Double-Walled Boron Nitride Nanotubes Grown by Floating Catalyst

Sep 13, 2008 - ... 1270 North Fairfield Road, Dayton, Ohio 45432, and School of Materials Engineering and Birck Nanotechnology Center, Purdue Universi...
0 downloads 0 Views 1MB Size
NANO LETTERS

Double-Walled Boron Nitride Nanotubes Grown by Floating Catalyst Chemical Vapor Deposition

2008 Vol. 8, No. 10 3298-3302

Myung Jong Kim,†,‡,§ Shahana Chatterjee,‡ Seung Min Kim,| Eric A. Stach,| Mark G. Bradley,‡ Mark J. Pender,† Larry G. Sneddon,*,‡ and Benji Maruyama*,† Materials and Manufacturing Directorate, Air Force Research Laboratory, 2941 Hobson Way, Wright Patterson Air Force Base, Ohio 45433, Department of Chemistry, UniVersity of PennsylVania, 231 South 34th Street, Philadelphia, PennsylVania 19104, UniVersal Technology Corporation, 1270 North Fairfield Road, Dayton, Ohio 45432, and School of Materials Engineering and Birck Nanotechnology Center, Purdue UniVersity, 1205 West State Street, West Lafayette, Indiana 47907 Received June 12, 2008; Revised Manuscript Received August 10, 2008

ABSTRACT One-dimensional nanostructures exhibit quantum confinement which leads to unique electronic properties, making them attractive as the active elements for nanoscale electronic devices. Boron nitride nanotubes are of particular interest since, unlike carbon nanotubes, all chiralities are semiconducting. Here, we report a synthesis based on the use of low pressures of the molecular precursor borazine in conjunction with a floating nickelocene catalyst that resulted in the formation of double-walled boron nitride nanotubes. As has been shown for carbon nanotube production, the floating catalyst chemical vapor deposition method has the potential for creating high quality boron nitride nanostructures with high production volumes.

Carbon nanotubes (CNTs) have been of intense research interest owing to their remarkable electronic, mechanical, and optical properties.1 Unfortunately, their use in many nanoelectronic applications has been limited by the fact that semiconducting nanotubes are intimately mixed with metallic nanotubes in as-grown samples. Boron nitride nanotubes (BNNTs) are structural analogues of carbon nanotubes, where alternating boron and nitrogen atoms replace carbon atoms in a hexagonal lattice structure. This elemental change results in a number of potential advantages over carbon nanotubes. These include a uniform band gap (∼5 eV) that is insensitive to either diameter or chirality,2,3 band gap tunability by the giant Stark effect,4,5 and oxidation resistance up to 800 °C. These properties make BNNTs attractive for nanoscale applications in, for example, nanoelectronics, optoelectronics, and nanocomposites. Carbon nanotubes can be synthesized with relative ease in laboratory and commercial settings via simple catalytic chemical vapor deposition (CCVD) techniques, thus allowing widespread access for research and applications. On the other

hand, BN nanotubes have proven more difficult to reliably synthesize on a large scale, and accordingly much less is known about their properties and utility in potential applications. Previous methods for BNNT syntheses have employed many of the routes that have already been developed for CNT syntheses, including arc discharge,6 laser ablation,7 ball milling,8 substitution reactions using carbon nanotubes,9 pyrolysis of polymeric precursors,10 and chemical vapor deposition.11,12 However, the floating catalyst CVD techniques that have been exploited to provide high quality, large scale syntheses of CNTs have not yet been achieved for BNNT synthesis.

* To whom correspondence should be addressed. (B.M.) E-mail: [email protected]. Tel: 937-255-0042. (L.G.S.) Tel: 215-8988632. E-mail: [email protected]. † Air Force Research Laboratory. ‡ University of Pennsylvania. § Universal Technology Corporation. | Purdue University.

As described in more detail in the Supporting Information (S1), the experimental CCVD setup consisted of a tube furnace equipped with an alumina pyrolysis-chamber, along with the appropriate mass flow controllers, bubblers, and pumping system that allow for the controlled delivery of the

10.1021/nl8016835 CCC: $40.75 Published on Web 09/13/2008

 2008 American Chemical Society

Here, we report that the use of low pressures of the molecular precursor borazine in conjunction with a floating nickelocene catalyst results in the predominant formation of double-walled BNNT structures. This floating catalyst method has the potential to produce BNNT materials continuously in a manner similar to the HiPco13 or CoMoCat14 processes used in the large scale production of carbon nanotubes.

