- Email: [email protected]
Macroscopic growth of carbon nanotube mats and their mechanical properties
Letters to the Editor / Carbon 45 (2007) 1105–1136 [3] Huang JF, Zeng XR, Li HJ, Xiong XB, Huang M. Influence of the preparing temperature on phase, microstructure and anti-oxidation property of SiC coating for C/C composites. Carbon 2004;42:1517–21. [4] Fu QG, Li HJ, Shi XH, Li KZ, Sun GD. Silicon carbide coating to protect carbon/carbon composites against oxidation. Scripta Mater 2005;52:923–7. [5] Kowbei W, Withers JC. CVD and CVR silicon-based functionally gradient coatings on C–C composites. Carbon 1995;33:415–26.
1133
[6] Fang HT, Jing CZ, Zhang DY, Jae HJ, Yoo DH. A Si–Mo fused slurry coating for oxidation protection of carbon–carbon composites. J Mater Sci Lett 2001;20:175–7. [7] Zhang YL, Li HJ, Fu QG, Li KZ, W J, Wang PY. A C/SiC gradient oxidation protective coating for carbon/carbon composites. Surf Coat Technol 2006;201(6):3491–5.
Macroscopic growth of carbon nanotube mats and their mechanical properties Simone Musso *, Samuele Porro, Mauro Giorcelli, Angelica Chiodoni, Carlo Ricciardi, Alberto Tagliaferro Politecnico di Torino, Dipartimento di Fisica, C.so Duca degli Abruzzi 24, 10129 Torino, Italy Received 12 September 2006; accepted 15 December 2006 Available online 27 December 2006
Since carbon nanotubes have been recognized as a ultrastrong material, both by simulations and experimental measurements [1], various applications have been proposed for their use in thermal and electrical enhancement [2,3] and in the area of strength and toughness as reinforcement for composite materials [4]. Nevertheless, significant progresses in developing carbon nanotube-composites technology are limited by the lack of a simple, economical and high conversion/deposition rate production method for carbon nanotubes. Because of the several advantages of the CVD techniques [5,6], much efforts have been devoted to the optimization of this method, though a real breakthrough in this field has yet to be achieved. Here we report the growth and the properties of self standing large area millimeters-thick sheets of as-deposited multiwall carbon nanotubes (MWCNTs). Such sheets were grown by catalytical chemical vapor deposition (CVD) at a growth rate up to 0.5 lm s1 over an area limited only by the deposition system size (up to 100 cm2 at present). With this method curved surfaces can be straightforwardly coated. The sheets showed good thermal stability and interesting mechanical properties such as elasticity and resistance under compression and were totally hydrophobic. These results show that the route towards the use of carbon nanotubes as a structural material is now open. The thick layers (over 2 mm) of MWCNTs, termed ‘‘mats’’ from here on, were grown in a CVD reactor formed by a horizontal quartz tube housed in a cylindrical furnace
*
Corresponding author. Fax: +39 011 5647399. E-mail address: [email protected] (S. Musso).
0008-6223/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2006.12.019
in which a constant nitrogen gas flow rate was maintained at pressures just above atmospheric [7]. The carbon (camphor) and the catalyst (ferrocene) sources were mixed in the ratio 20:1 in a pyrex flask. The precursors solid mixture, which contained no solvents, was melted and evaporated directly from the flask by heating it at 220 C with a heater plate. The flask connection between the quartz tube and the nitrogen gas source was realized with a T-joint. The gas evaporated from the flask was injected into the nitrogen flux, that carried it to the substrate region. Deposition of the MWCNTs took place on non pre-etched bare silicon substrates. The temperature in the deposition region was kept at 850 C. The reactor quartz tube internal diameter was 4.2 cm, and the length of the uniform temperature region of about 30 cm. The size of the substrate housed in such a tube was therefore 4 · 30 cm (substrate divided in two pieces 4 · 15 each, due to the size of commercial silicon wafers). The temperature gradient present at the edges of the uniform temperature region caused small gas turbulences reducing the uniform deposition area to 4 · 25 cm (100 cm2). As the current version of the system does not allow a continuous feed of the reagents, the process time lasted two hours at most. At the end of the process the substrates (up to 100 cm2 in total) were covered by a few millimeters thick uniform MWCNT mat. A consistent deposition of MWCNTs occurred also on every curved surface of the silicon substrate (see Fig. 1b) and on the inner surface of the quartz tube, although on the latter the deposition is in the form of non ordered compact and irregular tangles. The deposition rate reached 0.5 lm s1 and the mass production rate (considering only the mats grown on Si substrate) exceeded 1 g h1, while the total
1134
Letters to the Editor / Carbon 45 (2007) 1105–1136
Fig. 2. Thermo-gravimeter analysis. Thermo-oxidation experiments suggested the presence of secondary carbonaceous products (high T derivative shoulder).
