Fabrication of carbon nanoflowers by plasma-enhanced chemical vapor deposition

Fabrication of carbon nanoflowers by plasma-enhanced chemical vapor deposition

Applied Surface Science 255 (2009) 7846–7850 Contents lists available at ScienceDirect Applied Surface Science journal homepage: www.elsevier.com/lo...

629KB Sizes 0 Downloads 28 Views

Applied Surface Science 255 (2009) 7846–7850

Contents lists available at ScienceDirect

Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc

Fabrication of carbon nanoflowers by plasma-enhanced chemical vapor deposition Xiying Ma *, Baohe Yuan Institute of Optoelectronic Materials, Shaoxing University, The Around City West Road 508, Shaoxing 312000, Zhejiang, China



Article history: Received 14 December 2008 Received in revised form 15 March 2009 Accepted 15 March 2009 Available online 27 March 2009

Two and three-dimensional (2D and 3D) carbon nanoflowers have been prepared on silicon (1 1 1) substrates by plasma-enhanced chemical vapor deposition, using CH4, H2 and Ar as reactive gases in the presence of Fe catalyst. The flower patterns are controlled by the flux ratio of the carrier gas, the reaction pressure and the growth temperature. Through observation by scanning electron microscopy, we find that the 2D carbon nanoflowers are formed by various nanoleaves while the 3D flowers are composed of hundreds of nanofibers. The former is related closely to the flux ratio of gas and the reaction pressure, while the latter depended mainly on the growth temperature. The nucleation and growth process of the nanoflowers seem to be a vapor/liquid/solid mechanism. ß 2009 Elsevier B.V. All rights reserved.

PACS: 61.46+w 61.48+c Keywords: Carbon nanoflower PECVD SEM

1. Introduction Low-dimensional nanostructures, such as nanotubes [1], nanowires [2] and nanoribbons [3] have attracted much attention because of their novel physical properties and potential applications in constructing nanoscale electric and optoelectronic devices. Various new carbon nanostructures have been synthesized since the fullerene structure was first discovered in the 1980s [4]. For example, many fullerene derivatives and nanodiamond were synthesized and studied for potential applications [5,6]. At the same time, carbon nanofiber, nanowires and nanobelts were prepared using vapor-growth carbon fiber (VGCF) methods [7,8]. Single and multi-walled carbon nanotubes were fabricated using electric arc discharge [9], laser ablation [10] and plasma-enhanced chemical vapor deposition (PECVD) [11]. In particular, nanotubes and nanowires have been studied intensively for their potential applications in optical and electronic fields, including super-hard materials, field emitters [12], nanoprobes [13] and high-strength composites [14]. These new carbon nanostructures have enriched the research fields of carbon, and have supplied many indispensable new materials for modern industry.

* Corresponding author. Fax: +86 575 8342415. E-mail address: [email protected] (X. Ma). 0169-4332/$ – see front matter ß 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2009.03.061

Recently, with the progress of nanotechnology, more novel nanostructures, such as nanoflowers (SiC and InN) [15,16] and flower-like structures (ZnO and MgO) [17,18] have been synthesized. Here, we describe for the first time the growth of carbon nanoflowers, nanoleaves and nanotrees using the PECVD technique. These elegant nanoflower structures are obtained by controlling the flux of CH4, H2 and Ar, the reaction chamber pressure and the growth temperature. On the basis of scanning electro-microscope (SEM) observation and Raman spectroscopy, the growth mechanism of the structures is discussed. 2. Experiments The 2D and 3D carbon nanoflowers were synthesized on silicon (1 1 1) substrates by PECVD, using CH4 (99.999%), H2 (99.999%) and Ar (99.999%) as reactive gases. The system is made up of a vacuum reactive chamber, a plasma controller, a power and temperature controller, and a gas flowmeter, the sketch of the experimental setup is shown in Fig. 1. In which the plasma is a standard DC power that plays a key action of decomposition CH4 and recombination a new material rapidly during the deposition under low temperature [19,20]. The output power can be easily tuned by the applied electrical field between the anode and cathode. The substrates were cleaned ultrasonically with a sequence of acetone, ethanol and de-ionized water. Then they were coated uniformly with a Fe catalyst by dipping into a weak solution of FeCl2 and dried by a N2

X. Ma, B. Yuan / Applied Surface Science 255 (2009) 7846–7850


Fig. 1. The sketch of the experimental set-up.

