CHIN. PHYS. LETT. Vol. 26, No. 11 (2009) 116103
Synthesis and Properties of Magnetic Composites of Carbon Nanotubes/Fe Nanoparticle *
XU Mei-Hua(徐美华)1,2 , QI Xiao-Si(祁小四)1 , Z
HONG Wei(钟伟)1** , YE Xiao-Juan(叶小娟)1 , DENG Yu(邓昱)1 , AU Chaktong(区泽棠)3 , JIN Chang-Qing(靳长清)1 , YANG Zai-Xing(杨再兴)1 , DU You-Wei(都有为)1
National Laboratory of Solid State Microstructures and Department of Physics, Nanjing University, Nanjing 210093 2 Department of Applied Physics, Nanjing University of Technology, Nanjing 210009 3 Chemistry Department, Hong Kong Baptist University, Hong Kong
(Received 5 May 2009)
Magnetic composites of carbon nanotubes (CNTs) are synthesized by the in situ catalytic decomposition of benzene at temperatures as low as 400 ?C over Fe nanoparticles (mean grain size = 26 nm) produced by sol-gel fabrication and hydrogen reduction. The yield of CNT composite is up to about 3025% in a run of 6 h. FESEM and HRTEM investigations reveal that one-dimensional carbon species are produced in a large quantity. A relatively high value of magnetization is observed for the composite due to the encapsulation of ferromagnetic Fe3 C and/or -Fe. The method is suitable for the mass-production of CNT composites that contain magnetic nanoparticles.
61. 46. Fg, 68. 37. Ma, 68. 37. Og of high magnetization, it is necessary to remove the support and promoter that are not magnetic. The puri?cation processes can be costly, and if conditions were harsh, damages to the CNT materials (such as surface alteration and the formation of structural defects) could occur. It is hence desirable to develop a low-temperature route that involves a support-free catalyst. Recently, we reported the in situ synthesis of double helical carbon nano?bers and nanotubes as well as plait-like carbon nanocoils by the decomposition of acetylene over metal nanoparticles at relatively low temperatures (< 450? C). According to Kumar and Ando, benzene is ideal for CNT generation because of its hexagonal structure. Since benzene is cheap and abundant and is relatively safe to handle, it has been used as a carbon source for CNT generation. Fan et al. and Cheng et al. synthesized CNTs from benzene catalytically in the 1100–1200? C range via an improved “?oating catalyst” approach. Using chlorated benzene as a precursor, Ruitao et al. synthesized FeNi-?lled CNTs at 860? C. Recently, Mahanandia and Nanda reported a one-step technique for the preparation of aligned arrays of CNTs at 700? C from xylene, cyclohexane, camphor, hexane, toluene, pyridine and benzene. In terms of cost e?ectiveness, the methods reported so far for large-scale CNT synthesis from benzene cannot be considered to be satisfactory. In this Letter, we report a CVD method that is more e?ective. Using argon as a carrier and over catalytically active Fe nanoparticles, we obtain magnetic CNT composites
Carbon nanotubes (CNTs) have attracted enormous attention since the landmark paper by Iijima. The materials are unparalleled in Young’s modulus and strength-to-density ratio. They have a very high mechanical strength, excellent thermal conductivity, and a large elastic modulus and elastic strain, and potentially have wide applications in areas such as composites, nanoelectronics, spintronics, and nanoelectromechanical systems.[2?4] As for the carbon nanospecies that are decorated with magnetic materials, potential applications can be meaningful. It is envisaged that magnetic CNTs in the form of capsules or nanosubmarines can be used in the delivery of drugs to a desired location of the human body. Methods such as laser vaporization, arc discharge, and chemical vapor deposition (CVD) have been adopted for the synthesis of CNTs. The catalytic fabrication of CNTs by means of CVD was ?rst reported by Yacaman et al. In general, CNTs are produced over transition metals (e.g., Fe, Co, Ni or their alloys) at temperatures above 700? C[10?15] via the catalytic decomposition of hydrocarbons (e.g., acetylene, toluene and pyridine) or carbon monoxide. For the improvement of catalytic activity, materials such as Al2 O3 , SiO2 , TiO2 and zeolite are commonly used as catalyst supports. Also, for the promotion of CNT growth, a sulfur-containing additive such as thiophene was added to liquid toluene, xylene, and benzene.[16,17] The CVD methods, however, are associated with high reaction temperatures and a low CNT yield. To obtain CNT composites
