NANO LETTERS

Carbon Nanopipettes
Radhika C. Mani, Xiang Li, Mahendra K. Sunkara,*, and Krishna Rajan*,
Department of Chemical Engineering, UniVersity of LouisVille, LouisVille, Kentucky 40292, and Department of Materials Science and Engineering, Rensselaer Polytechnic Institute, Troy, New York 12180
Received February 28, 2003; Revised Manuscript Received March 12, 2003

2003 Vol. 3, No. 5 671-673

ABSTRACT
We report the growth and detailed structural investigation of a new morphological manifestation of carbon-based nanostructures in the form of tapered whiskers with uniform 13 nm hollowness. The base of the whiskers is in the submicron scale, tapering uniformly to form a pointed tip in the form of a pipette. These hollow nanopipettes have a shell containing helical graphitic sheets. We propose that these whiskers grow by competing phenomena of growth and etching that control their aspect ratio.

Carbon exhibits many interesting amorphous and crystalline phases, each showing a variety of one-, two-, and threedimensional morphologies. The discovery of fullerenes (C60 cage compounds) pioneered the research efforts on novel carbon nanostructures.1 A detailed structural investigation of carbon nanotubes was given by Iijima in 1991.2 Since then, a number of novel one-dimensional nanostructures have been reported such as helix-shaped graphite nanotubes,3 graphitic nanocones,4 nanohorns,5 conical crystals,6,7 and microtrees.8 In this letter we discuss another novel nanostructure of carbon termed as nanopipette, characterized by a hollow conical structure, whose hollowness remains constant throughout the length of the structure. Platinum substrates in the form of wires were immersed vertically into a microwave plasma in an ASTeX model 5010 chemical vapor deposition (CVD) reactor. The gas-phase composition was 1-2% of CH4/H2, which was atypical of carbon nanotube growth. At the end of the deposition experiment, some regions of the substrate were coated with a microcrystalline diamond film. In this region, along with diamond crystals, several whiskers that were about 200700 nm in diameter and 6 m in length emerged. These whiskers had a pointed tip (in the shape of a pipette), while the base was in the submicron regime (Figure 1). They also indicated some minor faceting on their surface. Transmission electron microscopy (TEM) was used for detailed structural investigation. These whiskers were analyzed in a JEOL model 2010 TEM. As seen in the bright field images (Figure 2), the whiskers are hundreds of nanometers long with a well-defined uniform hollow core. The core is approximately 1-3 nm in diameter, extending throughout the length of the whisker (Figure 2).
* Corresponding authors. M. Sunkara: phone 502-852-1558, fax 502852-6355, e-mail mahendra@louisville.edu. K. Rajan: phone 518-276-6126, fax 518-276-8554, e-mail rajank@rpi.edu. University of Louisville. Rensselaer Polytechnic Institute. 10.1021/nl034125o CCC: $25.00 Published on Web 04/01/2003 2003 American Chemical Society

Figure 1. High-resolution scanning electron microscope (SEM) image of the carbon nanowhiskers.

The dark field image in Figure 3A highlights this structure even further, showing the hollow core running across the entire length of a whisker. The details of the structure of the walls of the whisker require further work, but our preliminary observations do help provide some insight. At the tip of the whisker, where the thickness permitted a reasonable signal, an energy-filtered image using the sp2 core loss peak clearly illuminated the specimen (Figure 3A, inset). The dark region running down the axis of the whisker corresponds to the hollow core, which evidently does not contribute to any signal (in this case inelastically scattered core loss electrons). Based on the energy loss images, the walls of the whisker at least in the tip region appeared to be graphitic in nature. Basic basal plane lattice images confirmed this graphitic structure. Diffraction patterns from thicker regions of the whisker (Figure 3B), however, exhibited characteristic features of possible helical morphologies. The pitch angle associated with this type of structure can vary (in the case

Figure 2. Bright field image of whiskers, with a hollow core running axially down as indicated by arrows.