Fig. 1. Images of the MWCNT mat: (a) photograph of a block of the MWCNT mat evidencing its thickness (the Euro coin is 2.33 mm high); (b) SEM image of a CNT-coated substrate edge; (c) SEM image showing the macroscopic vertical alignment of the MWCNT; (d,e) SEM images with increasing magnification, showing the diameter distribution and the entanglement of nanotubes.
conversion rate is of over 30% in weight. The role of silicon in the mat formation remains unexplained, though literature suggests [8] that the morphology of the substrate wafers plays a role in the nucleation of the CNT and in the chemical/physical interaction with the catalyst particle. The MWCNT mat was mechanically delaminated from the substrate with a razor blade and blocks such as that shown in Fig. 1a were obtained. The figure shows that the MWCNT mat is thick and self standing. The substrate side of the mat blocks showed an optically flat surface. The density of the mat was measured by weighing a block of MWCNT versus its volume, and a value of 0.5 g/cm3 was obtained. This value is lower than the density expected from an isolated MWCNT. This is consistent with the fact that MWCNTs organized in mats were not closely packed, but spaced, as shown by SEM (Fig. 1d and e). Although Fig. 1c shows a macroscopic alignment of the MWCNTs, a closer look at the mats (Fig. 1d and e) shows that the MWCNTs were not fully aligned and somewhat entangled. As we shall detail below, the entanglement can improve mechanical properties. Thermo-oxidation experiments, carried out by heating the nanotubes in an oxygen rich atmosphere, showed that the blocks are stable up to 560 C. The broadening of the weight loss derivative curve over 650 C revealed the presence of a small amount of secondary carbonaceous products (Fig. 2). The TEM images reported in Fig. 3 confirmed that the structure is formed by multiwalled carbon nanotubes.
Fig. 3. TEM images: (a) the diameter of the MWCNTs was of the order of tens of nanometers; (b) the presence of iron particles was observed at the end of nanotubes; (c) the distance between walls in the MWCNT was ˚. about 3.4 A
Fig. 3a shows the presence of iron particles along the nanotubes. Fig. 3b shows the iron particles near the top end (i.e. the end far from the substrate) of the nanotubes. The presence of an amount of iron of about 6% in weight had been checked by EDX (energy dispersive X-ray) and thermogravimetric analyses. EDX showed that a higher concentration of iron was found in the top region of the MWCNT mat. These findings suggested that the growth
Letters to the Editor / Carbon 45 (2007) 1105–1136
Fig. 4. XRD analysis. The most significant graphite-like Bragg peaks are marked with Miller indices. Catalysts related peaks are indicated by (*).
