- Email: [email protected]
Synthesis and conductivity measurement of carbon spheres by catalytic CVD using non-magnetic metal complexes
Synthetic Metals 161 (2011) 1590–1595
Contents lists available at ScienceDirect
Synthetic Metals journal homepage: www.elsevier.com/locate/synmet
Synthesis and conductivity measurement of carbon spheres by catalytic CVD using non-magnetic metal complexes Chien-Ming Lei a , Wei-Li Yuan b , Hsin-Chen Huang a , Shao-Wei Ho c , Chi-Jung Su c,d,∗ a Department of Chemical & Materials Engineering and Graduate Institute of Nanomaterials, Chinese Culture University, 55, Huagang Rd., Shilin Dist., Taipei 11114, Taiwan, ROC b Department of Chemical Engineering, Feng Chia University, 100, Wenhua Rd., Xitun Dist., Taichung 40724, Taiwan, ROC c School of Applied Chemistry, Chung Shan Medical University No. 110, Sec. 1, Jianguo N. Rd., Taichung 40201, Taiwan, ROC d Department of Medical Research, Chung Shan Medical University Hospital, No. 110, Sec. 1, Jianguo N. Rd., Taichung 40201, Taiwan, ROC
a r t i c l e
i n f o
Article history: Received 2 February 2011 Received in revised form 6 May 2011 Accepted 18 May 2011 Available online 15 June 2011 Keywords: Carbon spheres Catalytic chemical vapor deposition Non-magnetic metal complexes Conductivity
a b s t r a c t Carbon spheres (CSs) have been an important subject of research in recent years. Catalytic chemical vapor deposition (CCVD) was carried out in this work to synthesize solid-core CSs at mild temperatures from 720 to 810 ◦ C. Non-magnetic metal complexes of La, Nb, and Ti, dispersed on porous kaolin support, were tried out as catalyst. X-ray diffraction patterns revealed the graphitic structures of CSs. TEM analysis showed no encapsulated transition-metal nanoparticles inside the CSs. It was found by Raman spectra that the La catalyst resulted in CSs with higher graphitization. To examine the potential applications of CSs to the fields such as catalysis, electrochemistry, and electronic device, values of the thermal and electrical conductivity of the prepared CSs using different catalysts were measured and found to be comparable to those of the commercial carbon black. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Novel carbon-based nanomaterials such as carbon spheres (CSs), carbon nanotubes, and carbon nanofibers have been attracting researchers from various disciplines to study their exciting chemical, physical, and mechanical properties [1–8]. CSs from nano- to micro-meters in size are becoming important due to their potential applications, like carbon black and nanotube, in many fields such as electro-conductive coatings, dye-sensitized solar cells, catalyst supports for impregnation of metal nanoclusters, electrochemical capacitors, battery anodes, and elements for electrostatic charge dissipation or electrical heating [9–15]. Recently, hollow spheres of metal oxides such as SiO2 , TiO2 , and SnO2 , prepared using nano/micro-sized carbon spheres as template have also been reported, which further extends the applicability of CSs [16–18]. Carbonaceous nanomaterials are frequently prepared by physical vapor deposition, electrical arc-discharge, and laser vaporization [19–22]. Chemical vapor deposition (CVD) involves carbonization of organic materials through a pyrolytic process at high temperatures above 1000 ◦ C [23,24]. Also popular, the catalytic CVD (CCVD) is a catalyzed carbonization process which often
uses transition-metal complexes of Fe, Ni, or Co to convert hydrocarbons such as methane, ethane, acetylene (C2 H2 ), and carbon monoxide into carbon spheres, tubes, or fibers at a temperature well below 1000 ◦ C [25–31]. The transition-metal nanoparticles are then derived from the metal complex precursors and sometimes found encapsulated by the carbonized material, judging from the sample’s ferrimagnetism or TEM images [32]. In most cases though, the metal nanoparticles are found trapped within carbon nanotubes but separate from the CSs [24,32–35]. In the present study, we report the fabrication of nano/microcarbon spheres by CCVD with non-magnetic complexes of La, Nb, and Ti as catalyst supported by kaolin. The microstructures of CSs were characterized by powder XRD, SEM, and TEM. The magnetic properties of CSs were tested by measuring the magnetic susceptibility at different temperatures. One advantage of using La, Nb, or Ti metal complex as catalyst is the potential for adding superconductivity to CSs if the non-magnetic metal is encapsulated within. In addition, the thermal and electrical conductivity of the CSs prepared via different catalysts were measured and compared with those of the commercial carbon black. 2. Experimental
∗ Corresponding author at: School of Applied Chemistry, Chung Shan Medical University No. 110, Sec. 1, Jianguo N. Rd., Taichung 40201, Taiwan, ROC. Fax: +886 4 24510890. E-mail address: [email protected] (C.-J. Su). 0379-6779/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.synthmet.2011.05.023
2.1. Catalysts for CSs To prepare the kaolin support, the kaolin powder (from Showa) was dissolved in distilled water in a weight ratio of 1:1 and stirred
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595
1591
Fig. 1. Experimental setup for synthesizing carbon spheres.
for 30 min. The kaolin solution was then poured into a mold to dry up in oven at 50 ◦ C overnight. The kaolin chunk was then carved into a boat with a trough. To prepare the catalysts, metal complexes such as lanthanum(III) nitrate hexahydrate (from Strem), ammonium hexafluoroniobate(V) (from Acros), and titanium(IV) (triethanolaminato)isopropoxide solution (80 wt.% in 2-propanol, from Aldrich) were purchased and used as received. Firstly 1 g lanthanum (III) nitrate hexahydrate was dissolved in 5 ml distilled water and stirred for 1 h at room temperature. Next, the well-mixed solution was dripped onto a kaolin boat and dried in oven at 50 ◦ C overnight. The drying process helps the metal complex catalyst adsorb strongly to the kaolin support. The same procedures were repeated to prepare the rest catalysts by dripping ammonium hexafluoroniobate (V) and titanium (IV) (triethanolaminato) isopropoxide on kaolin. 2.2. CCVD for CSs The CCVD schematic for synthesizing CSs is shown in Fig. 1. The kaolin boat with catalyst was placed upside down at the center of the quartz tube in an electrical furnace. The boat was firstly heated in N2 at 17 sccm from room temperature to 650 ◦ C in 1 h. Next a second stream of C2 H2 at 70 sccm was introduced into the reactor. After injection of the carbon source the temperature was further raised to the desired reaction temperature, 720 or 810 ◦ C, in 10 min. The reaction proceeded for 1 h at the reached temperature. In the end the reactor was cooled down to room temperature by continuous flushing of nitrogen. The CSs as fine black powder were collected by ceramic boats placed beneath the inverted catalyzing kaolin boat. Same schematic was used for all the three transition metal catalysts. The names of the CSs produced using different catalysts of lanthanum, niobium, and titanium are abbreviated as CS-La, CS-Nb, and CS-Ti, respectively.