of region of the whisker sample shown in Figure 3B, this angle is 9 as shown in Figure 3C) giving rise to a more complex morphology. Hence it is suggested that these nanopipettes are hollow whiskers possibly made up of helical

sheets of graphite. Clearly this is an area requiring further study, which is presently underway. Conical structures of graphite have been studied long before the knowledge of carbon nanotubes. Some of the earliest research on one-dimensional nanostructures of graphite described a model for the growth of graphite whiskers.9 According to this model, a sheet of graphite (consisting of several monolayers) overlaps itself and continues to wind around the whisker axis many times, rather than joining opposite edges and forming concentric tubes. This scroll structure gave rise to a simultaneous thickening and growth along the length. In another attempt, while heating silicon carbide crystals graphite whiskers were obtained.10 The explanation for columnar growth was given on the basis of the classical model of whisker growth initiated by nucleation on a dislocation with a screw component perpendicular to the surface. A time dependent impurity adsorption hampers layer spreading. Another explanation for these needle-shaped crystals was that a flat graphite sheet rolled into a cone, and each time it wrapped around by lifting itself to form a rotation over a small angle. Graphite sheets are known to grow along their edges, thus if subsequent growth takes place, a helical surface will be formed by growth along steps (to form a spiral shape).11 Graphite polyhedral crystals, synthesized from gas phase, had folded and closed graphene planes different from ordi-

Figure 3. (A) Dark field TEM image with 00.4 Bragg reflection peak (inset is the corresponding energy filtered TEM (EFTEM) image from sp2 carbon core loss peak). (B) Diffraction pattern from hollow whiskers, with indexing consistent with helical graphite sheets. (C) Same as Figure 3b, but outlining the multiple sets of hexagonal symmetry associated with helical sheets with angle 9.

Figure 4. (A) Conical structures with a central nanotube. (B) The aspect ratio of the conical structure increases, giving rise to (C) nanopipettes.
672 Nano Lett., Vol. 3, No. 5, 2003

nary graphite but similar to multiwalled carbon nanotubes.6 There were small angle conical crystals also,7 which were as large as 3 m and 300 nm in diameter. These conical crystals were not nucleated from fullerenes and fullerenelike structures. Some of these crystals were hollow, consisting of straight and well-defined 110 nm walls. Their hollow tips had a herringbone type structure; however, beyond the tip this herringbone structure transformed into a conical cavity. None of the above conical crystals had a structure similar to the one described in this work. Hence, none of the above mechanisms could explain the formation of a whisker that is conical on the outside and has a uniform hollow core at the center. The temperatures in the center of the plasma could reach as high as 2000 C. Since the substrate is placed vertically, the plasma tends to discharge at the tip, making it close to the melting point of platinum. These high temperatures and the peculiar conical shape made us assume that the growth mode could be due to an evaporating catalyst, i.e., a platinum catalyst particle initiates the growth of multiwalled carbon nanotubes. As the whisker grows, the catalyst droplet evaporates (the temperatures in the dense zone of the plasma are high enough to evaporate platinum). As the catalyst droplet evaporates, the outer diameter of the multiwalled shell reduces. This hypothesis could be easily tested using a sequential step experiment by growing for shorter time scales and seeing the catalyst at the tip. Hence, in another experiment, platinum thin films (50 nm) were electrodeposited on platinum wires coated with microcrystalline diamond films. Diamond films acted as a nonreactive stable substrate. The platinum coating on the diamond film melted into droplets when exposed to the plasma, due to temperature gradients. The growth experiments were performed over a period of 1 h or less. The results of this experiment are shown in Figure 4. As shown in this figure, there is a continuous gradient of one-dimensional structures along the wire. The region close to the tip of the wire has a conical structure, whose core contains a multiwalled (or single-walled) carbon nanotube (Figure 4A). The nanotube is surrounded by graphite deposit. As we move along the length of the wire, there is competition between the etching and growth of crystalline phase (sp2) of carbon. Hence the central nanotube remains, while the surrounding graphite material also grows rapidly. Thus, a short distance away from the conical structures, we obtain structures with a higher aspect ratio (shown in Figure 4B), and further away we obtain nanopipettes (Figure 4C). The density of these nanopipettes gradually reduces as we move to the end of the substrate. If the model of an evaporating catalyst were true, then we should not be seeing such a variation of structure discontinuously. The morphology seen in Figure 4A would mean that the rate of evaporation changes abruptly along the length of the whisker, which would be quite impossible. Hence, these experiments clearly indicate that selective etching and simultaneous growth is responsible for the growth of these conical nanopipettes with different aspect ratios along the length of the substrate. Overall, there is an initial open-ended
Nano Lett., Vol. 3, No. 5, 2003

