Top-Down Patterning and Self-Assembly for Regular Arrays of Semiconducting Single-Walled Carbon Nanotubes

摘要

Single-walled carbon nanotubes (SWNTs) are promising nano-electronic materials, but they face one primary challenge to their integration into electronics: the variability of their structure and their sometimes semiconducting or metallic nature. Whether through their diameter and chirality ("twist") or their density and orientation, control over the type of carbon nanotubes and where they are placed is key to eventual integration into practical electronics. Several approaches have been proposed to control these disparate aspects. They can be divided into dry and wet methods and according to whether sorting and alignment occurs during growth or post-growth. Dry methods rely on either the preferential growth of semiconducting tubes over metallic ones (governed by chirality), or on the removal of metallic tubes from the tubes grown on substrates. Positional control can be achieved by aligned growth, either through directional flow during the chemical vapor deposition (CVD) process or along crystal axes. These techniques are nevertheless limited in both density and purity of semiconducting SWNTs (s-SWNTs), even with new techniques to transfer and stack CVD-grown tubes and to remove the remaining metallic tubes. Wet approaches rely on solution phase separation of the grown tubes, and have produced some very pure solutions of s-SWNTs, and in some cases, even examples of controlled chirality.However, fewer processes exist for control over the positioning of the SWNTs during deposition from solution. Dielectrophoresis is one technique that has been repeatedly and reproducibly used for the assembly of SWNTs, although it is limited in density to 50 tubes per micrometer. A self-assembly approach was used to place SWNTs in patterned trenches, but the SWNT alignment within the trench was poor and the process was intended to deposit one carbon nanotube per trench.The Langmuir-Blodgett and Lang-muir-Schaeffer methods have been used to align SWNTs on a substrate with full coverage, but the proximity of the SWNTs resulting from these methods is too close to avoid multilayers, which give screening effects between tubes. Our process results in 2D nanostructures consisting of individual carbon nanotubes (CNTs) or thin bundles that, while still subject to the screening that lowers performance in CNTs in close proximity, show substantially better performance on a per-CNT basis than stacked layers of nanotubes.