flow, then placed on a graphite holder in a vacuum chamber. Before depositing the carbon flower films, the substrates were treated with H2 plasma for 3 min. Subsequently, the reactive gases were introduced separately into the reaction chamber at a base pressure of 10 5 Pa. Samples with different morphologies were deposited at the power of 500 W at the electrical field of 500 V for 2–4 h by changing the flux of CH4, H2 and Ar in a ratio of 1:3:3 to 1:3:6 (by vol.), the reaction pressure from 50 to 300 Pa, and the growth temperature in a range of 500–650 8C. Finally, the surface patterns of the samples were characterized by SEM. 3. Results and discussions 3.1. Fabrication of 2D carbon nanoflowers and nanoleaves When the gas flux ratio of CH4, H2 and Ar was 1:3:3 (by vol.), the reaction pressure was kept at 50 Pa, and the growth temperature was 550 8C, an SEM image of the deposited film is shown in Fig. 2. A hexangular carbon nanostructure, which looks like a flower, hence the term nanoflower, appears. The nanoflower is composed mainly of six ‘‘petals’’ emanating from the center and growing into a planar structure 7 mm long and 6.2 mm wide. The petals branch and some have a wavelike edge. In particular, the petals at the top left corner and at the right divide into two, which make the nanoflower appear to have an axis symmetry structure with respect to the diagonal line shown in the figure. Moreover, the length of the petals is diverse; the shortest is about 2 mm and the longest is up to 3 mm, reflecting the fact that the growth speed of the petals varies in different directions. Furthermore, the surface of the petals is very smooth and almost parallel with the substrate surface, indicating that growth is mainly in the plane of the surface. The morphology of our carbon nanoflower is unique: the pattern itself is like a flower, in contrast to the bushlike nanoflowers that have been reported [20,21], and the shape of the leaves edge appears as very smooth wavelike lace. Raman spectroscopy is the best way to characterise carbon products [22], since carbon materials, such as graphite, diamond, fullerenes and carbon nanotubules, have typical vibration modes that can be resolved by Raman spectroscopy. Therefore, the phase of this carbon nanoflower was analyzed by Raman spectroscopy at room temperature with a resolution of about 1 cm 1 and an excitation wavelength of 514.5 nm. In Fig. 2b, two peaks are obvious in the spectrum, the small one at 1208 cm 1 is from amorphous carbon [23] while the strong one at 1581 cm 1 corresponds to the typical G vibration peak of graphite, indicating

Fig. 2. (a) SEM image of a carbon nanoflower, formed a hexagonal planar structure with six petals and wavelike edges. (b) Raman spectroscopy shows a strong vibration peak at 1581 cm 1, indicating that the carbon nanoflower is a hexagonal graphite structure. (c) The schematic illustrates the nucleation and growth of the nanoflower.

that the carbon nanoflower has a hexagonal graphite structure. The broad peak at 1581 cm 1 results from the large size distribution. Fullerenes with spherical structure and closed nanotubules are both composed of hexacyclic carbon. Even though the growth mechanism of the planar carbon nanoflower is not clear, we can infer that it is developed by an unfolded layer of graphite, which has not been seen previously. For carbon materials grown by PECVD, Hatta et al. thought that there are many tiny carbon slices or clusters in the plasma vapor [24] that initially nucleate on Fe or Ni catalyst particles and then grow into different polygonal structures. In our experiment, considering that the growth of the carbon nanoflower is mainly parallel with the substrate surface, we propose the three-step growth model illustrated by Fig. 2c. In the first step, CH4 molecules are decomposed, and carbon slices and clusters are formed in the plasma under radiofrequency power and high temperature, in which carbon slices are dominant in the vapor. In the second step, Fe acts as a nucleation center to absorb carbon slices and then enhance the nucleation process of them. In the third step, when carbon slices on the Fe surface are saturated in


X. Ma, B. Yuan / Applied Surface Science 255 (2009) 7846–7850

stronger interaction of the carbon slices with an enhanced number and velocity with increasing reaction pressure and temperature. As a result, the leaves keep an orientated growth tendency, but prefer one or two directions. With the pressure and temperature increasing further, the pattern obtained is shown in Fig. 3b, where there are many nanoleaves 1 mm long and 0.5 mm wide. Like tree leaves dropped on the ground, the leaves are distributed randomly and some of them are partly overlapped. Also, they have a long thin mid-rib, making them resemble the symmetric structure of foliage. In this case we supposed that Fe particles were first assembled into lines that are distributed randomly on the substrate at high temperature, then the leaves are developed as carbon slices are nucleated on both sides of the line, and then arrayed with a tendency to short-range order due to the strong interaction. With pressure increasing to 300 Pa and temperature to 580 8C, the growth becomes completely direction-free, and forms the random flower pattern seen in Fig. 3c, which is formed by many small irregular interconnected leaves. The leaf structure is arbitrary, but the top surface is very smooth without any impurity. On the other hand, it can be considered to be a porous carbon film on which many pores with arbitrary shape are distributed uniformly on the substrate forming a carbon molecular sieve, and the porosity is estimated to be 40%. To our knowledge, porous carbon is usually synthesized by using porous templates. For example, Su et al. obtained microporous carbon by using an ammonium-form zeolite template [25]. In the present experiment a uniform nanoporous pattern was formed by self-assembly. Above the carbon nanoflower and nanoleaves, all growth is planar, but the shapes are affected significantly by the gas flux, pressure and temperature, of which the gas flux and pressure have the most important influence on the growth of nanoflowers, probably due to the larger change of these parameters than that of temperature. So far, zero, one and three-dimensional carbon structures (fullerene, nanotubule and diamond) have all been fabricated. However, large-scale 2D carbon structures are reported infrequently. These new planar structures will enrich the carbon family, and find great application in planar display and field emitter devices. 3.2. Fabrication of 3D carbon nanoflower