Fig. 2. SEM micrographs of (a) CS-La, (b) CS-Nb, and (c) CS-Ti; all synthesized at 720 ◦ C. The insets are enlarged portions.
2.3. Analyses for CSs SEM images of the synthesized CSs were taken by JEOL JSM 5200, while the TEM ones by JEOL II 1200EX. X-ray diffraction patterns of the sample powder were acquired by Bruker MXP3. Raman spectra were scanned by Jobin-Yvon T64000. Plots of magnetic susceptibility against temperature were obtained with Quantum Design PPMS-7. The thermal conductivity () of CSs was measured in a way similar to standard methods such as ASTM E1225-09 [36]. A simple system was set up with the sample sandwiched between two copper rods and enclosed in a heat insulating container. By flowing heat through the sample, can be calculated using the given dimensions of the sample, temperature difference across the sample, and the heat flowrate. The system was calibrated by standard samples with known before testing the samples of interest. To calculate the electrical conductivity (), the electrical conductance was measured first by tightly screwing the sample inside a PMMA tube and ramping a bias voltage across the sample to obtain the current–voltage (I–V) curve of which the slope was the
conductance. Using the known dimensions of the sample plus the measured electrical conductance, can be calculated from the relation = (conductance) (length)/(cross-sectional area). The PMMA tube connected with an electrical multimeter (Keithley 6430, USA) was calibrated by standard samples with known before testing the target samples. The commercial carbon black (Vulcan XC72 from Cabot, USA) was taken as the control sample. Its thermal and electrical conductivity were measured using the same methods and compared to the samples of concern. 3. Results and discussion 3.1. SEM analysis The SEM images of CSs, grown at 720 ◦ C on kaolin support with catalytic Ti, La, and Nb metal complexes, are shown in Fig. 2. The CSs observed in Fig. 2(a)–(c) resulted from catalyzed carbonization of C2 H2 by the metal complexes. Fig. 2(a) and (b) shows that
1592
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595
the obtained CSs using Ti and La as catalyst are similar in size. However, Fig. 2(c) shows the CSs with Nb as catalyst are much larger, indicating the Nb complex caused a higher rate of carbonization but a lower degree of graphitization (to be discussed below). The CSs observed in the micrographs often appear in singles, doubles, triples, or trains of merged spheres like soap bubbles with plane or curved inter-bubble surfaces. The mechanisms of forming such bubble-like CSs have been discussed in reports of pyrolytic CVD. Four processes are involved, namely, nucleation, surface diffusion, re-deposition, and sintering. Firstly, nucleation depends on substrate roughness and surface hydrophilicity. Striations on tungsten filament [37] and pores of anodic alumina [24] and kaolin [32–34] provide the nucleation sites. However, the surface hydrophilicity determines the wettability of carbonized material over the substrate, leading to either a 2D or 3D morphology. Since the kaolin support is relatively hydrophilic, it prevents the deposited carbon which is hydrophobic from wetting. The carbon took a 3D growth to form CSs. Otherwise, the final carbon product would have been carbon films, to which case a 2D growth model applies. To account for the sphericity of CSs, it is considered that the gaseous carbon precursor undergoes adsorption, decomposition, migration, and desorption unto successful nucleation. Unlike carbon nanotubes which often enclose the metal nanoparticles, in forming CSs the dispersed catalyst serves as carbon source while the nuclei as carbon sink. With re-deposition from the gas phase (a 3D transport) [35], and vivid surface diffusion (a 2D transport) between neighboring sources and sinks under a relatively high temperature, the pyrolyzed carbonaceous substance accretes at the nuclei to grow into single or coupled spheres. If the carbon nucleus were suspended in gas phase, it would grow into CSs isotropically by deposition. However, since the carbon nucleus was formed on the solid surface, it could grow only upwardly. When the contact area of the nucleus with the substrate is kept small due to low wettability, the continuous supply of carbon onto the nucleus will finally form a sphere via deposition and surface diffusion. After the CSs are blown off, new carbon nuclei appear on the catalyst boat. As long as the catalyst is not peeled off from kaolin with CSs, the synthesis can go on ceaselessly. However, the collected CSs should be removed constantly to avoid the gradual cementation near 700 ◦ C resulting from sintering by re-deposition and surface diffusion [32]. In contrast, the increase in size of the metal catalyst results from sintering and hinders the carbon nanotubes from formation [34]. A blank test has also been conducted showing no CSs collected in the absence of catalyst. However, this does not mean that no carbon species can be formed at the reaction temperatures used. The reaction rate is a key point. Since the catalyst functions to increase the decomposing rate of carbon precursor, when there is no catalyst, the decomposition of C2 H2 will be too slow to produce a high enough concentration for the pyrolyzed gaseous carbon radicals to re-deposit with a substantial amount on the kaolin support. Most carbon will be blown away such that almost no micrometer-sized CSs can be collected a little downstream of the catalyst boat inside the quartz tube. The catalytic nature of transition metals, surface area and hydrophilicity of kaolin, carbon source flowrate, and reaction temperature are considered the key parameters to give fast growth rate and form CSs. In this work, the yield of CSs was found best improved when they were synthesised at 810 ◦ C. The morphology is nonetheless similar to that at 720 ◦ C. Therefore, only the images and curves of CSs prepared at 720 ◦ C are shown here and in the following sections. If using lower reaction temperatures, it was found that CSs could not be formed below 700 ◦ C. As far as catalyst is concerned, our CSs
Fig. 3. X-ray patterns of (a) CS-La, (b) CS-Nb, and (c) CS-Ti; all synthesized at 720 ◦ C.
