Skip to content Skip to navigation

Metal-Catalyst-Free CVD Synthesis of Single Wall Carbon Nanotubes

Written by: 
Hyung Gyu Park
There are several advantages to metal-catalyst-free Carbon nanotubes, from use in a broader range of applications to easier purification, mitigated toxicity concerns, and lower prices of pure CNT materials. Synthesis of single wall carbon nanotubes (SWNT) on non-metallic catalysts, as recently carried out by Liu B. et al. and Huang S. et al., demonstrates that one could generalize standard synthesis processes to capture the fundamental physics underlying the CNT nucleation and growth.

A review by Hyung Gyu Park, Ph.D, Assistant Professor, D-MAVT, ETH Zurich, Zurich, Switzerland

Since the unique physical properties of Carbon nanotubes (CNT) were recognized, critical research has centered on tailoring the properties of nanotubes at the synthesis stage. The unique properties of CNT are believed to arise from their nanoscale morphology; therefore the primary synthesis issues include controlling nanotube geometry such as diameter, length, chirality, etc. A host of research during the last decade has pointed out that reactant gas composition, thermodynamic conditions of the reactor chamber, additives and catalysts are some of the most important control factors in the CVD synthesis of CNT. Of these, catalysts remain the primary but less thoroughly-explored factor. Until a few years ago, it was believed that CVD synthesis of single wall carbon nanotubes (SWNT) should incorporate transition metal nanoparticles, such as Fe, Ni, and Co, as catalysts. However, recent investigations on SWNT CVD are expanding this range of catalyst materials toward non-transition metals, noble metals, and most promisingly toward non-metallic materials.

CNT growth on non-metallic nanostructures can facilitate application of the CNT materials to semiconductor technology, pharmaceutical technology, membrane technology and basic sciences. Metal-free SWNT will remove a detrimental incompatibility issue in semiconductor electronics fabrication (CMOS) process that has not allowed adaptation of CNTs, as-synthesized or transplanted, onto an electronics substrate because of the contamination issues associated with metal catalyst particles. Metal-catalyst-free CNTs would enable much easier purification, mitigate a toxicity concerns, and lower the price of pure CNT materials, which may further extend the scope of these material applications. Membrane- or CNT-Nanofluidics-based research and industry could also benefit from the metal-free CNT in order to bypass one or more manufacturing steps, leading toward cost-effectiveness.        

Liu Figure 2a-b
Typical SEM and AFM images of the as-grown SWNTs on Si/SiO2 substrate covered with a 30-nm-thick sputtering deposited SiO2 film.
One widely accepted theory in the CNT synthesis on the Fe-family catalyst relies on a bottom growth process. In this technique, carbon-containing radicals or intermediate species are first broken out from precursor molecules through pyrolysis or decomposition on catalyst surfaces. Once they dissolve into a metal nanoparticle that becomes nearly molten due to a lower melting point effect at nanoscale, they would precipitate as a carbon cap around the nanoparticular catalyst surface and then push the cap up as a hollow pillar, i.e. a nanotube. This mechanism was later generalized to include other metals such as gold, silver, platinum and aluminum in a way that any metal nanoparticle can catalyze CNT growth as long as it turns to a nearly molten state by size reduction and sticks on the substrate during the growth. Switching gears, synthesis of single wall carbon nanotubes (SWNT) on non-metallic catalysts, recently carried out by Liu B. et al. and Huang S. et al., shows that one could further generalize this process to capture the fundamental physics underlying the CNT nucleation and growth. Detailed understanding of metal-catalyst-free CNT synthesis by coordination of experimental and theoretical endeavors will lead to better control of the morphology and properties of this material at its synthesis stage.

Huang Figure 2
SEM images of SWNTs from the scratched substrates by diamond blade after CVD at 900 °C for 10 min: (A-C) Si waferwith 1 ¼ SiO2 layer at different positions; (D-F) quartz plate with different magnifications.
In these communications, Liu et al. sputter deposit 30-nm-thick SiO2 nanoparticles onto a Si or thermal SiO2 surface. Subsequent thermal ramping/annealing to a CNT synthesis temperature (900 oC) creates sub-2-nm nanoparticles on the surface that pose catalytic activity for CNT growth. An intertwined SWNT mesh forms on top of this surface with a number density as high as 1010 cm-2. Another demonstration by the same group and Huang et al. synthesized individual SWNTs on smooth, thin film metal oxides, catalyzed in turn by scratched nanoparticle debris. One plausible mechanism proposed by Huang et al. assumes that sub-2-nm metal oxide nanoparticles have increased fluctuation at an atomic scale. The radical fluctuation of atoms may entail the generation of space holes and dislocations into which carbonaceous gas species could enter. Popping up of saturated carbon then nucleates a CNT. These authors claim that any material with a suitable size can serve as a nucleation center.

SWNTs free of metal catalysts push the envelope of CNT synthesis and their potential applications. A narrow size distribution of oxide nanoparticles can engender a narrow diameter distribution of as-grown SWNTs: 50% distributed within the ± 0.1-nm window; 84% within ±  0.3-nm. Unlike metal nanoparticle catalysts, oxide nanoparticles are not likely to diffuse on the surface or agglomerate on one another to enlarge the catalyst particle size. This effect implies that a precise control of CNT diameter could be possible at the catalyst preparation stage. As additional advantages, the metal-free characteristic can increase the compatibility with other micro/nanomanufacturing techniques; avoid generation of malicious defects during the purification process; and enable cost-effective simplification of any device fabrication by ruling out selective metal etching steps. However, challenges associated with the metal-free technique include catalyst density control, uniform distribution of CNT lengths, effect of additives such as water vapor during the CVD, and the role of hydrogen. Further systematic investigations will be able to address these challenges. Precise tuning of both SWNT diameter and chirality is on the horizon.

Images reproduced with permission from Liu B. et al. 2009. Metal-Catalyst-Free Growth of Single-Walled Carbon Nanotubes, J. Am. Chem. Soc. 131 (6): 2082-2083 and Huang S. et al. 2009. Metal-Catalyst-Free Growth of Single-Walled Carbon Nanotubes on Substrates, J. Am. Chem. Soc. 131 (6): 2094-2095. Copyright 2009 American Chemical Society.