Repeatable High-Yield Carbon Nanotube Growth by Moisture-Controlled Decoupled Chemical Vapor Deposition.

Abstract

Aligned carbon nanotubes (CNTs) have become well known due to their outstanding properties. However, there remains a lack of understanding of the multivariate nature of CNT synthesis, and particularly how to achieve consistent production of aligned CNTs on industrially relevant substrates. This thesis presents a series of techniques for improved manufacturing of CNT forests, including achievement of higher density CNT forests and improved consistency of production.

First, decoupling of the catalyst annealing and CNT nucleation/growth steps is shown to attenuate moisture transients in the CVD system, and results in improvement in average CNT forest height by 21% and reduction of standard deviation by 76%.

Second, it is found that moisture improves the activation of catalyst particles at the beginning of CNT growth. But during later stage of CNT growth a 28% slower growth rate of forest height is led by a moisture level of 370ppm for 10 minute growth. Moreover, deposition of a small amount of carbon onto the reactor wall improves the CNT forest density by a factor of 2~4. X-ray photoelectron spectroscopy and transmission electron microscopy show that the carbon deposition causes the formation of a graphitic layer on catalyst particles, which helps activate particles to achieve higher density CNT forests.

Then, these findings are translated to perform CNT synthesis on metal alloys. CNT forests are synthesized on Haynes 556 alloy. It is shown that air oxidation at 825 ºC and H2/He reduction at 775 ºC is required prior to hydrocarbon exposure in order to bring iron to surface as catalyst, and the presence of moisture during growth is required to achieve aligned forest. Finally, a hybrid filtration material is developed by synthesis of sparse tangled CNTs on sintered porous stainless steel filters. The CNT growth density on the porous substrate is tuned to give an increase in the particle capture efficiency of ~10^5 with a pressure drop of less than 15 psi. An analytical model is established to predict the relationship between pressure drop and flow rate, and rationalize the ideal CNT growth morphology on the filter.