Optoelectronics with Carbon Nanotubes
Optoelectronics with Carbon Nanotubes
Optoelectronics with Carbon Nanotubes
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Chapter II Methods<br />
1. Materials<br />
One of the most important characteristics of the CNTs for optoelectronic experiments is<br />
their diameter distribution since the E11 transition energy depends inversely on the diameter.<br />
Practically speaking, an appropriate distribution of diameters is required for the emission signal<br />
to fit <strong>with</strong>in the measurement energy window of about 0.55 eV to 1.2 eV. This energy range<br />
approximately corresponds to the diameter range of 0.7 nm to 1.5 nm. The empirical data show<br />
that the transition energy actually varies a great deal depending on the chirality, not just the<br />
diameter 15 , so it is likely that the practical diameter range extends further.<br />
Of the several standard ways to determine the CNT diameter, transmission electron<br />
microscopy (TEM) is the most direct and accurate, but it cannot be conducted on a wafer surface.<br />
Atomic force microscopy (AFM) has a measurement error of at least 0.2 nm (about 0.5 nm for<br />
the AFM system that was available to us), too large a fraction of a CNT diameter to measure it to<br />
any useful precision. A more promising approach is to use Raman spectroscopy to determine the<br />
phonon frequency of the tube’s radial breathing mode (RBM), which is inversely proportional to<br />
the diameter 37 . However, Raman signals, especially the RBM signal, are very weak from a CNT<br />
on a wafer surface (in contrast to suspended tubes), unless the laser excitation energy happens to<br />
be in resonance <strong>with</strong> the CNT’s E33 absorption energy. This means that the excitation energy<br />
needs to be tuned while searching for Raman signals, which is a very slow and labor-intensive<br />
process. Given such limitations, it is not practical to determine individual CNT diameters in<br />
advance, so the samples were chosen on the basis of their statistically-known diameter<br />
distribution. Attempts to measure individual diameters by Raman spectroscopy afterwards were<br />
not always successful.<br />
Several reliable methods have been developed to produce different types of carbon<br />
nanotubes (CNTs). For this work, the tubes used were grown using arc-discharge, laser-ablation<br />
and chemical-vapor deposition (CVD) methods. Of the three methods, laser ablation and arc-<br />
discharge produce bulk bundles of nanotubes in quantities of milligrams and even grams, while<br />
the CVD method grows individual nanotubes directly on a SiO2 surface from catalyst particles.<br />
28