were synthesized for the first time using non-magnetic La, Ti, and Nb metal complexes on kaolin. 3.2. XRD analysis The X-ray diffraction patterns of the collected powder were given in Fig. 3. The patterns only show hexagonal structures of graphite without the crystalline ones of La, Nb, or Ti. The results show that the catalyst remained on the kaolin support and did not depart with the fallen CSs. However, there is still some amor-
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595
1593
Fig. 5. Plot of magnetic susceptibility (g ) against temperature for (a) CS-La, (b) CSNb, and (c) CS-Ti; all synthesized at 720 ◦ C. The insets in (b) and (c) give curves of inverse g vs. temperature. Fig. 4. TEM micrographs of (a) CS-La, (b) CS-Nb, and (c) CS-Ti; all synthesized at 720 ◦ C. The insets are enlarged portions.
phous metal possibly embedded in the CSs. Hence, the catalyst loss demands the refill of catalyst for a long-term operation. In contrast, XRD peaks referring to metal crystallites normally appear in the presence of synthesized carbon nanotubes due to entrapment [32,34]. 3.3. TEM analysis The TEM images of CSs are shown in Fig. 4. The diameters of CSs catalyzed by La, Nb, and Ti metal catalysts are measured as 600–875, 425–1100, and 650–1250 nm, respectively. Again the images show that the CSs have a solid core and that they do not encapsulate any
transition-metal nanoclusters, which is commonly observed in CVD [24,32–35]. The insets clearly reveal the solid core. Concerning the size, however, it is currently under study to obtain CSs smaller than 50 nm by adjusting the gas flowrate, reaction temperature, heating time, and else. 3.4. Analysis of magnetic susceptibility To examine the superconductivity of possibly encapsulated metallic nanoparticles, the magnetic susceptibility was measured under an applied magnetic field of 100 or 500 Oe. The temperature dependence of magnetic susceptibility is plotted in Fig. 5. The insets in Fig. 5(b) and (c) are curves of reciprocal magnetic susceptibility (g −1 ) vs. temperature. Since there was a weak diamagnetic noise in the CS-La sample, negative g above 10 K was measured as shown
1594
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595
4. Conclusions Sub-micrometer CSs were successfully synthesized by CCVD using novel catalyst on kaolin support. It is reported for the first time that the non-magnetic metal (La, Nb, and Ti) complexes catalyzed the carbonization of C2 H2 into CSs. The favorable growth temperature falls between 720 and 810 ◦ C. The X-ray diffraction patterns of CSs show graphitic structure with transition-metal crystallites excluded. The diameters of solid-core CSs using La, Nb, and Ti as catalyst were determined by TEM as 600–875, 425–1100, and 650–1250 nm, respectively. The acquired curves of magnetic susceptibility for CSs show background noises only, indicating no superconductivity and encapsulation of transition-metal nanoparticles. The Raman spectra at room temperature showed a higher degree of graphitization for the La-catalyzed CSs, based on the relative intensity of the feature peaks at 1325 cm−1 and 1580 cm−1 . Further work with such CCVD system to produce nano-CSs is underway. Finally, the CS-Ti and CS-La are two promising candidates for traditional and novel applications due to their comparable thermal and electrical conductivity to those of commercial carbon black. Fig. 6. Raman spectra of CSs measured at room temperature. From top down, the curves correspond respectively to CS-La, CS-Nb, and CS-Ti; all synthesized at 720 ◦ C.
in Fig. 5(a). The corresponding g −1 vs. temperature curve becomes divergent and no inset is provided. Although La (fcc), Nb (bcc), and Ti (hcp) are superconducting elements with transition temperatures at 6.0, 9.5, and 0.39 K, respectively, the superconducting behavior is not observed for all the samples. The background-only signals in Fig. 5 indicate few transition-metal nanoclusters encapsulated within the CSs out of the CCVD process. With little encapsulation of catalyst in produced CSs, it may take long in such CCVD to replenish additional catalyst, on condition that the catalyst is not deactivated somehow. 3.5. Analysis of Raman spectra The Raman spectra of CSs in Fig. 6 show two prominent peaks at 1325 cm−1 and 1580 cm−1 , corresponding to the common D and G bands, respectively, for graphitic materials [24,32,34,38,39]. To compare the relative intensity of G band to D band (IG /ID ), ratios of peak areas are calculated as 1.579, 1.221, and 1.300 for CS-La, CSNb, and CS-Ti samples, respectively. The results show that CS-La gives the highest degree of graphitization than the rest. 3.6. Measurement of thermal and electrical conductivity Table 1 lists the thermal and electrical conductivity of the CCVD CSs via different catalyst and the commercial carbon black (Vulcan XC72). It is found that CS-Ti is of the same order of magnitude in both and as XC72. However, CS-Nb seems to have lower and , probably due to a smaller degree of graphitization. It is worth noting that CS-La has the largest close to that of XC72 and the smallest , confirming the Raman results that CS-La was graphitized the most.