occurred through a tip-growth mechanism in which the iron particle condensed and moved leaving behind the nanotubes [9]. However, as a recent report for a similar system suggested that a base-growth mechanism maybe at work [10], this issue needs further investigation. The TEM images (Fig. 3) showed that the diameter of the MWCNT was in the tens of nanometers range. The diameter can be tailored in the range 10–100 nm with a negligible variation of growth rate by modifying the deposition conditions. The various walls constituting the nanotube structures can be observed in Fig. 3c. The not complete graphitization of the walls visible in the TEM image and the presence of small amount of amorphous carbon, confirmed by X-ray diffraction analysis, can be ascribed to the low efficiency of iron catalyst in growing well-graphitized MWCNTs from camphor [11]. The XRD (Fig. 4) reported the typical graphite-like peaks and some feature related to the presence of catalyst, although the (0 0 2) peak showed a lower intensity and a broader shape than the defect free graphite. The described MWCNTs showed diameters, density and alignment comparable to results from previous works using camphor as carbon precursor. Nevertheless, the mats length, the camphor conversion efficiency (30%) and the growth rate were the highest reported so far. Indeed, a previous work [12] reported a 200 lm MWCNT mat with 220 nm s1 growth rate and 25% reagents conversion, whereas the present growth rate is about 500 nm s1 and the conversion rate is higher (see above). The hydrophobic properties of the blocks were checked by dynamic tests in which water drops were dropped onto the surface. A movie in which a macroscopic water drop (about 2.7 mm in diameter and 10 ll volume) bounces on a MWCNT mat is reported in the supplementary material. Such behavior had previously been reported for functionalized non-self standing nanotubes only [13]. Fig. 5 shows a time sequence, extracted from the movie, of the water drop bouncing and sliding with negligible friction over the surface. This witnesses the fully hydrophobic behavior of the MWCNT mat surface in dynamic condition. This prop-
1135
erty, connected to the availability of self-standing large area blocks of carbon nanotubes, can lead to several mechanical and coating applications in the field of fluidodynamics too. Outstanding mechanical characteristics have been reported for single nanotubes [14,15], but a recent work had shown that the axial and shear modulus decreased significantly as the diameter of the nanotube bundles increased [16]. Further investigations were carried out by several authors using standard [17] or modified AFM [18] set-ups. From an application point of view, however, such work suffered from two significant drawbacks, as all measurements were performed on the microscopic scale and the values obtained strongly influenced by the assumptions made. Hence the rather scattered values can give little information on what could be expected for large-scale systems. In order to overcome these drawbacks we have performed, for the first time to the best of our knowledge, a macroscopic investigation of the mechanical behavior of self-standing carbon nanotube mats. The compressive strength of several samples of free standing MWCNT blocks was tested by measuring the maximum stress sustainable by the material under crush loading until shattering of the specimen: the blocks can withstand a pressure of 25 MPa (i.e. about 250 atmospheres) without loss of integrity. Attempts were also made to measure the maximum tensile stress that the mat can withstand, in the direction parallel to the CNT growth direction, by pulling the upper and lower sides after having glued them to a holder. As in all cases the failure of glue was reached before that of the mat, more sophisticated measurements are planned to evaluate this property on a macroscopic scale. In our mats the more critical issue from a mechanical point of view was the weak resistance to traction in a direction parallel to the surface. The modulus of rupture (MOR), measured by third-point loading method, was determined as1 MOR ¼
1:5Pl wd 2
ð1Þ
where P was the maximum load at failure (expressed in kg), l was the span length between the two lower supports, d and w were respectively the depth and the width of the nanotube mat (dimensions in millimeters). The values obtained was 1.25 MPa. Although comparable to previous reports [6] it has to be improved for mechanical applications. We attributed this low value to the fact that the nanotubes were piled up almost vertically in the direction orthogonal to the substrate with a lack of interaction between neighboring nanotubes in the horizontal plane. On the other hand this peculiar structure is responsible of mat elasticity. Indeed hardness measurements performed on both sides of 1
‘‘American Society for Testing and Materials’’ standard test method for flexural strength of concrete (using simple beam with third-point loading) – ASTM C78.