Figure 1. (a) Low- and (b) high-resolution TEM images, (c) SEM image, and (d) EELS data taken from the boron nitride nanofibers produced via floating CCVD at a higher vapor pressure of borazine.

Figure 2. (a) SEM image, (b) low- and (c, d) high-resolution TEM images taken from the double-walled boron nitride nanotubes produced via floating CCVD at lower borazine pressures.

borazine, nickelocene, and ammonia vapors. Nickelocene was obtained from Strem, while borazine was synthesized and purified as previously reported.15 Under an ammonia flow, the supply bubbler containing the solid nickelocene (100 mg) was heated with a heating tape at 80 °C, while at the same time a separate bubbler containing the liquid borazine was Nano Lett., Vol. 8, No. 10, 2008

cooled under a nitrogen flow to the desired temperature using an acetone and dry ice cooling bath. The two gas streams were allowed to combine as they flowed into the hot zone of the furnace. The relative nickelocene to borazine ratio was controlled by both the temperatures of the bubblers and flow rates of the ammonia and nitrogen gases. 3299

In a typical run, the flow ratio of ammonia to nitrogen was 100 to 3 sccm, and the growth temperature was 1200 °C. After several hours of reaction under these conditions, grown materials were deposited on a collection grid or a filter located in the down stream of the pyrolysis tube. These as-grown samples were then ultrasonicated in ethanol, and the resulting solution dropped on a holey carbon TEM grid to enable study by transmission electron microscopy (TEM) and electron energy loss spectroscopy (EELS). The TEM and EELS analyses at University of Pennsylvania employed a JEOL 2010F, while those at Purdue University used a FEI Titan 80-300 S/TEM. Because of its volatility, ideal 1:1 B/N ratio, and preformed boron-nitride-like ring structure, borazine was selected as the chemical precursor for BN nanotube growth in our studies. Previously Lourie et al. had used a feed stream containing in situ generated borazine and hydrogen to produce multiwalled BNNT growth on nickel boride catalyst particles,16 but with in situ generation it is difficult to control the reactant pressure and ratios. We used borazine that had been synthesized and purified in advance, and by simply cooling the borazine at different temperatures, it was possible to carefully control its concentration in the reaction zone. As discussed below, such control proved important in the formation of different types of BN structures. As illustrated by the SEM and TEM images in Figure 1a-c, a reaction where the vapor pressure of borazine was maintained at 166 torr (20 °C, bubbler temperature, as detailed in the Supporting Information (S2)) produced a mixture of nanofibers and overcoated nanoparticle structures. The high-resolution TEM image in Figure 1b revealed that some of the nanostructures were similar to multiwalled nanotubes with a hollow core along the center, but had poor crystallinity. These nanofiber type structures were 500 nm to 1 µm long and had 20-30 nm diameters. The EELS data (Figure 1d) indicated they were pure BN, showing only the boron and nitrogen K-edges. The SEM images in Figure 2a revealed that many bundlelike structures were produced when the growth reaction was carried out with a borazine pressure of only 22 torr (-20 °C, bubbler temperature). The low resolution image in Figure 2b appears to be of single-walled nanotubes, but further analyses by high-resolution (HR) TEM (images in Figure 2c, d) showed that the structures are crystalline double-walled nanotubes with ∼2 nm diameters and lengths ranging from 100 to 200 nm. The mass yield of as-grown materials was approximately 20 mg/h with, as estimated from the HRTEM images, over 70% of the observed BN nanotubes being double-walled along with much smaller amounts of single- and few-walled nanotubes. Catalyst particles were frequently observed at the end of BN nanotubes (Supporting Information (S3)). As estimated from the low-resolution TEM images, one out of 20-30 nickel nanoparticles appeared to be activated for BN nanotube growth. Analogous control experiments carried out without nickelocene did not result in nanotube growth and thus provide further evidence for the catalytic activity. 3300

Figure 3. EELS data from (a) as-grown and (b) oxidized boron nitride double-walled nanotubes. (c) Inset shows magnified carbon K-edges from Figure 3a. Data demonstrate the removal of carbon by oxidation.