Table 1 Values of thermal and electrical conductivity of the CSs prepared by CCVD at 720 ◦ C in the presence of different catalyst. CCVD CSs
[W/m K]
[S/cm]
CS-Ti CS-Nb CS-La XC72
15.9 1.9 1.7 12.3
23.0 0.2 39.7 41.1
Acknowledgements The authors are indebted to Professor D.C. Ling of the Department of Physics at Tam Kang University for his help in experiments and insightful discussion. This work was financially supported by the National Science Council of Taiwan, ROC under grant NSC 95-2113-M-040-002 and Feng Chia University under grant FCU08C42279. References [1] F. Goutfer-Wurmser, H. Konno, Y. Kaburagi, K. Oshida, M. Inagaki, Synth. Met. 118 (2001) 33–38. [2] J. Przepiorski, B. Tryba, A.W. Morawski, Appl. Surf. Sci. 196 (2002) 296–300. [3] Z.X. Wang, L.P. Yu, W. Zhang, Z.Y. Zhu, G.W. He, Y. Chen, G. Hu, Phys. Lett. A 307 (2003) 249–252. [4] Y.Z. Jin, C. Gao, W.K. Hsu, Y.Q. Zhu, A. Huczko, M. Bystrzejewski, M. Roe, C.Y. Lee, S. Acquah, H. Kroto, D.R.M. Walton, Carbon 43 (2005) 1944–1953. [5] P.M. Ajayan, Chem. Rev. 99 (1999) 1787–1799. [6] H.J. Dai, Surf. Sci. 500 (2002) 218–241. [7] K.P. De Jong, J.W. Geus, Catal. Rev. -Sci. Eng. 42 (2000) 481–510. [8] W.K. Maser, A.M. Benito, M.T. Martinez, Carbon 40 (2002) 1685–1695. [9] C.C. Ting, W.S. Chao, Appl. Energy 87 (2010) 2500–2505. [10] B.X. Lei, W.J. Fang, Y.F. Hou, J.Y. Liao, D.B. Kuang, C.Y. Su, J. Photochem. Photobiol. A-Chem. 216 (2010) 8–14. [11] H.F. Xiong, M. Moyo, M.A.M. Motchelaho, L.L. Jewell, N.J. Coville, Appl. Catal. A-Gen. 388 (2010) 168–178. [12] D.S. Yuan, J.X. Chen, J.H. Zeng, S.X. Tan, Electrochem. Commun. 10 (2008) 1067–1070. [13] Y.Z. Jin, Y.J. Kim, C. Gao, Y.Q. Zhu, A. Huczko, M. Endo, H.W. Kroto, Carbon 44 (2006) 724–729. [14] F. El-Tantawy, K. Kamada, H. Ohnabe, Mater. Lett. 56 (2002) 112–126. [15] L. Gao, X.F. Zhou, Y.L. Ding, Chem. Phys. Lett. 434 (2007) 297–300. [16] Z.B. Lei, J.M. Li, Y.X. Ke, Y.G. Zhang, H.C. Zhang, F.Q. Li, J.Y. Xing, J. Mater. Chem. 11 (2001) 2930–2933. [17] Z.Y. Zhong, Y.D. Yin, B. Gates, Y.N. Xia, Adv. Mater. 12 (2000) 206–209. [18] F.L. Du, Z.Y. Guo, G.C. Li, Mater. Lett. 59 (2005) 2563–2565. [19] P.M. Ajayan, P. Redlich, M. Ruhle, J. Mater. Res. 12 (1997) 244–252. [20] T. Guo, P. Nikolaev, A. Thess, D.T. Colbert, R.E. Smalley, Chem. Phys. Lett. 243 (1995) 49–54. [21] D. Nishide, H. Kataura, S. Suzuki, S. Okubo, Y. Achiba, Chem. Phys. Lett. 392 (2004) 309–313. [22] T. Okazaki, T. Saito, K. Matsuura, S. Ohshima, M. Yumura, Y. Oyama, R. Saito, S. Iijima, Chem. Phys. Lett. 420 (2006) 286–290. [23] H.C. Wu, C.T. Hong, H.T. Chiu, Y.Y. Li, Diam. Relat. Mater. 18 (2009) 601–605. [24] H.S. Qian, F.M. Han, B. Zhang, Y.C. Guo, J. Yue, B.X. Peng, Carbon 42 (2004) 761–766. [25] M. Yudasaka, R. Kikuchi, Y. Ohki, E. Ota, S. Yoshimura, Appl. Phys. Lett. 70 (1997) 1817–1818. [26] K. Hernadi, A. Fonseca, J.B. Nagy, D. Bernaerts, A.A. Lucas, Carbon 34 (1996) 1249–1257. [27] D. Venegoni, P. Serp, R. Feurer, Y. Kihn, C. Vahlas, P. Kalck, Carbon 40 (2002) 1799–1807. [28] M. Perez-Cabero, I. Rodriguez-Ramos, A. Guerrero-Ruiz, J. Catal. 215 (2003) 305–316.
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595 [29] G.B. Zheng, K. Kouda, H. Sano, Y. Uchiyama, Y.F. Shi, H.J. Quan, Carbon 42 (2004) 635–640. [30] N. Toshima, Y. Shiraishi, T. Teranishi, M. Miyake, T. Tominaga, H. Watanabe, W. Brijoux, H. Bonnemann, G. Schmid, Appl. Organomet. Chem. 15 (2001) 178–196. [31] G. Schmid, Chem. Rev. 92 (1992) 1709–1727. [32] J.Y. Miao, D.W. Hwang, K.V. Narasimhulu, P.I. Lin, Y.T. Chen, S.H. Lin, L.P. Hwang, Carbon 42 (2004) 813–822. [33] J.Y. Miao, D.W. Hwang, C.C. Chang, S.H. Lin, K.V. Narasimhulu, L.P. Hwang, Diam. Relat. Mater. 12 (2003) 1368–1372.
1595
[34] Z.X. Xu, J.D. Lin, V.A.L. Roy, Y. Ou, D.W. Liao, Mater. Sci. Eng. B-Solid State Mater. Adv. Technol. 123 (2005) 102–106. [35] Z.L. Wang, Z.C. Kang, J. Phys. Chem. 100 (1996) 17725–17731. [36] ASTM E1225-09, Standard Test Method for Thermal Conductivity of Solids by Means of the Guarded-Comparative-Longitudinal Heat Flow Technique, 2009. [37] S.M.C. Vieira, C.A. Rego, P.R. Birkett, Diam. Relat. Mater. 17 (2008) 100–104. [38] Z. Mehraban, F. Farzaneh, V. Dadmehr, Mater. Lett. 63 (2009) 1653–1655. [39] A.A. Deshmukh, S.D. Mhlanga, N.J. Coville, Mater. Sci. Eng.: R: Rep. 70 (2010) 1–28.