1136
Letters to the Editor / Carbon 45 (2007) 1105–1136
Fig. 5. Hydrophobicity of nanotube mat. Time sequence images of a water droplet bouncing on the substrate side of the nanotube mat (impact speed of about 0.2 m s1). These frames were extracted from the movie provided in the supplementary material.
a self-standing CNT mat by a ‘‘shore A durometer’’2 revealed a value (60–70 units) comparable to that of the rubber in pneumatic tires. In order to overcome the limitation due to the limited modulus of rupture we are currently investigating growth conditions expected to lead to a much higher entanglement of the nanotubes. When this goal will be reached, the mechanical properties will be definitely improved and the resistance to horizontal traction increased. In summary, we have grown millimeters-thick large area self-standing blocks of carbon nanotubes having outstanding mechanical properties and a highly hydrophobic behavior. In particular, the possibility to homogeneously cover curved surfaces discloses new interesting future applications.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.carbon.2006. 12.019.
References [1] Li Y, Wang K, Wei J, Gu Z, Wang Z, Luo J, et al. Tensile properties of long aligned double-walled carbon nanotube strands. Carbon 2005;43:31–5. [2] Baughman RH, Zakhidov AA, De Heer WA. Carbon nanotubes – the route toward applications. Science 2002;297:787–92. [3] Nishio M, Sawaya S, Akita S, Nakayama Y. Carbon nanotube oscillators toward zeptogram detection. Appl Phys Lett 2005;86: 133111–3. [4] Jiang KL, Li QQ, Fan SS. Spinning continuous carbon nanotube yarns. Nature 2002;419:801.
2 ‘‘American Society for Testing and Materials’’ standard test method for rubber property (durometer hardness) – ASTM D2240-00.
[5] Hata K, Futaba DN, Mizunoc K, Namai T, Yumura M, Iijima S. Water-assisted highly efficient synthesis of impurity-free single-walled carbon nanotubes. Science 2004;306:1362–4. [6] Srivastava A, Srivastava ON, Talapatra S, Vajtai R, Ajayan PM. Carbon nanotube filters. Nat Mater 2004;3:610–4. [7] Musso S, Fanchini G, Tagliaferro A. Growth of vertically aligned carbon nanotubes by CVD by evaporation of carbon precursors. Diamond Relat Mater 2005;14:784–9. [8] Cao PJ, Zhu DL, Ma XC, Bai XJ. The effect of substrate morphology on the formation of large-scale well-aligned carbon nanotube film. Mater Chem Phys 2006;97:182–7. [9] Zhang X, Cao A, Wei B, Li Y, Wei J, Xu C, et al. Rapid growth of well-aligned carbon nanotube arrays. Chem Phys Lett 2002;362: 285–90. [10] Pinault M, Pichot V, Khodja H, Launois P, Reynaud C, Mayne-L’Hermite M. Evidence of sequential lift in growth of aligned multiwalled carbon nanotube multilayers. Nanoletters 2005;5:2394–8. [11] Kumar M, Ando Y. Controlling the diameter distribution of carbon nanotubes grown from camphor on a zeolite support. Carbon 2005;43:533–40. [12] Kumar M, Ando Y. A simple method of producing aligned carbon nanotubes from an unconventional precursor – Camphor. Chem Phys Lett 2003;374:521–6. [13] Lau KKS, Bico J, Teo KBK, Chhowalla M, Amaratunga GAJ, Milne WI, et al. Superhydrophobic carbon nanotube forests. Nanoletters 2003;3:1701–5. [14] Treacy MMJ, Ebbesen TW, Gibson TM. Exceptionally high Young’s modulus observed for individual carbon nanotubes. Nature 1996;381: 678–80. [15] Wong EW, Sheehan PE, Lieber CM. Nanobeam mechanics: elasticity, strength, and toughness of nanorods and nanotubes. Science 1997;277:1971–5. [16] Salvetat JP, Briggs GAD, Bonard JM, Bacsa RR, Kulik AJ, Sto¨ckli T, et al. Elastic and shear moduli of single-walled carbon nanotube ropes. Phys Rev Lett 1999;82(5):944–7. [17] Walters DA, Ericson LM, Casavant MJ, Liu J, Colbert DT, Smith KA, et al. Elastic strain of freely suspended single-wall carbon nanotubes ropes. Appl Phys Lett 1999;74:3803–5. [18] Yu MF, Lourie O, Dyer M, Moloni K, Kelly T. Strength and breaking mechanism of multi-walled carbon nanotubes under tensile load. Science 2000;287:637–40.