* Supported by the National Natural Science Foundation of China under Grant No 10674059, the National High Technology Research and Development Program of China under Grant No 2007AA021805, and the National Basic Research Program of China under Grant No 2005CB623605. ** Email: email@example.com c ○ 2009 Chinese Physical Society and IOP Publishing Ltd
CHIN. PHYS. LETT. Vol. 26, No. 11 (2009) 116103
from benzene at 400? C. The method is simple and energy e?cient and there is no need to use a promoter or a catalyst support. There were two steps involved in CNT generation: (a) the preparation of the catalyst and (b) the synthesis of CNTs from benzene. The schematic of the experimental system is depicted in Fig. 1. To prepare the ferric oxide precursor, 0.03 mol FeCl2 · 4H2 O and 0.045 mol citric acid monohydrate were mixed with 250 ml absolute ethanol, and the mixture was stirred at 60? C for 6 h. With the evaporation of ethanol at 80? C and heating of the residual at 500? C in air for 4 h, the xerogel was turned to a powder of ferric oxide. To obtain carbon composites in a large quantity, a quartz tube that was 6 cm in diameter and 75 cm in length (equipped with temperature and gas-?ow controls) was used as the reactor. The ferric oxide powder (40 mg, spread on a ceramic plate) was placed inside the reactor, and in situ reduced to Fe nanoparticles under H2 at 500? C for 4 h. Then benzene (kept in a three-necked ?ask maintained at 70? C over a water bath) was carried into the reactor by a ?ow of argon. The decomposition of benzene was conducted at 400? C for 6 h over the Fe nanoparticles, and after cooling to room temperature (RT), about 0.905 g of “as-prepared” CNT composite (black in color) was collected. To obtain a puri?ed CNT sample, the as-prepared CNT composite was immersed in aqueous HCl solution at 80? C for 18 h. The as-obtained CNTs were then washed with distilled water and absolute ethanol (three cycles) followed by drying in air at 80? C.
H2 Ar Ceramic plate Quartz tube Valve Furnace Waste gas Thermocouple
yield of CNTs using benzene as a carbon source has not been reported before. The thermogravimetric (TG, Perkin-Elmer TGA7 series, heated in air at a rate of 10? C/min) curve of asprepared CNT composite shows a weight loss in the temperature range from RT to 500? C and the leftover is about 97.67% (Fig. 2). The loss is ascribed to amorphous carbon oxidation and the residue may be attributed to CNTs and Fe oxide. Higher than 500? C, the weight lost sharply and the residue is 3.14 wt%. The loss can be ascribed to the CNTs oxidation and the residue is attributed to ferric oxide. Therefore, we could estimate that the as-prepared CNT composite contains 3.27% amorphous carbon. Accordingly, we recorded a yield of carbon products up to about 3025% when we deduct amorphous carbon from carbon products.
Fig. 2. Thermogravimetric (TG) curve of the as-prepared CNT composite.
Intensity (arb. units)
Graphite + α-Fe ? Fe3C
Benzene Water bath
, , ,
Fig. 1. Schematic of experimental system.
In this study, 40 mg of the ferric oxide powder (contain about 28 mg Fe catalyst) was used and about 905 mg of carbon products was obtained in each run. The percentage of carbon yield is de?ned as follows: Carbon yield(%) = 100 × tot ? cat %, cat (1)
Fig. 3. XRD pattern of the as-prepared CNT composite.
where cat is the initial amount of the catalyst and tot the total mass after reaction. Accordingly, we recorded a yield of carbon products up to about 3130%. To the best of our knowledge, such a high
The x-ray di?raction (XRD) of samples was conducted at RT on a di?ractometer (Model D/Max-RA, Rigaku, Japan) for phase identi?cation using Cu radiation. The XRD pattern of as-prepared CNT composite is shown in Fig. 3. The di?raction peaks can be indexed to phases of graphite, -Fe, and Fe3 C. The size of the catalyst nanoparticles is estimated to be about 26 nm according to the Sherrer formula. The results of elemental analysis of the as-prepared CNT
CHIN. PHYS. LETT. Vol. 26, No. 11 (2009) 116103
composite reveal the absence of hydrogen (Elementar Vario MICRO, Germany). Accordingly, we deduce that there is complete decomposition of benzene during the formation of graphitic carbon.