Contents lists available at ScienceDirect
Synthetic Metals journal homepage: www.elsevier.com/locate/synmet
Synthesis and conductivity measurement of carbon spheres by catalytic CVD using non-magnetic metal complexes Chien-Ming Lei a , Wei-Li Yuan b , Hsin-Chen Huang a , Shao-Wei Ho c , Chi-Jung Su c,d,∗ a Department of Chemical & Materials Engineering and Graduate Institute of Nanomaterials, Chinese Culture University, 55, Huagang Rd., Shilin Dist., Taipei 11114, Taiwan, ROC b Department of Chemical Engineering, Feng Chia University, 100, Wenhua Rd., Xitun Dist., Taichung 40724, Taiwan, ROC c School of Applied Chemistry, Chung Shan Medical University No. 110, Sec. 1, Jianguo N. Rd., Taichung 40201, Taiwan, ROC d Department of Medical Research, Chung Shan Medical University Hospital, No. 110, Sec. 1, Jianguo N. Rd., Taichung 40201, Taiwan, ROC
a r t i c l e
i n f o
Article history: Received 2 February 2011 Received in revised form 6 May 2011 Accepted 18 May 2011 Available online 15 June 2011 Keywords: Carbon spheres Catalytic chemical vapor deposition Non-magnetic metal complexes Conductivity
a b s t r a c t Carbon spheres (CSs) have been an important subject of research in recent years. Catalytic chemical vapor deposition (CCVD) was carried out in this work to synthesize solid-core CSs at mild temperatures from 720 to 810 ◦ C. Non-magnetic metal complexes of La, Nb, and Ti, dispersed on porous kaolin support, were tried out as catalyst. X-ray diffraction patterns revealed the graphitic structures of CSs. TEM analysis showed no encapsulated transition-metal nanoparticles inside the CSs. It was found by Raman spectra that the La catalyst resulted in CSs with higher graphitization. To examine the potential applications of CSs to the fields such as catalysis, electrochemistry, and electronic device, values of the thermal and electrical conductivity of the prepared CSs using different catalysts were measured and found to be comparable to those of the commercial carbon black. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Novel carbon-based nanomaterials such as carbon spheres (CSs), carbon nanotubes, and carbon nanofibers have been attracting researchers from various disciplines to study their exciting chemical, physical, and mechanical properties [1–8]. CSs from nano- to micro-meters in size are becoming important due to their potential applications, like carbon black and nanotube, in many fields such as electro-conductive coatings, dye-sensitized solar cells, catalyst supports for impregnation of metal nanoclusters, electrochemical capacitors, battery anodes, and elements for electrostatic charge dissipation or electrical heating [9–15]. Recently, hollow spheres of metal oxides such as SiO2 , TiO2 , and SnO2 , prepared using nano/micro-sized carbon spheres as template have also been reported, which further extends the applicability of CSs [16–18]. Carbonaceous nanomaterials are frequently prepared by physical vapor deposition, electrical arc-discharge, and laser vaporization [19–22]. Chemical vapor deposition (CVD) involves carbonization of organic materials through a pyrolytic process at high temperatures above 1000 ◦ C [23,24]. Also popular, the catalytic CVD (CCVD) is a catalyzed carbonization process which often
uses transition-metal complexes of Fe, Ni, or Co to convert hydrocarbons such as methane, ethane, acetylene (C2 H2 ), and carbon monoxide into carbon spheres, tubes, or fibers at a temperature well below 1000 ◦ C [25–31]. The transition-metal nanoparticles are then derived from the metal complex precursors and sometimes found encapsulated by the carbonized material, judging from the sample’s ferrimagnetism or TEM images [32]. In most cases though, the metal nanoparticles are found trapped within carbon nanotubes but separate from the CSs [24,32–35]. In the present study, we report the fabrication of nano/microcarbon spheres by CCVD with non-magnetic complexes of La, Nb, and Ti as catalyst supported by kaolin. The microstructures of CSs were characterized by powder XRD, SEM, and TEM. The magnetic properties of CSs were tested by measuring the magnetic susceptibility at different temperatures. One advantage of using La, Nb, or Ti metal complex as catalyst is the potential for adding superconductivity to CSs if the non-magnetic metal is encapsulated within. In addition, the thermal and electrical conductivity of the CSs prepared via different catalysts were measured and compared with those of the commercial carbon black. 2. Experimental
∗ Corresponding author at: School of Applied Chemistry, Chung Shan Medical University No. 110, Sec. 1, Jianguo N. Rd., Taichung 40201, Taiwan, ROC. Fax: +886 4 24510890. E-mail address: [email protected] (C.-J. Su). 0379-6779/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.synthmet.2011.05.023
2.1. Catalysts for CSs To prepare the kaolin support, the kaolin powder (from Showa) was dissolved in distilled water in a weight ratio of 1:1 and stirred
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595
1591
Fig. 1. Experimental setup for synthesizing carbon spheres.
for 30 min. The kaolin solution was then poured into a mold to dry up in oven at 50 ◦ C overnight. The kaolin chunk was then carved into a boat with a trough. To prepare the catalysts, metal complexes such as lanthanum(III) nitrate hexahydrate (from Strem), ammonium hexafluoroniobate(V) (from Acros), and titanium(IV) (triethanolaminato)isopropoxide solution (80 wt.% in 2-propanol, from Aldrich) were purchased and used as received. Firstly 1 g lanthanum (III) nitrate hexahydrate was dissolved in 5 ml distilled water and stirred for 1 h at room temperature. Next, the well-mixed solution was dripped onto a kaolin boat and dried in oven at 50 ◦ C overnight. The drying process helps the metal complex catalyst adsorb strongly to the kaolin support. The same procedures were repeated to prepare the rest catalysts by dripping ammonium hexafluoroniobate (V) and titanium (IV) (triethanolaminato) isopropoxide on kaolin. 2.2. CCVD for CSs The CCVD schematic for synthesizing CSs is shown in Fig. 1. The kaolin boat with catalyst was placed upside down at the center of the quartz tube in an electrical furnace. The boat was firstly heated in N2 at 17 sccm from room temperature to 650 ◦ C in 1 h. Next a second stream of C2 H2 at 70 sccm was introduced into the reactor. After injection of the carbon source the temperature was further raised to the desired reaction temperature, 720 or 810 ◦ C, in 10 min. The reaction proceeded for 1 h at the reached temperature. In the end the reactor was cooled down to room temperature by continuous flushing of nitrogen. The CSs as fine black powder were collected by ceramic boats placed beneath the inverted catalyzing kaolin boat. Same schematic was used for all the three transition metal catalysts. The names of the CSs produced using different catalysts of lanthanum, niobium, and titanium are abbreviated as CS-La, CS-Nb, and CS-Ti, respectively.
