Rapid Growth of Single-Wall Carbon Nanotubes by Pulsed Laser CVD
Z. Liu, D. J. Styers-Barnett, A. A. Puretzky, C. M. Rouleau, I. N. Ivanov, K. Xiao, and
D. B. Geohegan (CNMS Staff); D. Yuan and J. Liu (CNMS Users, Duke University)
The nucleation and rapid growth of single-wall carbon nanotubes (SWNTs) were explored by pulsed-laser assisted chemical vapor deposition (PLA-CVD). A special high-power, Nd:YAG laser system with tunable pulse width (> 0.5 ms) was implemented to rapidly heat (>30,000 °C/s) metal catalyst-covered substrates to different growth temperatures for very brief (sub-second) and controlled time periods as measured by in situ optical pyrometry. Utilizing growth directly on transmission electron microscopy grids, exclusively SWNTs were found to grow under rapid heating conditions, with a minimum nucleation time of > 0.10 s. By measuring the length of nanotubes grown by single laser pulses, extremely fast growth rates (up to 100 microns/s) were found to result from the rapid heating and cooling induced by the laser treatment. Subsequent laser pulses were found not to incrementally continue the growth of these nanotubes, but instead activate previously inactive catalyst nanoparticles to grow new nanotubes. Localized growth of nanotubes with variable density was demonstrated through this process, and was applied for the reliable direct-write synthesis of SWNTs onto pre-patterned, catalyst-covered metal electrodes for the synthesis of SWNT field-effect transistors.
Understanding the rapid growth kinetics of single-wall carbon nanotubes by chemical vapor deposition is currently important to enable strategies to scale their synthesis. More importantly, however, their important physical properties are believed to be determined at nucleation, when their diameter and chirality (and thus, their metallic vs. semiconducting properties) are set. Theoretical studies on nanotube nucleation and growth mechanisms predict that once nanotube nuclei are formed, growth should proceed through further incorporation of carbon at the catalyst-tube interface. One major experimental limitation has been the inability to sufficiently limit growth times to understand and control nucleation and early growth. The PLA-CVD technique developed here is a powerful tool to investigate the fundamental mechanisms of nanowire and nanotube nucleation and growth as well as a promising method for the localized growth of nanostructures for device applications.
“Pulsed Laser CVD Investigations of Single-Wall Carbon Nanotube Growth Dynamics,” Z. Liu, D. J. Styers-Barnett, A. A. Puretzky, C. M. Rouleau, I. N. Ivanov, K. Xiao, D. Yuan, J. Liu and D. B. Geohegan, Applied Physics A (in press).
This research was conducted in part at the Center for Nanophase Materials Sciences, which is sponsored by the Division of Scientific User Facilities, and in part by the Division of Materials Sciences and Engineering (Measurements of NT growth), U. S. Department of Energy.
(a) Schematic of PLA-CVD vacuum chamber. (b) Time-dependent temperature profiles of a Si/SiO2 wafer irradiated by (a) a single 50ms-wide laser pulse, and (b) from a macropulse of 25 pulses of 5 ms width. (c) SEM of nanotubes grown using twenty 5-ms laser pulses on a Fe/Al2O3 catalyst-coated TEM grid under methane, hydrogen, ethylene, and argon atmosphere. Only SWNTs were found. TEM image in inset shows a 3.8-nm-diameter SWNT. (d) Typical Raman spectrum of SWNTs grown on Si, inset showing detail of the radial breathing modes. (e) AFM image and diameter distribution (inset) of SWNTs grown on Si from 6000 laser pulses. (f) Statistics of nanotube length resulting from irradiation by (200-pulse) macropulses. Boxes represent 50% of measured values, lines are 5-95%, while the open circles are outliers. The horizontal line is the median length calculated for each growth experiment, and does not change appreciably with repeated irradiation, despite increasing nanotube density.
(a) Schematic of local direct-writing of SWNTs for device applications. (b) An in-situ contacted nanotube bridging two electrodes inside the local hot zone. Arrow identifies the nanotube. (c) No nanotube growth from adjacent electrodes outside the heating zone. (d) Schematic diagram of a back-gated SWNT field-effect transistor. (e) I-Vg curve of the transistor exhibiting ambipolar behavior. Vds = 0.5 V.