1133
[6] Fang HT, Jing CZ, Zhang DY, Jae HJ, Yoo DH. A Si–Mo fused slurry coating for oxidation protection of carbon–carbon composites. J Mater Sci Lett 2001;20:175–7. [7] Zhang YL, Li HJ, Fu QG, Li KZ, W J, Wang PY. A C/SiC gradient oxidation protective coating for carbon/carbon composites. Surf Coat Technol 2006;201(6):3491–5.
Macroscopic growth of carbon nanotube mats and their mechanical properties Simone Musso *, Samuele Porro, Mauro Giorcelli, Angelica Chiodoni, Carlo Ricciardi, Alberto Tagliaferro Politecnico di Torino, Dipartimento di Fisica, C.so Duca degli Abruzzi 24, 10129 Torino, Italy Received 12 September 2006; accepted 15 December 2006 Available online 27 December 2006
Since carbon nanotubes have been recognized as a ultrastrong material, both by simulations and experimental measurements [1], various applications have been proposed for their use in thermal and electrical enhancement [2,3] and in the area of strength and toughness as reinforcement for composite materials [4]. Nevertheless, significant progresses in developing carbon nanotube-composites technology are limited by the lack of a simple, economical and high conversion/deposition rate production method for carbon nanotubes. Because of the several advantages of the CVD techniques [5,6], much efforts have been devoted to the optimization of this method, though a real breakthrough in this field has yet to be achieved. Here we report the growth and the properties of self standing large area millimeters-thick sheets of as-deposited multiwall carbon nanotubes (MWCNTs). Such sheets were grown by catalytical chemical vapor deposition (CVD) at a growth rate up to 0.5 lm s1 over an area limited only by the deposition system size (up to 100 cm2 at present). With this method curved surfaces can be straightforwardly coated. The sheets showed good thermal stability and interesting mechanical properties such as elasticity and resistance under compression and were totally hydrophobic. These results show that the route towards the use of carbon nanotubes as a structural material is now open. The thick layers (over 2 mm) of MWCNTs, termed ‘‘mats’’ from here on, were grown in a CVD reactor formed by a horizontal quartz tube housed in a cylindrical furnace
*
Corresponding author. Fax: +39 011 5647399. E-mail address: [email protected] (S. Musso).
0008-6223/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2006.12.019
in which a constant nitrogen gas flow rate was maintained at pressures just above atmospheric [7]. The carbon (camphor) and the catalyst (ferrocene) sources were mixed in the ratio 20:1 in a pyrex flask. The precursors solid mixture, which contained no solvents, was melted and evaporated directly from the flask by heating it at 220 C with a heater plate. The flask connection between the quartz tube and the nitrogen gas source was realized with a T-joint. The gas evaporated from the flask was injected into the nitrogen flux, that carried it to the substrate region. Deposition of the MWCNTs took place on non pre-etched bare silicon substrates. The temperature in the deposition region was kept at 850 C. The reactor quartz tube internal diameter was 4.2 cm, and the length of the uniform temperature region of about 30 cm. The size of the substrate housed in such a tube was therefore 4 · 30 cm (substrate divided in two pieces 4 · 15 each, due to the size of commercial silicon wafers). The temperature gradient present at the edges of the uniform temperature region caused small gas turbulences reducing the uniform deposition area to 4 · 25 cm (100 cm2). As the current version of the system does not allow a continuous feed of the reagents, the process time lasted two hours at most. At the end of the process the substrates (up to 100 cm2 in total) were covered by a few millimeters thick uniform MWCNT mat. A consistent deposition of MWCNTs occurred also on every curved surface of the silicon substrate (see Fig. 1b) and on the inner surface of the quartz tube, although on the latter the deposition is in the form of non ordered compact and irregular tangles. The deposition rate reached 0.5 lm s1 and the mass production rate (considering only the mats grown on Si substrate) exceeded 1 g h1, while the total
1134
Letters to the Editor / Carbon 45 (2007) 1105–1136
Fig. 2. Thermo-gravimeter analysis. Thermo-oxidation experiments suggested the presence of secondary carbonaceous products (high T derivative shoulder).