Fig. 2. SEM micrographs of (a) CS-La, (b) CS-Nb, and (c) CS-Ti; all synthesized at 720 ◦ C. The insets are enlarged portions.
2.3. Analyses for CSs SEM images of the synthesized CSs were taken by JEOL JSM 5200, while the TEM ones by JEOL II 1200EX. X-ray diffraction patterns of the sample powder were acquired by Bruker MXP3. Raman spectra were scanned by Jobin-Yvon T64000. Plots of magnetic susceptibility against temperature were obtained with Quantum Design PPMS-7. The thermal conductivity () of CSs was measured in a way similar to standard methods such as ASTM E1225-09 [36]. A simple system was set up with the sample sandwiched between two copper rods and enclosed in a heat insulating container. By flowing heat through the sample, can be calculated using the given dimensions of the sample, temperature difference across the sample, and the heat flowrate. The system was calibrated by standard samples with known before testing the samples of interest. To calculate the electrical conductivity (), the electrical conductance was measured first by tightly screwing the sample inside a PMMA tube and ramping a bias voltage across the sample to obtain the current–voltage (I–V) curve of which the slope was the
conductance. Using the known dimensions of the sample plus the measured electrical conductance, can be calculated from the relation = (conductance) (length)/(cross-sectional area). The PMMA tube connected with an electrical multimeter (Keithley 6430, USA) was calibrated by standard samples with known before testing the target samples. The commercial carbon black (Vulcan XC72 from Cabot, USA) was taken as the control sample. Its thermal and electrical conductivity were measured using the same methods and compared to the samples of concern. 3. Results and discussion 3.1. SEM analysis The SEM images of CSs, grown at 720 ◦ C on kaolin support with catalytic Ti, La, and Nb metal complexes, are shown in Fig. 2. The CSs observed in Fig. 2(a)–(c) resulted from catalyzed carbonization of C2 H2 by the metal complexes. Fig. 2(a) and (b) shows that
1592
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595
the obtained CSs using Ti and La as catalyst are similar in size. However, Fig. 2(c) shows the CSs with Nb as catalyst are much larger, indicating the Nb complex caused a higher rate of carbonization but a lower degree of graphitization (to be discussed below). The CSs observed in the micrographs often appear in singles, doubles, triples, or trains of merged spheres like soap bubbles with plane or curved inter-bubble surfaces. The mechanisms of forming such bubble-like CSs have been discussed in reports of pyrolytic CVD. Four processes are involved, namely, nucleation, surface diffusion, re-deposition, and sintering. Firstly, nucleation depends on substrate roughness and surface hydrophilicity. Striations on tungsten filament [37] and pores of anodic alumina [24] and kaolin [32–34] provide the nucleation sites. However, the surface hydrophilicity determines the wettability of carbonized material over the substrate, leading to either a 2D or 3D morphology. Since the kaolin support is relatively hydrophilic, it prevents the deposited carbon which is hydrophobic from wetting. The carbon took a 3D growth to form CSs. Otherwise, the final carbon product would have been carbon films, to which case a 2D growth model applies. To account for the sphericity of CSs, it is considered that the gaseous carbon precursor undergoes adsorption, decomposition, migration, and desorption unto successful nucleation. Unlike carbon nanotubes which often enclose the metal nanoparticles, in forming CSs the dispersed catalyst serves as carbon source while the nuclei as carbon sink. With re-deposition from the gas phase (a 3D transport) [35], and vivid surface diffusion (a 2D transport) between neighboring sources and sinks under a relatively high temperature, the pyrolyzed carbonaceous substance accretes at the nuclei to grow into single or coupled spheres. If the carbon nucleus were suspended in gas phase, it would grow into CSs isotropically by deposition. However, since the carbon nucleus was formed on the solid surface, it could grow only upwardly. When the contact area of the nucleus with the substrate is kept small due to low wettability, the continuous supply of carbon onto the nucleus will finally form a sphere via deposition and surface diffusion. After the CSs are blown off, new carbon nuclei appear on the catalyst boat. As long as the catalyst is not peeled off from kaolin with CSs, the synthesis can go on ceaselessly. However, the collected CSs should be removed constantly to avoid the gradual cementation near 700 ◦ C resulting from sintering by re-deposition and surface diffusion [32]. In contrast, the increase in size of the metal catalyst results from sintering and hinders the carbon nanotubes from formation [34]. A blank test has also been conducted showing no CSs collected in the absence of catalyst. However, this does not mean that no carbon species can be formed at the reaction temperatures used. The reaction rate is a key point. Since the catalyst functions to increase the decomposing rate of carbon precursor, when there is no catalyst, the decomposition of C2 H2 will be too slow to produce a high enough concentration for the pyrolyzed gaseous carbon radicals to re-deposit with a substantial amount on the kaolin support. Most carbon will be blown away such that almost no micrometer-sized CSs can be collected a little downstream of the catalyst boat inside the quartz tube. The catalytic nature of transition metals, surface area and hydrophilicity of kaolin, carbon source flowrate, and reaction temperature are considered the key parameters to give fast growth rate and form CSs. In this work, the yield of CSs was found best improved when they were synthesised at 810 ◦ C. The morphology is nonetheless similar to that at 720 ◦ C. Therefore, only the images and curves of CSs prepared at 720 ◦ C are shown here and in the following sections. If using lower reaction temperatures, it was found that CSs could not be formed below 700 ◦ C. As far as catalyst is concerned, our CSs
Fig. 3. X-ray patterns of (a) CS-La, (b) CS-Nb, and (c) CS-Ti; all synthesized at 720 ◦ C.
