Growth and field emission properties of tubular carbon cones

Growth and field emission properties of tubular carbon cones

ARTICLE IN PRESS Ultramicroscopy 107 (2007) 861–864 www.elsevier.com/locate/ultramic Growth and field emission properties of tubular carbon cones J.J...

458KB Sizes 2 Downloads 22 Views

ARTICLE IN PRESS

Ultramicroscopy 107 (2007) 861–864 www.elsevier.com/locate/ultramic

Growth and field emission properties of tubular carbon cones J.J. Li, Q. Wang, C.Z. Gu Beijing National Laboratory for Condensed Matter Physics, Microfabrication Laboratory, Institute of Physics, Chinese Academy of Sciences, Beijing 100080, PR China

Abstract New forms of tubular carbon cone (TCC) were grown on gold wires by hot-filament chemical vapor deposition (HFCVD). They have a long-cone-shaped appearance with a herringbone hollow interior, surrounded by helical sheets of graphite that are coiled around it. It is considered that TCC formation results because the size of the catalyst particle located in the top of the TCC decreases continuously during growth, due to etching effects in the CVD plasma, reflecting competition between the growth and etching processes in the plasma. In addition, field emission measurements show that TCCs have a very low-threshold field of 0.27 V/mm, and that a stable macroscopic emitting current density of 1 mA/cm2 can be obtained at only 0.5 V/mm. TCCs have good field emission properties, compared to other forms of carbon field emitter, and may be good candidates for use in field emission display devices. r 2007 Elsevier B.V. All rights reserved. PACS: 61.46; 79.70 Keywords: CVD; Microstructure; Field emission; Carbon cone

1. Introduction The formation of cone-shaped carbon nanostructures is an attractive research field, due to the potential applications for practical nanoscale devices in nanoelectronics, scanning microscopy, atomic force microscopy and field emission devices. So far, some novel cone-shaped carbon nanostructures have been reported, such as graphitic nanocones [1,2], nanohorns [3], nanopipettes [4] and tubular graphite cones [5]. Characterization of their various properties has mostly concentrated on their appearance and microstructure as measured by scanning electron microscopy (SEM) and transmission electron microscopy (TEM); other properties such as electron field emission are seldom studied. In addition, despite a similar cone-shaped appearance, these carbon nanostructures have different aspect ratios, which are likely to be related to the growth mechanism. Although the growth mechanism of cone-shaped carbon nanostructures has been discussed and some viewpoints Corresponding author. Tel./fax: +86 10 82648198.

E-mail address: [email protected] (J.J. Li). 0304-3991/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.ultramic.2007.02.021

proposed [4,6], understanding the exact growth mechanism is still a challenge. Among the reported cone-shaped carbon nanostructures, nanopipettes and tubular graphite cones have a special structure with a hollow interior, differing in this respect from other carbon nanocones. Nanopipettes are hundreds of nanometers long, with a well-defined uniform 1–3 nm hollow core and a shell containing helical graphitic sheets. Another long conical structure, called a tubular graphite cone, is composed of cylindrical graphite sheets; continuous shortening of graphite layers from the interior to the exterior makes them cone shaped. Although the growth mechanism of the long conical structure is still unclear, these conical graphitic structures have potential uses as scanning probe tips for STM and AFM measurements, but with greater rigidity and easier mounting than the carbon nanotubes currently used. On the other hand, this long conical graphitic structure also combines the stability of a cone with the cylinder’s advantage for field emission at low applied fields, and it may be suitable as a cold cathode in field emission devices. In our work, new tubular carbon cones (TCCs) similar to the reported nanopipette and tubular graphite cones are

ARTICLE IN PRESS 862

J.J. Li et al. / Ultramicroscopy 107 (2007) 861–864

grown successfully, the growth mechanism is studied (based on their appearances and microstructure characteristics), and the field emission properties are measured.

distance between the anodic probe and the sample used as the cathode is kept at 4 mm. When the vacuum is pumped down to 2  10–6 Pa, the voltage is applied gradually while the emission current is recorded by an electrometer.