Fig. 3. SEM images of the carbon products with different morphologies. (a) Cactus ‘‘leaves’’, (b) small random leaves, and (c) nanoporous film, deposited by varying the flux ratio of CH4, H2 and Ar from 1:3:4 to 1:3:6, the pressure from 100 to 200 Pa, and the growth temperature from 550 to 580 8C.

one direction, they start to extend along the directions under lower vapor pressure. During normal growth, carbon slices always array along the substrate surface. Eventually, the carbon nanoflower with a hexagonal snowflake-like structure is developed, which agrees with the hexagonal structure of graphite. As the growth temperature was increased from 550 to 580 8C, the gas flux ratio of CH4, H2 and Ar was varied from 1:3:4 to 1:3:6 (by vol.), the reactive pressure was adjusted from 100 to 300 Pa, the SEM images of the carbon products are shown in Fig. 3a–c. In Fig. 3a, the carbon nanoleaves are arrayed in parallel along the northwest direction and take on an axis symmetry with respect to the red line. Both the shapes and their growth behavior are like a cactus that reproduce from a leaf end. Each leaf is about 2 mm long and 250 nm wide, and the total length of the nanoleaves is up to 6– 7 mm. We note that there are many Fe particles on the leaves, which makes a carbon–Fe composite material. Compared with Fig. 2, the pattern of the leaves changes greatly, resulting from the

Fig. 4a–c shows SEM images of the films deposited at 650 8C on Fe-coated Si (1 1 1) substrates over 2–4 h. Fig. 4a shows a 3D carbon nanoflower composed of hundreds of outward facing bushy petals grown for 2 h. The closer to the center, the greater is the density of the petals. As a result, it seems to have a ‘‘heart’’ in the center. The corolla features a center symmetry with a diameter of about 5.5 mm. With growth time increased to 2.5 h, the petals spread out in all directions (Fig. 4b) and the corolla extends to 6.5 mm. Finally, the nanoflower is completely open when the time is increased to 4 h (Fig. 4c), consisting of a large number of outward facing carbon nanofibers with a mean diameter about 30 nm and a length of 0.5–0.8 mm. So far, several mechanisms have been proposed to account for the growth of nanowires and nanofibers, which includes vapor/ liquid/solid (VLS) [26], and vapor/solid (VS) [27]. Our experiment is performed in the presence of Fe catalyst, even though there is catalyst-free on top of the carbon nanofibers; the growth appears to be a VLS process. Initially, carbon clusters are absorbed onto the fused Fe catalyst surface and nucleation at a temperature of 650 8C. Short carbon nanofibers are nucleated when carbon clusters on the Fe surface became supersaturated, and then the carbon nanofibers grow away from the substrate. Finally, a 3D carbon nanoflower is developed by spreading the aggregated nanofibers over the deposition time.

X. Ma, B. Yuan / Applied Surface Science 255 (2009) 7846–7850


Fig. 5. SEM images of a carbon nanotree formed by a thick carbon layer covering the nanofibers. (a and b) Images taken in different areas of the film, while (c) is a high magnification of (b).

Fig. 4. SEM images of 3D carbon nanoflowers. (a) A nanoflower composed of hundreds of outward facing petals for deposition 2 h, (b) the nanoflower starts to ‘‘bloom’’ for deposition 2.5 h and (c) a completely open nanoflower, consisting of large number of outward facing nanofibers with a mean diameter of 30 nm for growth 4 h.