were synthesized for the first time using non-magnetic La, Ti, and Nb metal complexes on kaolin. 3.2. XRD analysis The X-ray diffraction patterns of the collected powder were given in Fig. 3. The patterns only show hexagonal structures of graphite without the crystalline ones of La, Nb, or Ti. The results show that the catalyst remained on the kaolin support and did not depart with the fallen CSs. However, there is still some amor-
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595
1593
Fig. 5. Plot of magnetic susceptibility (g ) against temperature for (a) CS-La, (b) CSNb, and (c) CS-Ti; all synthesized at 720 ◦ C. The insets in (b) and (c) give curves of inverse g vs. temperature. Fig. 4. TEM micrographs of (a) CS-La, (b) CS-Nb, and (c) CS-Ti; all synthesized at 720 ◦ C. The insets are enlarged portions.
phous metal possibly embedded in the CSs. Hence, the catalyst loss demands the refill of catalyst for a long-term operation. In contrast, XRD peaks referring to metal crystallites normally appear in the presence of synthesized carbon nanotubes due to entrapment [32,34]. 3.3. TEM analysis The TEM images of CSs are shown in Fig. 4. The diameters of CSs catalyzed by La, Nb, and Ti metal catalysts are measured as 600–875, 425–1100, and 650–1250 nm, respectively. Again the images show that the CSs have a solid core and that they do not encapsulate any
transition-metal nanoclusters, which is commonly observed in CVD [24,32–35]. The insets clearly reveal the solid core. Concerning the size, however, it is currently under study to obtain CSs smaller than 50 nm by adjusting the gas flowrate, reaction temperature, heating time, and else. 3.4. Analysis of magnetic susceptibility To examine the superconductivity of possibly encapsulated metallic nanoparticles, the magnetic susceptibility was measured under an applied magnetic field of 100 or 500 Oe. The temperature dependence of magnetic susceptibility is plotted in Fig. 5. The insets in Fig. 5(b) and (c) are curves of reciprocal magnetic susceptibility (g −1 ) vs. temperature. Since there was a weak diamagnetic noise in the CS-La sample, negative g above 10 K was measured as shown
1594
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595
4. Conclusions Sub-micrometer CSs were successfully synthesized by CCVD using novel catalyst on kaolin support. It is reported for the first time that the non-magnetic metal (La, Nb, and Ti) complexes catalyzed the carbonization of C2 H2 into CSs. The favorable growth temperature falls between 720 and 810 ◦ C. The X-ray diffraction patterns of CSs show graphitic structure with transition-metal crystallites excluded. The diameters of solid-core CSs using La, Nb, and Ti as catalyst were determined by TEM as 600–875, 425–1100, and 650–1250 nm, respectively. The acquired curves of magnetic susceptibility for CSs show background noises only, indicating no superconductivity and encapsulation of transition-metal nanoparticles. The Raman spectra at room temperature showed a higher degree of graphitization for the La-catalyzed CSs, based on the relative intensity of the feature peaks at 1325 cm−1 and 1580 cm−1 . Further work with such CCVD system to produce nano-CSs is underway. Finally, the CS-Ti and CS-La are two promising candidates for traditional and novel applications due to their comparable thermal and electrical conductivity to those of commercial carbon black. Fig. 6. Raman spectra of CSs measured at room temperature. From top down, the curves correspond respectively to CS-La, CS-Nb, and CS-Ti; all synthesized at 720 ◦ C.
in Fig. 5(a). The corresponding g −1 vs. temperature curve becomes divergent and no inset is provided. Although La (fcc), Nb (bcc), and Ti (hcp) are superconducting elements with transition temperatures at 6.0, 9.5, and 0.39 K, respectively, the superconducting behavior is not observed for all the samples. The background-only signals in Fig. 5 indicate few transition-metal nanoclusters encapsulated within the CSs out of the CCVD process. With little encapsulation of catalyst in produced CSs, it may take long in such CCVD to replenish additional catalyst, on condition that the catalyst is not deactivated somehow. 3.5. Analysis of Raman spectra The Raman spectra of CSs in Fig. 6 show two prominent peaks at 1325 cm−1 and 1580 cm−1 , corresponding to the common D and G bands, respectively, for graphitic materials [24,32,34,38,39]. To compare the relative intensity of G band to D band (IG /ID ), ratios of peak areas are calculated as 1.579, 1.221, and 1.300 for CS-La, CSNb, and CS-Ti samples, respectively. The results show that CS-La gives the highest degree of graphitization than the rest. 3.6. Measurement of thermal and electrical conductivity Table 1 lists the thermal and electrical conductivity of the CCVD CSs via different catalyst and the commercial carbon black (Vulcan XC72). It is found that CS-Ti is of the same order of magnitude in both and as XC72. However, CS-Nb seems to have lower and , probably due to a smaller degree of graphitization. It is worth noting that CS-La has the largest close to that of XC72 and the smallest , confirming the Raman results that CS-La was graphitized the most.