Fig. 1. Images of the MWCNT mat: (a) photograph of a block of the MWCNT mat evidencing its thickness (the Euro coin is 2.33 mm high); (b) SEM image of a CNT-coated substrate edge; (c) SEM image showing the macroscopic vertical alignment of the MWCNT; (d,e) SEM images with increasing magnification, showing the diameter distribution and the entanglement of nanotubes.
conversion rate is of over 30% in weight. The role of silicon in the mat formation remains unexplained, though literature suggests [8] that the morphology of the substrate wafers plays a role in the nucleation of the CNT and in the chemical/physical interaction with the catalyst particle. The MWCNT mat was mechanically delaminated from the substrate with a razor blade and blocks such as that shown in Fig. 1a were obtained. The figure shows that the MWCNT mat is thick and self standing. The substrate side of the mat blocks showed an optically flat surface. The density of the mat was measured by weighing a block of MWCNT versus its volume, and a value of 0.5 g/cm3 was obtained. This value is lower than the density expected from an isolated MWCNT. This is consistent with the fact that MWCNTs organized in mats were not closely packed, but spaced, as shown by SEM (Fig. 1d and e). Although Fig. 1c shows a macroscopic alignment of the MWCNTs, a closer look at the mats (Fig. 1d and e) shows that the MWCNTs were not fully aligned and somewhat entangled. As we shall detail below, the entanglement can improve mechanical properties. Thermo-oxidation experiments, carried out by heating the nanotubes in an oxygen rich atmosphere, showed that the blocks are stable up to 560 C. The broadening of the weight loss derivative curve over 650 C revealed the presence of a small amount of secondary carbonaceous products (Fig. 2). The TEM images reported in Fig. 3 confirmed that the structure is formed by multiwalled carbon nanotubes.
Fig. 3. TEM images: (a) the diameter of the MWCNTs was of the order of tens of nanometers; (b) the presence of iron particles was observed at the end of nanotubes; (c) the distance between walls in the MWCNT was ˚. about 3.4 A
Fig. 3a shows the presence of iron particles along the nanotubes. Fig. 3b shows the iron particles near the top end (i.e. the end far from the substrate) of the nanotubes. The presence of an amount of iron of about 6% in weight had been checked by EDX (energy dispersive X-ray) and thermogravimetric analyses. EDX showed that a higher concentration of iron was found in the top region of the MWCNT mat. These findings suggested that the growth
Letters to the Editor / Carbon 45 (2007) 1105–1136
Fig. 4. XRD analysis. The most significant graphite-like Bragg peaks are marked with Miller indices. Catalysts related peaks are indicated by (*).