2. Experimental details 3. Results and discussion The TCCs were grown directly on Au wire (0.5 mm in diameter) by a hot-filament plasma-enhanced CVD process. The CVD gases used in the process were H2 and CH4 with a flow ratio of 10:50. The base pressure and reactive pressure were 0.5 Pa and 25 Torr, respectively. The DC plasma discharge was operated at 50 mA and 500 V. The distance between filament and substrate was 1 cm, and the filament temperature was controlled at 1800 1C by changing the filament current. During growth, the substrate temperature was kept at 750 1C, while the growth time was 30 min. The Au wires used as a substrate were put perpendicular to the filament. Pt wires were also put near the Au wires, because Pt catalyzes the growth of the TCCs during the plasma-enhanced CVD process. The surface morphologies and microstructures of asgrown TCCs were characterized by using SEM and TEM, respectively. Field emission measurements on TCCs were carried out in a home-made test system with an anodic probe with a tip area of about 1 mm2. The anodic probe moves only in the direction of the X-axis, while the sample can be adjusted in the directions of the Y- and Z-axes. Therefore, we can adjust the relative distance between anodic probe and the Au wire so that the tip of the anodic probe goes close to the Au wire. During testing, the

Figs. 1(a), (b) and (c) show different surface morphologies from the top along the Au wire substrate, which are named by region A, B and C, respectively. It can be seen that in region A there is a low density of short TCCs; a large number of long- and higher-density TCCs are grown in region B; but no TCCs are observed in the following region C. So, the middle region (B) is more suitable for growing TCCs than the other two regions, due to an appropriate distribution of electric field and temperature. The TCCs grown in region B are aligned perpendicular to the substrate with an average length in the range of 2–3 mm, but a few reach a length of 4–5 mm, as shown in Fig. 1(b). Fig. 2(a) shows a TEM image of a single TCC, which has a long conical shape, with a herringbone hollow core clearly observed. The hollow core is approximately 5–10 nm in diameter and extends throughout the whole length of the cone. Fig. 2(b) exhibits an ultrahighresolution TEM image, which shows that the outermost layer is amorphous carbon about 2–3 nm thick, and that between the hollow core and the amorphous carbon is a graphite layer with typical graphitic interlayer spacing of about 0.343 nm. The thickness of the graphite layer increases along the whole length of the cone from top to

Fig. 1. The surface morphologies. SEM images, from the top, along the Au wire. (a) Region A, with a low density of short TCCs. (b) Region B, with a high density of long TCCs. (c) Region C, without any TCCs.

Fig. 2. (a) SEM and (b) TEM images of a single tubular carbon cone.

ARTICLE IN PRESS J.J. Li et al. / Ultramicroscopy 107 (2007) 861–864

863

Fig. 3. (a) EDX spectra and (b) the diffraction pattern of a single tubular carbon cone.

root. Fig. 3(a) shows the typical chemical composition of a single cone. Strong C and Pt peaks can be seen in the energy dispersive X-ray (EDX) spectrum. The Cu peak and the weak Ta peaks arise from the Cu grids and CVD filament, respectively. Thus, the TCCs contain both carbon (graphite) and Pt phases. Apparently Pt, after serving as a catalyst, ends up at the top of the cone. As evidence, the diffraction pattern of the top of a single TCC is shown in Fig. 3(b). The region marked by a dotted circle can be indexed as that of Pt crystalline particle, recorded from the [0 1 1] zone axis. This diffraction pattern does not clearly show information about the graphite structure because a stronger signal from crystalline Pt conceals that of the graphite. There is a question as to how Pt reaches the Au wire substrate from the nearby Pt wire. For this, an evaporation transformation process is proposed. During the growth process of TCCs, it is assumed that the Pt wire near the Au wire is placed vertically; the plasma tends to discharge at the tip, heating it to close to the melting point of platinum (1772 1C). Such a high temperature will lead to the evaporation of a Pt droplet, which then condenses onto the surface of the Au wire. The evaporated Pt droplet then acts as a catalyst particle for TCC growth. Reported CVD growth processes for cone-shaped carbon nanostructures use two types of gas source. One is a mixture of NH3 and C2H2 gas [6,7]; the other is a mixture of H2 and CH4, as used in our work [4,5]. For the carbon cones grown using NH3 and C2H2 gas, it has been suggested that there are two growth directions rather than one: vertical growth (i.e., catalytic growth via diffusion of C through the catalyst particle); and lateral growth (by precipitation of C from the discharge at the outer wall of the carbon cone). Consequently, a conical structure, or a carbon nanocone, can be formed. For the gas mixture H2 and CH4, a different growth mechanism is proposed. This is that the formation of a TCC is a result of competition between the growth (associated with the CH4) and etching (associated with the H2) of graphite sheets surrounding the central nanotube in the plasma. This is based on the initial nanotube maintaining the inner hollow core. This nano-

Ion etching direction Etched Pt particle Pt particle Graphite layer

Before etching

After etching

Fig. 4. Schematic of the gradual reduction in size of the Pt particle causing TCC formation, due to plasma etching.