The EELS data in Figure 3a collected from the as-grown double-walled nanotubes again showed the sharp boron and nitrogen K-edges that are characteristic of the sp2 hybridized atoms of boron nitride, but unlike in the nanofiber EELS analyses, the spectra also showed small carbon K edges. This carbon most likely came from decomposed nickelocene catalyst. The amount of carbon impurity seen in the BN nanofibers was negligible, due to the higher pressure of borazine that was used during their syntheses. On the basis of previous reports,17,18 the observed broadly distributed carbon K edges (Figure 3c), corresponding to π* transitions, can be assigned to amorphous carbon. The high resolution TEM image (Figure 4a) also supports this interpretation, showing that the coatings on the outer surface of the tube appear amorphous. Similar carbon K-edges were also observed by Arenal et al. for single-walled BN nanotubes grown by laser vaporization and were interpreted as arising from carbon contamination during the growth process.19 To further investigate whether the carbon detected in EELS (Figure 3a) was part of the amorphous coating or carbon incorporated into the BN-graphene lattice, a sample was heated in air to remove the carbon. Oxidation of the as-grown materials in air at 600 °C for 30 min in a tube furnace Nano Lett., Vol. 8, No. 10, 2008

Figure 4. TEM images of BNNTs: (a) As-grown double-walled boron nitride nanotube showing amorphous carbon contaminants coating the nanotube, and (b) double-walled boron nitride nanotube after oxidation in air at 600 °C for 30 min showing that the removal of carbon contaminants while the structural integrity of the tube was preserved. This indicates that carbon is not incorporated to any significant extent in the BN lattice.

changed their color from black to gray (Supporting Information (S4)). Carbon-containing materials including carbon nanotubes20 would have been oxidized under these conditions. Therefore, if the as-grown tubes contained any structural carbon (i.e., BxCyNz nanotubes), the tubes would have exhibited structural damage following oxidation.21 The TEM images taken from the gray materials (Figure 4b) showed that the structural integrity of double-walled nanotubes was preserved after oxidation. In addition, the amorphous contaminants on the outer tube surface were largely removed, and the carbon K-edges disappeared from the EELS data (Figure 3b). These results prove that carbon atoms were not incorporated into the BN network of the grown nanotubes. The predominant production of double-walled BN nanotubes was previously observed by Cumings et al. using a plasma-arc method22 and by Golberg et al.23 in the metal oxide promoted reaction of B2O3 with carbon nanotubes in a nitrogen atmosphere. Initially, it was proposed that “lip-lip” interactions were important in stabilizing the growing open edges of carbon nanotubes by forming bridge bonds between adjacent layers.24 In contrast to a C-C bond, a B-N bond is highly polar. Thus, “lip-lip” interactions such as those described above should be even more important in the formation of BN nanotubes, even during catalyzed growth, and should favor the formation of double-walled nanotubes.25 Kinetics could also play a role. The fact that double-walled BN nanotubes were produced at low borazine pressures suggests that the growth rate is limited by the supply of BN precursor to the catalyst particles, whereas multiwalled nanotube growth is thought to be limited by the diffusion through the catalyst particles, analogous to the growth processes for single-walled and doubled-walled carbon nanotube growth.26 The presence of catalyst particles with diameters larger than the associated BN nanotubes could also be evidence that nanotube growth is supply limited. However, a clear mechanism for double-walled nanotube formation has not been clearly elucidated in any nanotube system and remains an open question. Nano Lett., Vol. 8, No. 10, 2008

In conclusion, BN nanotubes were synthesized via catalytic CVD using the floating catalyst, nickelocene. Depending on the borazine pressure, the products are either BN nanotubes (lower pressure) or BN nanofibers (higher pressure). TEM and EELS data indicate that the as-grown BN nanotubes are double-walled and are pure BN nanotubes with a high crystalline quality. It is significant that BN nanotubes can be produced with a well-defined BN precursor and a floating catalyst in a manner similar to carbon nanotube CCVD synthesis, since the CCVD method has the potential to be carried out under the continuous conditions needed for the high volume production of BN nanotubes. Acknowledgment. Authors gratefully acknowledge support from the AFOSR, Air Force Office of Scientific Research through the UTC contract (Collaborative Research and Development (CR&D) F33615-03-D-5801). S.M.K and E.A.S. acknowledge additional support from the Army Research Office. Supporting Information Available: (S1) Schematic diagram of the CCVD setup for BN nanotube growth; (S2) the vapor pressure of borazine; (S3) the catalytic effect of nickel nanoparticles; and (S4) pictures taken from (a) asgrown and (b) oxidized materials. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Dresselhaus M. S.; Dresselhaus, G.; Avouris, P. Carbon Nanotubes Synthesis, Structure, Properties and Applications; Springer: New York, 2001. (2) Blase, X.; Rubio, A.; Louie, S. G.; Cohen, M. L. Europhys. Lett. 1994, 28, 335. (3) Czerw, R.; Webster, S.; Carroll, D. L.; Vieira, S. M. C.; Birkett, P. R.; Rego, C. A.; Roth, S. Appl. Phys. Lett. 2003, 83, 1617. (4) Ishigami, M.; Sau, J. D.; Aloni, S.; Cohen, M. L.; Zettl, A. Phys. ReV. Lett. 2005, 94, 56804. (5) Khoo, K. H.; Mazzoni, M. S. C.; Louie, S. G. Phys. ReV. B 2004, 69, 201401. (6) Loiseau, A.; Willaime, F.; Demoncy, N.; Schramchenko, N.; Hug, G. Carbon 1998, 36, 743–752. (7) Golberg, D.; Bando, Y.; Eremets, K.; Takemura, K.; Kurashima, K.; Yusa, H. App. Phys. Lett. 1996, 69, 2045–2047. 3301

