AmericanChemicalSociety.com

Carbon nanotubes (CNTs) are potentially ideal materials for nanoelectronics and alternative energy devices due to their fast charge carrier transport, ultra-high surface area, and great chemical and thermal stabilities. Prototype electronic devices and alternative energy devices incorporating CNTs include field effect transistors, nano-antennas, photovoltaic cells, photoelectrochemical cells, fuel cells, batteries, and electrolytic cells. A bottleneck that currently limits the realization of CNT’s full potential in these applications is the fact that as-synthesized CNTs are mixtures of metallic and semiconducting species. For example, the presence of semiconducting tubes lowers the conductance of CNT electrodes in fuel cells, batteries and electrolytic cells. On the other hand, the presence of metallic tubes in photovoltaic and photoelectrochemical cells increases the carrier recombination and may even short the device if the concentration is greater than the percolation threshold.

The separation of semiconducting and metallic tubes remains a major challenge. Two recent approaches have attracted much interest. First, Krupke et al. constructed a dielectrophoretic device and demonstrated that metallic single-walled carbon nanotubes (SWNTs) are preferentially adsorbed to the electrodes. Second, Zheng et al. showed that ion-exchange chromatography of SWNT/DNA complexes yields fractions of SWNT enriched with semiconducting or metallic tubes at different elution times. A mechanism for the ion-exchange separation has been proposed where the electronic density of states in SWNT screens the charge density on the DNA backbone; thus, the stationary phase of the ion-exchange column interacts with different effective charge density on the SWNT-DNA complexes with different dielectric screening properties.

These promising techniques, however, are still in the discovery phase and the mechanisms are not yet fully understood. Since both techniques aim at separating CNTs based on their drastically different dielectric properties, an in-depth investigation on CNT dielectrics is critically important to understand the CNT responses to external fields and external charge perturbations.

This grant supports our investigation on the dielectric properties of CNTs and the mechanism of dielectric property-based separation of carbon nanotubes. The study on these fundamental questions will lead to efficient separation of metallic from semiconducting tubes and ultimately improve alternative energy devices.  Our latest progress has resulted in a publication in JPCC, as summarized below.

We found that the dielectric response of metallic nanotubes decreases with decreasing length, and becomes indistinguishable from that of semiconducting nanotubes below 200 nm. The results demonstrate that length also needs to be taken into consideration when separating carbon nanotubes.

It is challenging to measure the properties of very short nanotubes and obtain statistically meaningful data. However, using a contactless approach — electrostatic force microscopy (EFM) — we managed for the first time to measure the length effect of dielectric properties for individual nanotubes. It was demonstrated that whereas the dielectric response of metallic nanotubes shows a strong length dependence, the length response of the semiconducting tubes was less evident.

We investigated the source of this dependence of metallic nanotube length. We found that defects at the ends of the nanotubes reduce the electrical conductivity and dielectric response, and because the ends play a proportionally greater role at shorter tube lengths, the dielectric effect is reduced as the nanotube become shorter.

The implications of these results go beyond separation. It is highly relevant in various problems such as the downscaling of channel lengths in carbon nanotube transistors, or the thickness in carbon-nanotube-based nanofiltration membranes and so on.