3.3. Fabrication of carbon nanotree Fig. 5a–c shows SEM images of the film formed by a two-step growth strategy, which was initially nucleated at 650 8C for 40 min, followed by normal growth at 500 8C for 80 min. The images in Fig. 5a and b were taken in different areas of the film,

while Fig. 5c is a high magnification of Fig. 5b. As a whole, they look like a tree covered by a thick layer of snow. The petals here are about 200 nm, on which there are many small irregular pores (Fig. 5c). Similar to the growth of nanofibers, in the first stage the nanotree is formed by carbon clusters nucleated on the Fe surface and grow into short carbon nanofibers at high temperature. As the temperature falls in the second stage, amorphous carbon clusters are dominant in the plasma vapor. The flowers are developed gradually when amorphous carbon accumulates on the fiber surface and combined by a weak van der Waals force. 4. Conclusions We have successfully fabricated 2D carbon nanoflowers and nanoleaves, and 3D nanofiber flowers and nanotrees on Si (1 1 1) substrates by PECVD. We found that the reaction pressure and temperature were important influences on their morphology. On the basis of the analysis of the growth processes of carbon nanoflowers and SEM observation, we conclude that the growth of


X. Ma, B. Yuan / Applied Surface Science 255 (2009) 7846–7850

carbon nanoflowers via a VLS mechanism is possible. This synthetic strategy provides a simple method for synthesis of various carbon nanoflower structures, and offers an opportunity for the extensive investigation of other nanostructures. Acknowledgements This work was supported, in part, by the national natural science foundation of China (No. 60776004) and the undergraduate bud project of Zhejiang Province (No. 2008R40G2180006). References [1] [2] [3] [4] [5] [6]

S. Lijiama, T. Ichihashi, Nature 363 (1993) 603. P. Gao, Z.L. Wang, J. Phys. Chem. B 106 (2002) 1265. Z.R. Dai, Z.W. Pan, Z.L. Wang, J. Phys. Chem. B 106 (2002) 902. H.W. Kroto, J.R. Heath, S.C. O’Brien, R.F. Crul, R.E. Smalley, Nature 318 (1985) 162. D. Ugarte, Nature 359 (1992) 707. J.G. Hou, J.L. Yang, H.Q. Wang, Q.X. Li, C.G. Zeng, L.F. Yuan, B. Wang, D.M. Chen, Q.S. Zhu, Nature 409 (6818) (2001) 304. [7] S. Motojima, I. Hasegawa, S. Kagaya, J. Mater. Sci. 30 (1995) 5049.

[8] B. Vigolo, A. Penicaud, C. Coulon, C. Sauder, R. Pailler, C. Journet, P. Bernier, P. Poulin, Science 290 (2000) 1331. [9] S. Iijima, T. Ichihashi, Y. Pentagons, Nature 356 (1992) 776. [10] E.G. Gamaly, W.T. Ebbesen, Phys. Rev. B 52 (1995) 2083. [11] A. Thess, R. Lee, P. Nikolaev, H. Dai, P. Petit, J. Robert, C. Xu, Y.H. Lee, S.G. Kim, G.E. Scuseria, D. Tomanek, J.E. Fischer, R.E. Smalley, Science 273 (1996) 483. [12] L.C. Qin, D. Zhou, A.R. Krauss, D.M. Gruen, Appl. Phys. Lett. 72 (1998) 3437. [13] S.P. Jarvis, T. Uchihashi, T. Ishida, H. Tokumoto, J. Phys. Chem. B 104 (2000) 6091. [14] P.M. Ajayan, L.S. Schadler, C. Giannaris, A. Rubio, Adv. Mater. 12 (2000) 750. [15] F.Q. He, Y.P. Zhao, Appl. Phys. Lett. 88 (2006) 193113. [16] X.S. Fang, C.H. Ye, L.D. Zhang, J.X. Zhang, J.W. Zhao, P. Yan, Small 1 (2005) 422. [17] T.T. Kang, X.L. Liu, R.Q. Zhang, W.G. Hu, G. Cong, F.A. Zhao, Appl. Phys. Lett. 89 (2006) 071113. [18] G.W. Ho, A.S.W. Wong, D.J. Kang, M.E. Well, Nanotechnology 15 (2004) 996. [19] K.O. Denysenko, Appl. Phys. Lett. 90 (2007) 251501. [20] I. Denysenko, K. Ostrikov, J. Phys. D 42 (2009) 015208. [21] G. Shen, Y. Bando, C.J. Lee, J. Phys. Chem. B 109 (2005) 10779. [22] B.S. Elman, M. Shayegan, M.S. Dresselhaus, H. Mazurek, G. Dresselhaus, Phys. Rev. B 25 (1982) 4142. [23] A.M. Rao, E. Richter, S. Bandow, B. Chase, P.C. Eklund, K.A. Williams, S. Fang, G.U. Sumanasekera, A.L. Loper, E.C. Dickey, Science 275 (1997) 187. [24] N. Hatta, K. Murata, Chem. Phys. Lett. 217 (1994) 398. [25] F. Su, X.S. Zhao, L. Lu, Z. Zhou, Carbon 42 (2004) 2821. [26] M. Endo, T. Toyama, Oberlin, Jpn. J. Appl. Phys. 16 (1977) 1519. [27] C.H. Ye, G.W. Meng, Y.H. Wang, Z. Jiang, L. Zhang, J. Phys. Chem. B 106 (2002) 10338.