(8) Chen, Y.; Gerald, J. F.; Williams, J. S.; Bulcock, S. Chem. Phys. Lett. 1999, 299, 260–264. (9) Han, W. Q.; Mickelson, W.; Cumings, J.; Zettl, A. Appl. Phys. Lett. 2002, 81, 1100–1112. (10) Bechelany, M.; Bernard, S.; Brioude, A.; Cornu, D.; Stadelmann, P.; Charcosset, C.; Fiaty, K.; Miele, P. J. Phys. Chem. C 2007, 111, 13378–13384. (11) Ma, R.; Bando, Y.; Sato, T. Chem. Phys. Lett. 2001, 337, 61–64. (12) Guo, L.; Singh, R. N. Nanotechnolgy 2008, 19, 065601. (13) Bronikowski, M. J.; Willis, P. A.; Colbert, D. T.; Smith, K. A.; Smalley, R. E. J. Vac. Sci. Technol., A 2001, 19, 1800–1805. (14) Kitiyanan, B.; Alvarez, W. E.; Harwell, J. H.; Resasco, D. E. Chem. Phys. Lett. 2000, 317, 497–503. (15) Wideman, T.; Sneddon, L. G. Inorg. Chem. 1995, 34, 1002–1003. (16) Lourie, O. R.; Jones, C. R.; Bartlett, B. M.; Gibbons, P. C.; Ruoff, R. S.; Buhro, W. E. Chem. Mater. 2000, 12, 1808–1810. (17) Stephan, O.; Ajayan, P. M.; Colliex, C.; Cyrot-Lackmann, F.; Sandre, E. Phys. ReV. B 1996, 53, 13824–13829.

3302

(18) Wang, W. L.; Bai, X. D.; Liu, K. H.; Xu, Z.; Golberg, D.; Bando, Y.; Wang, E. G. J. Am. Chem. Soc. 2006, 128, 6530–6531. (19) Arenal, R.; Stephan, O.; Cochon, J.-L.; Loiseau, A. J. Am. Chem. Soc. 2007, 129, 16183–16189. (20) Xu, Y.-Q.; Peng, H.; Hauge, R. H.; Smalley, R. E. Nano Lett. 2004, 5, 163–168. (21) Han, W.-Q.; Mickelson, W.; Cumings, J.; Zettl, A. Appl. Phys. Lett. 2002, 81, 1110–1112. (22) Cumings, J.; Zettl, A. Chem. Phys. Lett. 2000, 316, 211–216. (23) Golberg, D.; Bando, Y. Appl. Phys. Lett. 2001, 79, 415–417. (24) Kwon, Y.-K.; Lee, Y. H.; Kim, S.-G.; Jund, P.; Tomanek, D.; Smalley, R. E. Phys. ReV. Lett. 1997, 79, 2065–2068. (25) Blase, X.; Vita, A. D.; Charlier, J.-C.; Car, R. Phys. ReV. Lett. 1998, 80, 1666–1669. (26) Hafner, J. H.; Bronikowski, M. J.; Azamian, B. R.; Nikolaev, P.; Rinzler, A. G.; Colbert, D. T.; Smith, K. A.; Smalley, R. E. Chem. Phys. Lett. 1998, 296, 195–202.

NL8016835

Nano Lett., Vol. 8, No. 10, 2008