occurred through a tip-growth mechanism in which the iron particle condensed and moved leaving behind the nanotubes [9]. However, as a recent report for a similar system suggested that a base-growth mechanism maybe at work [10], this issue needs further investigation. The TEM images (Fig. 3) showed that the diameter of the MWCNT was in the tens of nanometers range. The diameter can be tailored in the range 10–100 nm with a negligible variation of growth rate by modifying the deposition conditions. The various walls constituting the nanotube structures can be observed in Fig. 3c. The not complete graphitization of the walls visible in the TEM image and the presence of small amount of amorphous carbon, confirmed by X-ray diffraction analysis, can be ascribed to the low efficiency of iron catalyst in growing well-graphitized MWCNTs from camphor [11]. The XRD (Fig. 4) reported the typical graphite-like peaks and some feature related to the presence of catalyst, although the (0 0 2) peak showed a lower intensity and a broader shape than the defect free graphite. The described MWCNTs showed diameters, density and alignment comparable to results from previous works using camphor as carbon precursor. Nevertheless, the mats length, the camphor conversion efficiency (30%) and the growth rate were the highest reported so far. Indeed, a previous work [12] reported a 200 lm MWCNT mat with 220 nm s1 growth rate and 25% reagents conversion, whereas the present growth rate is about 500 nm s1 and the conversion rate is higher (see above). The hydrophobic properties of the blocks were checked by dynamic tests in which water drops were dropped onto the surface. A movie in which a macroscopic water drop (about 2.7 mm in diameter and 10 ll volume) bounces on a MWCNT mat is reported in the supplementary material. Such behavior had previously been reported for functionalized non-self standing nanotubes only [13]. Fig. 5 shows a time sequence, extracted from the movie, of the water drop bouncing and sliding with negligible friction over the surface. This witnesses the fully hydrophobic behavior of the MWCNT mat surface in dynamic condition. This prop-
1135
erty, connected to the availability of self-standing large area blocks of carbon nanotubes, can lead to several mechanical and coating applications in the field of fluidodynamics too. Outstanding mechanical characteristics have been reported for single nanotubes [14,15], but a recent work had shown that the axial and shear modulus decreased significantly as the diameter of the nanotube bundles increased [16]. Further investigations were carried out by several authors using standard [17] or modified AFM [18] set-ups. From an application point of view, however, such work suffered from two significant drawbacks, as all measurements were performed on the microscopic scale and the values obtained strongly influenced by the assumptions made. Hence the rather scattered values can give little information on what could be expected for large-scale systems. In order to overcome these drawbacks we have performed, for the first time to the best of our knowledge, a macroscopic investigation of the mechanical behavior of self-standing carbon nanotube mats. The compressive strength of several samples of free standing MWCNT blocks was tested by measuring the maximum stress sustainable by the material under crush loading until shattering of the specimen: the blocks can withstand a pressure of 25 MPa (i.e. about 250 atmospheres) without loss of integrity. Attempts were also made to measure the maximum tensile stress that the mat can withstand, in the direction parallel to the CNT growth direction, by pulling the upper and lower sides after having glued them to a holder. As in all cases the failure of glue was reached before that of the mat, more sophisticated measurements are planned to evaluate this property on a macroscopic scale. In our mats the more critical issue from a mechanical point of view was the weak resistance to traction in a direction parallel to the surface. The modulus of rupture (MOR), measured by third-point loading method, was determined as1 MOR ¼
1:5Pl wd 2
ð1Þ
where P was the maximum load at failure (expressed in kg), l was the span length between the two lower supports, d and w were respectively the depth and the width of the nanotube mat (dimensions in millimeters). The values obtained was 1.25 MPa. Although comparable to previous reports [6] it has to be improved for mechanical applications. We attributed this low value to the fact that the nanotubes were piled up almost vertically in the direction orthogonal to the substrate with a lack of interaction between neighboring nanotubes in the horizontal plane. On the other hand this peculiar structure is responsible of mat elasticity. Indeed hardness measurements performed on both sides of 1
‘‘American Society for Testing and Materials’’ standard test method for flexural strength of concrete (using simple beam with third-point loading) – ASTM C78.
1136
Letters to the Editor / Carbon 45 (2007) 1105–1136
Fig. 5. Hydrophobicity of nanotube mat. Time sequence images of a water droplet bouncing on the substrate side of the nanotube mat (impact speed of about 0.2 m s1). These frames were extracted from the movie provided in the supplementary material.
a self-standing CNT mat by a ‘‘shore A durometer’’2 revealed a value (60–70 units) comparable to that of the rubber in pneumatic tires. In order to overcome the limitation due to the limited modulus of rupture we are currently investigating growth conditions expected to lead to a much higher entanglement of the nanotubes. When this goal will be reached, the mechanical properties will be definitely improved and the resistance to horizontal traction increased. In summary, we have grown millimeters-thick large area self-standing blocks of carbon nanotubes having outstanding mechanical properties and a highly hydrophobic behavior. In particular, the possibility to homogeneously cover curved surfaces discloses new interesting future applications.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.carbon.2006. 12.019.