Table 1 Values of thermal and electrical conductivity of the CSs prepared by CCVD at 720 ◦ C in the presence of different catalyst. CCVD CSs
[W/m K]
[S/cm]
CS-Ti CS-Nb CS-La XC72
15.9 1.9 1.7 12.3
23.0 0.2 39.7 41.1
Acknowledgements The authors are indebted to Professor D.C. Ling of the Department of Physics at Tam Kang University for his help in experiments and insightful discussion. This work was financially supported by the National Science Council of Taiwan, ROC under grant NSC 95-2113-M-040-002 and Feng Chia University under grant FCU08C42279. References [1] F. Goutfer-Wurmser, H. Konno, Y. Kaburagi, K. Oshida, M. Inagaki, Synth. Met. 118 (2001) 33–38. [2] J. Przepiorski, B. Tryba, A.W. Morawski, Appl. Surf. Sci. 196 (2002) 296–300. [3] Z.X. Wang, L.P. Yu, W. Zhang, Z.Y. Zhu, G.W. He, Y. Chen, G. Hu, Phys. Lett. A 307 (2003) 249–252. [4] Y.Z. Jin, C. Gao, W.K. Hsu, Y.Q. Zhu, A. Huczko, M. Bystrzejewski, M. Roe, C.Y. Lee, S. Acquah, H. Kroto, D.R.M. Walton, Carbon 43 (2005) 1944–1953. [5] P.M. Ajayan, Chem. Rev. 99 (1999) 1787–1799. [6] H.J. Dai, Surf. Sci. 500 (2002) 218–241. [7] K.P. De Jong, J.W. Geus, Catal. Rev. -Sci. Eng. 42 (2000) 481–510. [8] W.K. Maser, A.M. Benito, M.T. Martinez, Carbon 40 (2002) 1685–1695. [9] C.C. Ting, W.S. Chao, Appl. Energy 87 (2010) 2500–2505. [10] B.X. Lei, W.J. Fang, Y.F. Hou, J.Y. Liao, D.B. Kuang, C.Y. Su, J. Photochem. Photobiol. A-Chem. 216 (2010) 8–14. [11] H.F. Xiong, M. Moyo, M.A.M. Motchelaho, L.L. Jewell, N.J. Coville, Appl. Catal. A-Gen. 388 (2010) 168–178. [12] D.S. Yuan, J.X. Chen, J.H. Zeng, S.X. Tan, Electrochem. Commun. 10 (2008) 1067–1070. [13] Y.Z. Jin, Y.J. Kim, C. Gao, Y.Q. Zhu, A. Huczko, M. Endo, H.W. Kroto, Carbon 44 (2006) 724–729. [14] F. El-Tantawy, K. Kamada, H. Ohnabe, Mater. Lett. 56 (2002) 112–126. [15] L. Gao, X.F. Zhou, Y.L. Ding, Chem. Phys. Lett. 434 (2007) 297–300. [16] Z.B. Lei, J.M. Li, Y.X. Ke, Y.G. Zhang, H.C. Zhang, F.Q. Li, J.Y. Xing, J. Mater. Chem. 11 (2001) 2930–2933. [17] Z.Y. Zhong, Y.D. Yin, B. Gates, Y.N. Xia, Adv. Mater. 12 (2000) 206–209. [18] F.L. Du, Z.Y. Guo, G.C. Li, Mater. Lett. 59 (2005) 2563–2565. [19] P.M. Ajayan, P. Redlich, M. Ruhle, J. Mater. Res. 12 (1997) 244–252. [20] T. Guo, P. Nikolaev, A. Thess, D.T. Colbert, R.E. Smalley, Chem. Phys. Lett. 243 (1995) 49–54. [21] D. Nishide, H. Kataura, S. Suzuki, S. Okubo, Y. Achiba, Chem. Phys. Lett. 392 (2004) 309–313. [22] T. Okazaki, T. Saito, K. Matsuura, S. Ohshima, M. Yumura, Y. Oyama, R. Saito, S. Iijima, Chem. Phys. Lett. 420 (2006) 286–290. [23] H.C. Wu, C.T. Hong, H.T. Chiu, Y.Y. Li, Diam. Relat. Mater. 18 (2009) 601–605. [24] H.S. Qian, F.M. Han, B. Zhang, Y.C. Guo, J. Yue, B.X. Peng, Carbon 42 (2004) 761–766. [25] M. Yudasaka, R. Kikuchi, Y. Ohki, E. Ota, S. Yoshimura, Appl. Phys. Lett. 70 (1997) 1817–1818. [26] K. Hernadi, A. Fonseca, J.B. Nagy, D. Bernaerts, A.A. Lucas, Carbon 34 (1996) 1249–1257. [27] D. Venegoni, P. Serp, R. Feurer, Y. Kihn, C. Vahlas, P. Kalck, Carbon 40 (2002) 1799–1807. [28] M. Perez-Cabero, I. Rodriguez-Ramos, A. Guerrero-Ruiz, J. Catal. 215 (2003) 305–316.
C.-M. Lei et al. / Synthetic Metals 161 (2011) 1590–1595 [29] G.B. Zheng, K. Kouda, H. Sano, Y. Uchiyama, Y.F. Shi, H.J. Quan, Carbon 42 (2004) 635–640. [30] N. Toshima, Y. Shiraishi, T. Teranishi, M. Miyake, T. Tominaga, H. Watanabe, W. Brijoux, H. Bonnemann, G. Schmid, Appl. Organomet. Chem. 15 (2001) 178–196. [31] G. Schmid, Chem. Rev. 92 (1992) 1709–1727. [32] J.Y. Miao, D.W. Hwang, K.V. Narasimhulu, P.I. Lin, Y.T. Chen, S.H. Lin, L.P. Hwang, Carbon 42 (2004) 813–822. [33] J.Y. Miao, D.W. Hwang, C.C. Chang, S.H. Lin, K.V. Narasimhulu, L.P. Hwang, Diam. Relat. Mater. 12 (2003) 1368–1372.
1595
[34] Z.X. Xu, J.D. Lin, V.A.L. Roy, Y. Ou, D.W. Liao, Mater. Sci. Eng. B-Solid State Mater. Adv. Technol. 123 (2005) 102–106. [35] Z.L. Wang, Z.C. Kang, J. Phys. Chem. 100 (1996) 17725–17731. [36] ASTM E1225-09, Standard Test Method for Thermal Conductivity of Solids by Means of the Guarded-Comparative-Longitudinal Heat Flow Technique, 2009. [37] S.M.C. Vieira, C.A. Rego, P.R. Birkett, Diam. Relat. Mater. 17 (2008) 100–104. [38] Z. Mehraban, F. Farzaneh, V. Dadmehr, Mater. Lett. 63 (2009) 1653–1655. [39] A.A. Deshmukh, S.D. Mhlanga, N.J. Coville, Mater. Sci. Eng.: R: Rep. 70 (2010) 1–28.