nanotube that grows out. Surrounding this nanotube, helical sheets of graphite coil around. Due to a continuous coiling around one another, the outer layer appears conical, while the initial nanotube maintains the inner hollow core. Depending on the placement in the plasma, there is a competitive growth and etching of the graphite sheets surrounding the central nanotube. At the tip of the substrate, etching seems to dominate giving rise to a low aspect ratio. As we move away from the plasma, growth seems to dominate, giving rise to high aspect ratio whiskers. These hollow onedimensional structures could show further enhanced properties beyond those of microneedles used currently for drug delivery12 and nanotubes used for field emission.13 We previously found that the nanocomposite carbon deposit found at the tip of platinum wire immersed directly in hydrocarbon plasma exhibits nearly reversible response to dopamine and other neurological solutes.14 Similarly, the novel carbon nanopipettes reported in this letter could be ideal candidates for simultaneous drug delivery and in-vivo detection of neurotransmitters. Such novel applications in drug delivery, neurological solute detection, and field emission based devices are the focus of our current efforts. In summary, we present a novel growth morphology of carbon grown from the gas phase with noble metals. This pipette-like shape grows due to the simultaneous growth and etching modes occurring in the plasma. Acknowledgment. One of the authors (M.K.S.) greatly appreciates partial financial support from NSF through CAREER grant (CTS 9876251). One of the authors (R.C.M.) gratefully acknowledges Speed Scientific School at University of Louisville for Grosscurth fellowship. Authors also appreciate Dr. Shashank Sharma at University of Louisville and Dr. Roy Gat at Coating Technology Solutions Inc., MA, for helpful discussions during the course of this study. References
(1) Kroto, H. W.; Heath, J. R.; OBrien, S. C.; Curl, R. F.; Smalley, R. E. Nature 1985, 318, 162. (2) Iijima, S. Nature 1991, 354, 56. (3) Amelinckx, S.; Zhang, X. B.; Bernaerts, D.; Zhang, X. F.; Ivanov, V.; Nagy, J. B. Science 1994, 265, 635. (4) Krishnan, A.; Dujardin, E.; Treacy, M. M. J.; Hugdahl, J.; Lynum, S.; Ebbesen, T. W. Nature 1997, 388, 451. (5) Iijima, S.; Yudasaka, M.; Yamada, R.; Bandow, S.; Suenaga, K.; Kokai, F.; Takahashi, K. Chem. Phys. Lett. 1999, 309, 165. (6) Gogotsi, Y.; Libera, J. A.; Kalashnikov, N.; Yoshimura, M. Science 2000, 290, 317. (7) Gogotsi, Y.; Dimovski, S.; Libera, J. A. Carbon 2002, 40, 2263. (8) Ajayan, P. M.; Nugent, J. M.; Siegel, R. W.; Wei, B.; Kohler-Redlich, Ph. Nature 2000, 404, 243. (9) Bacon, R. J. Appl. Phys. 1960, 31, 283. (10) Haanstra, H. B.; Verspui, G.; Knippenberg, W. F. J. Cryst. Growth 1972, 16, 71. (11) Amelinckx, S.; Luyten, W.; Krekels, T.; Van Tendeloo, G.; Van Landuyt, J. J. Cryst. Growth 1992, 121, 543. (12) Henry, S.; McAllister, D. V.; Allen, M. G.; Prausnitz, M. R. J. Pharm. Sci. 1998, 87, 922. (13) Tsai, C. L.; Chen, C. F.; Wu, L. K. Appl. Phys. Lett. 2002, 81, 721. (14) Mani, R. C.; Sharma, S.; Sunkara, M. K.; Gullapalli, J.; Baldwin, R. P.; Rao, R.; Rao, A. M.; Cowley, J. M. Electrochem. Solid State Lett. 2002, 5, E32-E35.

NL034125O
673