tube is surrounded by helical sheets of graphite coiling around it; consequently, the outer layer appears conical due to continuous coiling of sheets around one another. At the Au wire substrate, when growth is dominant in comparison with etching, a high-aspect ratio TCC can be formed; otherwise, a low-aspect ratio cone will be formed. Another crucial factor is the change in the size of the Pt catalyst particle at the TCC tip during CVD growth. Fig. 4 shows a schematic of the gradual reduction in size of the Pt particle causing TCC formation. At the beginning of growth, the size of the catalyst particle does not change, and the result should be a carbon nanotube of constant diameter. When the CVD plasma is enhanced, the size of the catalyst particle at the TCC tip is gradually reduced, due to a plasma-etching effect. The gradually diminishing catalyst size causes the nanotube diameter to change with growth time, resulting in a long-cone configuration. This is the key mechanism for obtaining the TCCs. Field emission characteristics from the three different regions A, B and C on the Au wire were measured, as shown in Fig. 5. Owing to most TCCs being grown on the region of 1.4 mm in length form the top along Au wire, we defined this region facing perpendicularly to the anode probe as the effective emission area, which is estimated to be about 1.1 mm2. Obviously, the TCCs grown in region B have the best field emission properties, with a very lowthreshold field of 0.27 V/mm, corresponding to a current

ARTICLE IN PRESS J.J. Li et al. / Ultramicroscopy 107 (2007) 861–864

864

4. Conclusions

2.5

J (mA/cm2)

2.0

B

1.5

A

1.0

0.5

C

0.0 0.1

0.2

0.3

0.4 0.5 E (V/µm)

0.6

0.7

0.8

Fig. 5. Field emission measurements, from the regions A, B and C, on asgrown TCCs.

density of 1 mA/cm2. When the applied electric field is increased to 0.5 V/mm, the emission current density can reach 1 mA/cm2. This is the minimum current density required to obtain a brightness of 300 cd/m2 from a VGA display with a typical high-voltage phosphor screen efficiency of 9 lm/W. We also find that these high emission current densities are stable. These emission characteristics of TCCs are better than those of normal carbon nanotubes and carbon nanofibres, that typically have a threshold field of 1–2 V/mm and a current density of 1 mA/cm2. These good field emission properties of TCCs are, firstly, attributed to having highaspect ratio, which causes a high-field enhancement factor. Secondly, their conical bases reduce the screening effects because there is sufficient distance between adjacent tubular cones. Both factors are favorable to field emission enhancement [8].

TCCs with the shape of a long cone and with a herringbone hollow interior, surrounded by helical sheets of graphite coiling around this, have been grown successfully. A growth mechanism has been proposed, namely that the gradually diminishing catalyst-particle size (induced by plasma-etching effects during growth) is the key mechanism in TCC formation. Field emission measurements show that as-grown TCCs have excellent field emission properties with a low-threshold field and stable high-current density at higher fields. We consider them as having potential use in field emission display devices, and conclude that further investigations should be made into their growth mechanism and into their field emission properties. Acknowledgments This work was supported by National Natural Science Foundation of China (Grant nos. 5047207 and 60671048) and National Center for Nanoscience and Technology, China. References [1] A. Krishnan, E. Dujardin, M.M.J. Treacy, J. Hugdahl, S. Lynum, T.W. Ebbesen, Nature 388 (1997) 451. [2] Y. Gogotsi, S. Dimovski, J.A. Libera, Carbon 20 (2002) 2263. [3] S. Iijima, M. Yudasaka, R. Yamada, S. Bandow, K. Suenaga, F. Kokai, K. Takahashi, Chem. Phys. Lett. 309 (1999) 165. [4] C.M. Radhika, X. Li, K.S. Mahendra, R. Krishna, Nanoletters 3 (2003) 671. [5] G.Y. Zhang, X. Jiang, E.G. Wang, Science 300 (2003) 472. [6] V.I. Merkulov, M.A. Guillorn, D.H. Lowndes, M.L. Simpson, Appl. Phys. Lett. 79 (2001) 1178. [7] M. Chhowalla, K.B.K. Teo, C. Ducati, N.L. Rupesinghe, G.A.J. Amaratunga, A.C. Ferrari, D. Roy, J. Robertson, W.I. Milne, J. Appl. Phys. 90 (2001) 5308. [8] C.J. Edgcombe, U. Valdre´, Philos. Mag. B 82 (2002) 987.