Fig. 4. FESEM micrograph of the as-prepared CNT composite.
As shown in Fig. 5(b), there is the encapsulation of an Fe nanoparticle at the CNT tip. The CNT wall consists of many layers of carbon sheets, and the lattice fringes can clearly be seen. The average value of spacing of the graphitic sheets is about 0.34 nm, consistent with the result of XRD investigation. The average outer diameter of the CNTs is about 28 nm and the average wall thickness is about 8.2 nm. It is clear that despite a reaction temperature of 400? C, the CNTs obtained are crystalline. It is apparent that the crystal orientation of the inner wall is more regular than that of the outer wall. Nevertheless, some CNTs are with a hollow core (Fig. 5(c)), plausibly due to the detachment of Fe nanoparticles during the ultrasonic treatment. Figure 5(c) shows two MWCNTs of about a 30 nm outer diameter. In Fig. 5(d), one can see the entrapment of Fe particles inside an MWCNT segment.
50 nm 10 nm (c) (d)
Fig. 5. (a) HRTEM micrograph of the as-prepared CNT composite, (b) high-magnification of a tip containing an encapsulated Fe particle (inset: the magnified image of the area indicated in (b)), (c) HRTEM image of the graphitic wall of an MWCNT segment with a hollow core, and (d) an MWCNT segment with entrapped Fe particles.
Fig. 6. Raman spectrum of as-prepared CNTs.
The morphologies of samples were examined over a high-resolution transmission electron microscope (HRTEM, model JEOL-2010, Japan) operated at an accelerating voltage of 200 kV, and ?eld-emission scanning electron microscope (FE-SEM model 1530VP, LEO, Germany; equipped with a JEOL JFC-1600 auto ?ne-coater) operated at an accelerating voltage of 5 kV. A representative FESEM image of asprepared CNT composite is shown in Fig. 4. One can see that CNTs of one-dimensional nanostructures were obtained in a large quantity. There are tangled CNTs along with catalyst nanoparticles and amorphous carbon. The outer diameters of CNTs range from 30 to 90 nm, indicative of multi-walled CNTs (MWCNTs). The tubular structures were further characterized by HRTEM (Fig. 5). For the HRTEM investigation, the sample was treated ultrasonically for about 5 min in alcohol solution (for good dispersion) before being supported on a copper grid. Figures 5(a) and 5(b) reveal that the carbon materials are indeed MWCNTs.
Fig. 7. Typical magnetization curves of the as-prepared CNT composite.
Raman spectroscopy investigations were performed over a Jobin-Yvon Labram HR800 instrument with 514.5 nm Ar+ -laser excitation. Raman spectroscopy is a non-destructive technique that provides information about the crystal structure and disorder of carbon samples.[25,26] The Raman spectrum of asprepared CNT composite is shown in Fig. 6. Two strong peaks typical of graphitized CNTs are observed, one at about 1356.1 cm?1 and the other at about 1581.7 cm?1 , and there is no Raman feature be-
CHIN. PHYS. LETT. Vol. 26, No. 11 (2009) 116103
low 400 cm?1 . It is clear that the CNTs synthesized are mostly MWCNTs, in agreement with the result of the HRTEM investigation (Fig. 5). The intensities of the two bands are roughly the same. In other words, there are defects in the as-prepared CNT composite. Furthermore, compared to the Raman peaks of crystalline graphite, the two peaks in Fig. 6 are obviously up-shifted by several cm?1 . The phenomenon suggests that there is the presence of compress stress. The magnetic properties of samples were measured at 300 K over a Quantum Design MPMS SQUID magnetometer (Quantum Design MPMS-XL, U.S.A.). Figure 7 shows the magnetization dependence of asprepared CNT composite on a ?eld measured at RT. The saturation magnetization and the coercivity C are 4.56 emu/g and 284.6 Oe, respectively. According to the yield of CNTs, the Fe content in the sample is deduced to be about 3.09 wt%. Given that Fe3 C is the major Fe-containing phase (as revealed in the XRD study, Fig. 3) the amount of Fe3 C in the sample is deduced to be about 3.32 wt%. Based on the Fe3 C content and the fact that the magnetization of Fe3 C is about 128 emu/g at 300 K, we estimate that the saturation magnetization of the sample is about 4.24 emu/g, not far from the 4.56 emu/g value detected experimentally. The XRD pattern of Fig. 3 indicates that the sample contains three primary phases, that of graphite, -Fe, and Fe3 C. In other words, besides Fe3 C, there is the presence of -Fe particles. The magnetization of -Fe is about 220 emu/g at 300 K. As a result, the value obtained experimentally is higher than the one estimated solely based on the Fe3 C presence. The magnetization reaches saturation at about 1.5 T magnetic ?eld. Compared to the CNTs synthesized over nonmagnetic metal (such as Cu), the CNT composite containing ferromagnetic Fe3 C and/or -Fe nanoparticles exhibits a higher magnetization. In summary, magnetic CNT composites have been synthesized e?ciently at 400? C using benzene as a carbon source and Fe nanoparticles as a catalyst without the use of catalyst support and a sulfur-containing promoter. We obtain a yield of carbon products of about 3025% in a run of 6 h. The method is suitable for the mass synthesis of magnetic CNT composites from benzene.