References [1] Li Y, Wang K, Wei J, Gu Z, Wang Z, Luo J, et al. Tensile properties of long aligned double-walled carbon nanotube strands. Carbon 2005;43:31–5. [2] Baughman RH, Zakhidov AA, De Heer WA. Carbon nanotubes – the route toward applications. Science 2002;297:787–92. [3] Nishio M, Sawaya S, Akita S, Nakayama Y. Carbon nanotube oscillators toward zeptogram detection. Appl Phys Lett 2005;86: 133111–3. [4] Jiang KL, Li QQ, Fan SS. Spinning continuous carbon nanotube yarns. Nature 2002;419:801.
2 ‘‘American Society for Testing and Materials’’ standard test method for rubber property (durometer hardness) – ASTM D2240-00.
[5] Hata K, Futaba DN, Mizunoc K, Namai T, Yumura M, Iijima S. Water-assisted highly efficient synthesis of impurity-free single-walled carbon nanotubes. Science 2004;306:1362–4. [6] Srivastava A, Srivastava ON, Talapatra S, Vajtai R, Ajayan PM. Carbon nanotube filters. Nat Mater 2004;3:610–4. [7] Musso S, Fanchini G, Tagliaferro A. Growth of vertically aligned carbon nanotubes by CVD by evaporation of carbon precursors. Diamond Relat Mater 2005;14:784–9. [8] Cao PJ, Zhu DL, Ma XC, Bai XJ. The effect of substrate morphology on the formation of large-scale well-aligned carbon nanotube film. Mater Chem Phys 2006;97:182–7. [9] Zhang X, Cao A, Wei B, Li Y, Wei J, Xu C, et al. Rapid growth of well-aligned carbon nanotube arrays. Chem Phys Lett 2002;362: 285–90. [10] Pinault M, Pichot V, Khodja H, Launois P, Reynaud C, Mayne-L’Hermite M. Evidence of sequential lift in growth of aligned multiwalled carbon nanotube multilayers. Nanoletters 2005;5:2394–8. [11] Kumar M, Ando Y. Controlling the diameter distribution of carbon nanotubes grown from camphor on a zeolite support. Carbon 2005;43:533–40. [12] Kumar M, Ando Y. A simple method of producing aligned carbon nanotubes from an unconventional precursor – Camphor. Chem Phys Lett 2003;374:521–6. [13] Lau KKS, Bico J, Teo KBK, Chhowalla M, Amaratunga GAJ, Milne WI, et al. Superhydrophobic carbon nanotube forests. Nanoletters 2003;3:1701–5. [14] Treacy MMJ, Ebbesen TW, Gibson TM. Exceptionally high Young’s modulus observed for individual carbon nanotubes. Nature 1996;381: 678–80. [15] Wong EW, Sheehan PE, Lieber CM. Nanobeam mechanics: elasticity, strength, and toughness of nanorods and nanotubes. Science 1997;277:1971–5. [16] Salvetat JP, Briggs GAD, Bonard JM, Bacsa RR, Kulik AJ, Sto¨ckli T, et al. Elastic and shear moduli of single-walled carbon nanotube ropes. Phys Rev Lett 1999;82(5):944–7. [17] Walters DA, Ericson LM, Casavant MJ, Liu J, Colbert DT, Smith KA, et al. Elastic strain of freely suspended single-wall carbon nanotubes ropes. Appl Phys Lett 1999;74:3803–5. [18] Yu MF, Lourie O, Dyer M, Moloni K, Kelly T. Strength and breaking mechanism of multi-walled carbon nanotubes under tensile load. Science 2000;287:637–40.