nanotube
Polymer Composites by Magnetic Field Processing. Journal of Applied Physics 94(2003). 11. Garmestani, H., Al-Haik, M.S. & K. Dahmen, e.a. Epoxy Mediated Alignment of Single Wall Carbon Nanotubes
Under High Magnetic Fields. Advanced Materials 15 (2003). 12. Jin, L., Bower, C. & Zhou, O. Alignment of Carbon Nanotubes in a Polymer Matrix by Mechanical Stretching. Applied
Physics Letters 73, 1197-1199 (1998). 13. Gojny, F.H., Wichmann, M.H.G., Köpke, U., Fiedler, B. & Schulte, K. Carbon Nanotube-Reinforced Epoxy-
Composites: Enhanced Stiffness and Fracture Toughness at Low Nanotube Content. Composites Science and Technology, 64, 2363-2371 (2004).
14. Garcia, E., et al. Aligned Carbon Nanotube Reinforcement of Advanced Composite Ply Interfaces. in 49th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference 16t (Schaumburg, IL, 2008).
15. Xia, Z., et al. Direct observation of toughening mechanisms in carbon nanotube ceramic matrix composites. Acta Materialia 52, 931-944 (2004).
16. Li, Y.L., Kinloch, I.A. & Windle, A.H. Science 304(2004). 17. García, E.J., Hart, A.J. & Wardl, B.L. Electrical and Thermal Properties of Hybrid Woven Composites Reinforced with
Aligned Carbon Nanotubes. AIAA Journal 46, 1405-1412 (2008). 18. Yamamoto, N. & Wardl, B.L. Electrical and Thermal Properties of Hybrid Woven Composites Reinforced with
Aligned Carbon Nanotubes. in 49th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference 7 - 10 April 2008, (Schaumburg, IL, 2008).
19. Otsuka, K., Abe, Y. & al., N.K.e. Synthesis of carbon nanotubes on Ni/carbon-fiber catalysts under mild conditions. Carbon 42 (2004).
20. Zhua, S., Sub, C.-H., Lehoczkyb, S.L. & al, e. Carbon nanotube growth on carbon fibers. Diamond and Related Materials 12(2003).
21. Zhua, S., Sub, C.-H. & S.L. Lehoczkyb, e.a. Carbon nanotube growth on carbon fibers. Diamond and Related Materials 12, 1825-1828 (2003).
22. Makris, T.D., Giorgi, R. & N. Lisi, e.a. Carbon Nanotube Growth on PAN- and Pitch-Based Carbon Fibres by HFCVD. Fullerenes, Nanotubes, and Carbon Nanostructures, 13, 383-392 (2005).
23. Qu, L., Zhao, Y. & Dai, L. Carbon Microfibers Sheathed with Aligned Carbon Nanotubes: Towards Multidimensional, Multicomponent, and Multifunctional Nanomaterials. Small 2 1052 - 1059 (2006).
24. J. Phillips, T.S., M. Nemer & Lester, K. Graphitic Structures by Design. Langmuir 22, 9694-9703 (2006).