 Iijima S 1991 Nature 354 56  Xu N, Ding J W, Chen H B and Ma M M 2009 Chin. Phys. Lett. 26 076102  Qiu W, Kang Y L, Lei Z K, Qin Q H and Li Q 2009 Chin. Phys. Lett. 26 080701  Baughman R H, Zakhidov A A and de Heer W A 2002 Science. 297 787  Korneva G, Ye H H, Gogotsi Y, Halverson D, Friedman G, Bradley J C and Kornev K G 2005 Nano Lett. 5 879  Thess A, Lee R, Nikolaev P, Dai H J, Petit P, Robert J, Xu C H, Lee Y H, Kim S G, Rinzler A G, Colbert D T, Scuseria G E, Tomanek D, Fischer J E and Smalley R E 1996 Science 273 483  Journet C, Maser W K, Bernier P, Loiseau A, delaChapelle M L, Lefrant S, Deniard P, Lee R and Fischer J E 1997 Nature 388 756  Terranova M L, Sessa V and Rossi M 2006 Chem. Vap. Deposit. 12 315  Yacam?n M J, Yoshida M M and Rend?n L M 1993 Appl. a o Phys. Lett. 62 202  Piedigrosso P, Konya Z, Colomer J F, Fonseca A, Van Tendeloo G and Nagy J B 2000 Phys. Chem. Chem. Phys. 2 163  Sarangi D and Karimi A 2003 Nanotechnology 14 109  Xie J N, Mukhopadyay K, Yadev J and Varadan V K 2003 Smart. Mater. Struct. 12 744  Cheng J P, Zhang X B, Liu F, Tu J P, Ye Y, Ji Y J and Chen C P 2003 Carbon 41 1965  Cheng J P, Zhang X B, Tu H, Tao X Y, Ye Y and Liu F 2006 Mater. Chem. Phys. 95 12  Ivanov V, Fonseca A, Nagy J B, Lucas A, Lambin P, Bernaerts D and Zhang X B 1995 Carbon 33 1727  Fan Y Y, Li F, Cheng H M, Su G, Yu Y D and Shen Z H 1998 J. Mater. Res. 13 2342  Cheng H M, Li F, Su G, Pan H Y, He L L, Sun X and Dresselhaus M S 1998 Appl. Phys. Lett. 72 3282  Tang N J, Zhong W, Gedanken A and Du Y W 2006 J. Phys. Chem. B 110 11772  Tang N J, Zhong W, Au C T, Gedanken A, Yang Y and Du Y W 2007 Adv. Funct. Mater. 17 1542  Tang N J, Yang Y, Lin K J, Zhong W, Au C T and Du Y W 2008 J. Phys. Chem. C 112 10061  Kumar M and Ando Y 2003 Chem. Phys. Lett. 374 521  Lv R T, Kang F Y, Wang W X, Wei J Q, Gu J L, Wang K L and Wu D H 2007 Carbon 45 1433  Mahanandia P and Nanda K K 2008 Nanotechnology 19 155602  Chai S P, Zein S H S and Mohamed A R 2007 Diamond Relat. Mater. 16 1656  Rao A M, Richter E, Bandow S, Chase B, Eklund P C, Williams K A, Fang S, Subbaswamy K R, Menon M, Thess A, Smalley R E, Dresselhaus G and Dresselhaus M S 1997 Science275 187  Li W Z, Zhang H, Wang C Y, Zhang Y, Xu L W, Zhu K and Xie S S 1997 Appl. Phys. Lett. 70 2684